diff --git a/.github/workflows/check-help-docs.yml b/.github/workflows/check-help-docs.yml new file mode 100644 index 0000000..02e6437 --- /dev/null +++ b/.github/workflows/check-help-docs.yml @@ -0,0 +1,26 @@ +name: Check Help Docs + +# Validates the NLP++ Help markdown on every PR / push that touches it: +# fails on broken .md links, broken image links, or table/index drift. +# (Runs on windows-latest because Help/check-help-docs.ps1 uses Windows path logic.) + +on: + pull_request: + paths: + - 'Help/markdown/**' + - 'Help/check-help-docs.ps1' + - '.github/workflows/check-help-docs.yml' + push: + branches: [ main ] + paths: + - 'Help/markdown/**' + - 'Help/check-help-docs.ps1' + +jobs: + check: + runs-on: windows-latest + steps: + - uses: actions/checkout@v4 + - name: Validate help docs (links, images, table/index drift) + shell: pwsh + run: ./Help/check-help-docs.ps1 diff --git a/Help/check-help-docs.ps1 b/Help/check-help-docs.ps1 index 046f153..6bf8469 100644 --- a/Help/check-help-docs.ps1 +++ b/Help/check-help-docs.ps1 @@ -8,10 +8,13 @@ is reported only as INFORMATION, never as an error to auto-fix. Reports: - 1. Broken links - any ](*.md) reference in the corpus that does not resolve. - 2. Orphaned pages - *.md files not referenced from index.md (legacy candidates). - 3. Table/index drift- pages linked from a Table_of_* but absent from index.md. - 4. Engine coverage - (info) engine function/action names with no help page. + 1. Broken links - any ](*.md) reference in the corpus that does not resolve. + 1b. Broken images - any ![..](img) reference that does not resolve. + 2. Orphaned pages - *.md files not referenced from index.md (legacy candidates). + 3. Table/index drift- pages linked from a Table_of_* but absent from index.md. + 4. Engine coverage - (info) engine function/action names with no help page. + 5. Empty stubs - (-Stubs) current pages with empty Returns/Example/etc. + 6. Orphaned images - (-Images) image files on disk referenced by no page. .PARAMETER HelpDir Path to the Help/markdown directory. Defaults to ./markdown relative to this script. @@ -26,7 +29,8 @@ param( [string]$HelpDir = (Join-Path $PSScriptRoot 'markdown'), [string]$EnginePath = 'c:\git\nlp-engine\lite', - [switch]$Stubs # also report current pages with empty Returns/Example/Remarks sections + [switch]$Stubs, # also report current pages with empty Returns/Example/Remarks sections + [switch]$Images # also list orphaned image assets (on disk but referenced by no page) ) if (-not (Test-Path $HelpDir)) { Write-Error "HelpDir not found: $HelpDir"; exit 2 } @@ -54,6 +58,23 @@ Write-Host "== 1. Broken .md links ($($broken.Count)) ==" -ForegroundColor Cyan $broken | Sort-Object | ForEach-Object { Write-Host " $_" } $issues += $broken.Count +# --- 1b. Broken image links --------------------------------------------------- +$brokenImg = @() +foreach ($f in $allMd) { + $txt = Get-Content $f.FullName -Raw + foreach ($m in [regex]::Matches($txt, '!\[[^\]]*\]\(([^)]+)\)')) { + $tgt = $m.Groups[1].Value + if ($tgt -match '^https?://') { continue } + $tgt = ([uri]::UnescapeDataString($tgt) -split '#')[0].Trim() + if (-not $tgt) { continue } + if (-not (Test-Path (Join-Path $f.DirectoryName $tgt))) { + $brokenImg += '{0} -> {1}' -f $f.FullName.Substring($HelpDir.Length+1), $tgt + } + } +} +Write-Host "`n== 1b. Broken image links ($($brokenImg.Count)) ==" -ForegroundColor Cyan +$brokenImg | Sort-Object | ForEach-Object { Write-Host " $_" } + # --- 2. Orphaned flat pages (not referenced from index.md) -------------------- $flat = $allMd | Where-Object { $_.DirectoryName -eq $HelpDir -and $_.Name -ne 'index.md' } $orphans = $flat | Where-Object { $indexTxt -notmatch [regex]::Escape('(' + $_.Name + ')') } | @@ -126,7 +147,29 @@ if ($Stubs) { $rows | Sort-Object Page | ForEach-Object { Write-Host (" {0,-22} {1}" -f $_.Page, $_.Empty) } } +# --- 6. Orphaned image assets (opt-in via -Images) ---------------------------- +if ($Images) { + $assetRoot = Split-Path $HelpDir # Help/ + $imgFiles = Get-ChildItem -Path $assetRoot -Recurse -File -Include *.gif,*.png,*.jpg,*.jpeg + $refset = @{} + foreach ($f in $allMd) { + $txt = Get-Content $f.FullName -Raw + foreach ($m in [regex]::Matches($txt, '!\[[^\]]*\]\(([^)]+)\)')) { + $tgt = $m.Groups[1].Value + if ($tgt -match '^https?://') { continue } + $tgt = ([uri]::UnescapeDataString($tgt) -split '#')[0].Trim() + if (-not $tgt) { continue } + try { $r = [System.IO.Path]::GetFullPath((Join-Path $f.DirectoryName $tgt)) } catch { continue } + $refset[$r.ToLower()] = $true + } + } + $orphanImg = $imgFiles | Where-Object { -not $refset.ContainsKey($_.FullName.ToLower()) } + Write-Host "`n== 6. Orphaned image assets: on disk but referenced by no page ($($orphanImg.Count)) ==" -ForegroundColor Cyan + Write-Host " (informational; paths relative to Help/)" + $orphanImg | ForEach-Object { Write-Host (" " + $_.FullName.Substring($assetRoot.Length+1)) } +} + Write-Host "`n--- summary ---" -ForegroundColor Green -Write-Host ("broken links: {0} orphaned pages: {1} table/index drift: {2}" -f $broken.Count, $orphans.Count, $tableDrift.Count) -# Exit non-zero only on hard errors (broken links / table-index drift), so it can gate CI. -if (($broken.Count + $tableDrift.Count) -gt 0) { exit 1 } else { exit 0 } +Write-Host ("broken links: {0} broken images: {1} orphaned pages: {2} table/index drift: {3}" -f $broken.Count, $brokenImg.Count, $orphans.Count, $tableDrift.Count) +# Exit non-zero on hard errors (broken .md links / broken images / table-index drift) so it can gate CI. +if (($broken.Count + $brokenImg.Count + $tableDrift.Count) -gt 0) { exit 1 } else { exit 0 } diff --git a/Help/helps/$allcaps.htm b/Help/helps/$allcaps.htm deleted file mode 100644 index e970c84..0000000 --- a/Help/helps/$allcaps.htm +++ /dev/null @@ -1,89 +0,0 @@ - - - - - - - - - - -$allcaps - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$allcaps

- -

Purpose

- -

Check if the token underlying the node is all uppercase.

- -

Syntax

- -

returnedBoolean = variableType("$allcaps")

- -

returnedBoolean - type: int
-variableType - type: N

- -

Returns

- -

Returns 1 if the token is all - uppercase and 0 if not. If multiple - words (even if all words are capitalized) returns 0.

- -

Remarks

- -

None.

- -

Example

- -

 

- -

See Also

- -

$cap, $lowercase, - $mixcap, $uppercase, - Special Variables

- -

 

- - - - diff --git a/Help/helps/$anaspath.htm b/Help/helps/$anaspath.htm deleted file mode 100644 index a0224d0..0000000 --- a/Help/helps/$anaspath.htm +++ /dev/null @@ -1,84 +0,0 @@ - - - - - - - - - - -$anaspath - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$anaspath

- -

Purpose

- -

Get the parent folder for the analyzer folders.

- -

Syntax

- -

returnedString = variableType("$anaspath")

- -

returnedString - type: str
-variableType - type: G

- -

Returns

- -

Remarks

- -

Example

- -

$anaspath returns "D\apps\Resume" assuming the app directory is "D\apps\".

- -

See Also

- -

$apppath, $kbpath, $input, - $inputhead, $inputname, - $inputtail, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$apppath.htm b/Help/helps/$apppath.htm deleted file mode 100644 index ef69fe3..0000000 --- a/Help/helps/$apppath.htm +++ /dev/null @@ -1,82 +0,0 @@ - - - - - - - - - - -$apppath - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$apppath

- -

Purpose

- -

Get the base directory path for the current application.

- -

Syntax

- -

returnedString = variableType("$apppath")

- -

returnedString - type: str
-variableType - type: G

- -

Returns

- -

Remarks

- -

Example

- -

"D:\dev\apps\Resume".

- -

See Also

- -

$inputpath, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$cap.htm b/Help/helps/$cap.htm deleted file mode 100644 index 7bf6416..0000000 --- a/Help/helps/$cap.htm +++ /dev/null @@ -1,114 +0,0 @@ - - - - - - - - - - -$cap - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$cap

- -

Purpose

- -

Check if the token underlying the node is a capitalized word.

- -

Syntax

- -

returnedBoolean = variableType("$cap")

- -

returnedBoolean - type: int
-variableType - type: N

- -

Returns

- -

Returns 1 if the token is a - capitalized word and 0 if not.

- -

Remarks

- - - - - - -
-

WARNING: $cap - is not identical to the @PRE action cap( ). $cap will traverse down to the leafiest node - it can find, in order to perform its check. The @PRE action only chains - down if the [s] flag is present in the associated phrase element(s). Also, - $cap is not stopped by a [base] attribute in a node.

- -

Example

- -

See Also

- -

$allcaps, $lowercase, - $mixcap, $uppercase, - Special Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/$end.htm b/Help/helps/$end.htm deleted file mode 100644 index 957fbf9..0000000 --- a/Help/helps/$end.htm +++ /dev/null @@ -1,109 +0,0 @@ - - - - - - - - - - -$end - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$end

- -

Purpose

- -

Check if the referenced node has a right sibling in the - parse tree.

- -

Syntax

- -

returnedBoolean = variableType("$end"[, number])

- -

returnedBoolean - type: bool
-variableType - type: N or X
-number - type: int

- -

Returns

- -

Evaluates to 0 if the node has - a right sibling and 1 if the referenced - node does not have a right sibling.

- -

Remarks

- -

Example

- -

@RULES

- -

@@RULES

- -

See Also

- -

$start, Special - Variables

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/$exists.htm b/Help/helps/$exists.htm deleted file mode 100644 index 111c37b..0000000 --- a/Help/helps/$exists.htm +++ /dev/null @@ -1,104 +0,0 @@ - - - - - - - - - - -$exists - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$exists

- -

Purpose

- -

Check if the node exists in the parse tree.

- -

Syntax

- -

returnedBoolean = variableType("$exists"[, number])

- -

returnedBoolean - type: bool
-variableType - type: N
-number - type: int

- -

Returns

- -

Returns 1 if the node exists - and 0 if the node does not exist.

- -

Remarks

- -

Example

- -

@RULES

- -

@@RULES

- -

See Also

- -

$unknown, Special - Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/$input.htm b/Help/helps/$input.htm deleted file mode 100644 index 085d7f5..0000000 --- a/Help/helps/$input.htm +++ /dev/null @@ -1,84 +0,0 @@ - - - - - - - - - - -$input - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$input

- -

Purpose

- -

Get the full input path for the file.

- -

Syntax

- -

returnedString = variableType("$input")

- -

returnedString - type: str
-variableType - type: G

- -

Returns

- -

Remarks

- -

Example

- -

$input returns "D:\apps\Resume\input\Dev1\rez.txt" if the - input file is "rez.txt".

- -

See Also

- -

$inputhead, $inputname, - $inputpath, $inputtail, - Special Variables

- -

 

- - - - diff --git a/Help/helps/$inputhead.htm b/Help/helps/$inputhead.htm deleted file mode 100644 index 966abb0..0000000 --- a/Help/helps/$inputhead.htm +++ /dev/null @@ -1,84 +0,0 @@ - - - - - - - - - - -$inputhead - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$inputhead

- -

Purpose

- -

Get the head (name) of the input file.

- -

Syntax

- -

returnedString = variableType("$inputhead")

- -

returnedString - type: str
-variableType - type: G

- -

Returns

- -

Remarks

- -

Example

- -

$inputhead returns "rez" if the input file is "rez.txt".

- -

See Also

- -

$apppath, $input, - $inputname, $inputpath, - $inputtail, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$inputname.htm b/Help/helps/$inputname.htm deleted file mode 100644 index 8f5517f..0000000 --- a/Help/helps/$inputname.htm +++ /dev/null @@ -1,84 +0,0 @@ - - - - - - - - - - -$inputname - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$inputname

- -

Purpose

- -

Get the name of the input file.

- -

Syntax

- -

returnedString = variableType("$inputname")

- -

returnedString - type: str
-variableType - type: G

- -

Returns

- -

Remarks

- -

Example

- -

$inputname returns "rez.txt" if the input file is "rez.txt".

- -

See Also

- -

$apppath, $input, - $inputhead, $inputpath, - $inputtail, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$inputparent.htm b/Help/helps/$inputparent.htm deleted file mode 100644 index f305660..0000000 --- a/Help/helps/$inputparent.htm +++ /dev/null @@ -1,84 +0,0 @@ - - - - - - - - - - -$inputparent - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$inputparent

- -

Purpose

- -

Get the name of the parent directory of the input file.

- -

Syntax

- -

returnedString = variableType("$inputparent")

- -

returnedString - type: str
-variableType - type: G

- -

Returns

- -

Remarks

- -

Example

- -

$inputparent returns "Dev1" assuming input file is "D\apps\Resume\input\Dev1\text.txt".

- -

See Also

- -

$apppath, $input, - $inputhead, $inputname, - $inputtail, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$inputpath.htm b/Help/helps/$inputpath.htm deleted file mode 100644 index 52fb545..0000000 --- a/Help/helps/$inputpath.htm +++ /dev/null @@ -1,85 +0,0 @@ - - - - - - - - - - -$inputpath - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$inputpath

- -

Purpose

- -

Get the path of the input file.

- -

Syntax

- -

returnedString = variableType("$inputpath")

- -

returnedString - type: str
-variableType - type: G

- -

Returns

- -

Remarks

- -

Example

- -

$inputpath returns "D\apps\Resume\input\Dev1" assuming input - file "rez.txt" resides in Dev1 directory.

- -

See Also

- -

$apppath, $input, - $inputhead, $inputname, - $inputtail, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$inputtail.htm b/Help/helps/$inputtail.htm deleted file mode 100644 index f2a3575..0000000 --- a/Help/helps/$inputtail.htm +++ /dev/null @@ -1,84 +0,0 @@ - - - - - - - - - - -$inputtail - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$inputtail

- -

Purpose

- -

Get the tail (extension type) of the input file.

- -

Syntax

- -

returnedString = variableType("$inputtail")

- -

returnedString - type: str
-variableType - type: G

- -

Returns

- -

Remarks

- -

Example

- -

$inputtail returns "txt" if the input file is "rez.txt".

- -

See Also

- -

$apppath, $input, - $inputhead, $inputname, - $inputpath, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$isdirrun.htm b/Help/helps/$isdirrun.htm deleted file mode 100644 index e577fea..0000000 --- a/Help/helps/$isdirrun.htm +++ /dev/null @@ -1,82 +0,0 @@ - - - - - - - - - - -$isdirrun - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$isdirrun

- -

Purpose

- -

Check if the engine is processing a directory of files instead of an individual file. Returns 1 if it is analyzing a directory, else 0.

- -

Syntax

- -

returnedBoolean = variableType("$isdirrun")

- -

returnedBoolean - type: int
-variableType - type: G

- -

Returns

- -

Returns 1 if processing a directory, 0 if not.

- -

Remarks

- -

Example

- -

See Also

- -

$isfirstfile, $islastfile, $inputpath, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$isfirstfile.htm b/Help/helps/$isfirstfile.htm deleted file mode 100644 index f56f00f..0000000 --- a/Help/helps/$isfirstfile.htm +++ /dev/null @@ -1,82 +0,0 @@ - - - - - - - - - - -$isfirstfile - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$isfirstfile

- -

Purpose

- -

Checks to see if the current file is the first file being analzyed in a directory.

- -

Syntax

- -

returnedBoolean = variableType("$isfirstfile")

- -

returnedBoolean - type: bool
-variableType - type: G

- -

Returns

- -

Returns 1 if the current file is the first file, 0 if not.

- -

Remarks

- -

Example

- -

See Also

- -

$isdirrun, $islastfile, $inputpath, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$islastfile.htm b/Help/helps/$islastfile.htm deleted file mode 100644 index 489adf0..0000000 --- a/Help/helps/$islastfile.htm +++ /dev/null @@ -1,82 +0,0 @@ - - - - - - - - - - -$islastfile - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$islastfile

- -

Purpose

- -

Checks to see if the current file is the last file being analzyed in a directory.

- -

Syntax

- -

returnedBoolean = variableType("$islastfile")

- -

returnedBoolean - type: int
-variableType - type: G

- -

Returns

- -

Returns 1 if the current file is the last file, 0 if not.

- -

Remarks

- -

Example

- -

See Also

- -

$isdirrun, $isfirstfile, $inputpath, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$kbpath.htm b/Help/helps/$kbpath.htm deleted file mode 100644 index d658c35..0000000 --- a/Help/helps/$kbpath.htm +++ /dev/null @@ -1,85 +0,0 @@ - - - - - - - - - - -$kbpath - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$kbpath

- -

Purpose

- -

Get the path for the kb / user.

- -

Syntax

- -

returnedString = variableType("$kbpath")

- -

returnedString - type: str
-variableType - type: G

- -

Returns

- -

Remarks

- -

Example

- -

$kbpath returns "D\apps\Resume\input\Dev1\kb\user" assuming the app directory - is "D\apps\Resume\input\Dev1" resides in Dev1 directory.

- -

See Also

- -

$apppath, $input, - $inputhead, $inputname, - $inputtail, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$length.htm b/Help/helps/$length.htm deleted file mode 100644 index 15dcb79..0000000 --- a/Help/helps/$length.htm +++ /dev/null @@ -1,100 +0,0 @@ - - - - - - - - - - -$length - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$length

- -

Purpose

- -

Get the length of node's text.

- -

Syntax

- -

returnedNumber = variableType("$length"[, number])

- -

returnedNumber - type: int
-variableType - type: N or X
-number - type: int

- -

Returns

- -

Remarks

- -

Example

- -

@RULES

- -

@@RULES

- -

See Also

- -

Special Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/$lowercase.htm b/Help/helps/$lowercase.htm deleted file mode 100644 index 9bf10f2..0000000 --- a/Help/helps/$lowercase.htm +++ /dev/null @@ -1,114 +0,0 @@ - - - - - - - - - - -$lowercase - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$lowercase

- -

Purpose

- -

Check if the token underlying the node is all lowercase.

- -

Syntax

- -

returnedBoolean = variableType("$lowercase")

- -

returnedBoolean - type: int
-variableType - type: N

- -

Returns

- -

Returns 1 if the token is all - lowercase and 0 if not. If multiple - words (even if all words are lowercase) returns 0.

- -

Remarks

- - - - - - -
-

WARNING: $lowercase is not identical to - the @PRE action lowercase( ). - $lowercase will traverse down to the leafiest node it can find, in order - to perform its check. The @PRE action only chains down if the [s] flag - is present in the associated phrase element(s). Also, $lowercase is not - stopped by a [base] attribute in a node.

- -

Example

- -

See Also

- -

$allcaps, $cap, $mixcap, $uppercase, - Special Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/$mixcap.htm b/Help/helps/$mixcap.htm deleted file mode 100644 index 9b647e0..0000000 --- a/Help/helps/$mixcap.htm +++ /dev/null @@ -1,89 +0,0 @@ - - - - - - - - - - -$mixcap - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$mixcap

- -

Purpose

- -

Check if the token underlying the node is a mixed-capitalized word. -

- -

Syntax

- -

returnedBoolean = variableType("$mixcap")

- -

returnedBoolean - type: int
-variableType - type: N

- -

Returns

- -

Returns 1 if the token is a - mixed-capitalized word and 0 if - not.

- -

Remarks

- -

Example

- -

MIchigan, or abcD

- -

See Also

- -

$allcaps, $cap, $lowercase, $uppercase, - Special Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/$oend.htm b/Help/helps/$oend.htm deleted file mode 100644 index 97c0e7c..0000000 --- a/Help/helps/$oend.htm +++ /dev/null @@ -1,98 +0,0 @@ - - - - - - - - - - -$oend - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$oend

- -

Purpose

- -

End offset of the referenced node in the input text.

- -

Syntax

- -

variableType("$oend"[, number])

- -

variableType - type: N or X
-number - type: int

- -

Returns

- -

Remarks

- -

Example

- -

@RULES

- -

@@RULES

- -

See Also

- -

$ostart, Special - Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/$ostart.htm b/Help/helps/$ostart.htm deleted file mode 100644 index e12ead8..0000000 --- a/Help/helps/$ostart.htm +++ /dev/null @@ -1,98 +0,0 @@ - - - - - - - - - - -$ostart - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$ostart

- -

Purpose

- -

Start offset of the referenced node in the input text.

- -

Syntax

- -

variableType("$ostart"[, number])

- -

variableType - type: N or X
-number - type: int

- -

Returns

- -

Remarks

- -

Example

- -

@RULES

- -

@@RULES

- -

See Also

- -

$oend, Special - Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/$raw.htm b/Help/helps/$raw.htm deleted file mode 100644 index e868146..0000000 --- a/Help/helps/$raw.htm +++ /dev/null @@ -1,104 +0,0 @@ - - - - - - - - - - -$raw - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$raw

- -

Purpose

- -

Fetch 'cleaned up' text covered by the node from the text buffer. Leading - and trailing spaces are removed and separators are converted to single - space.

- -

Syntax

- -

returnedString = variableType("$raw"[, number])

- -

returnedString - type: str
-variableType - type: N or X
-number - type: int

- -

Returns

- -

Remarks

- -

Example

- -

@RULES

- -

@@RULES

- -

See Also

- -

$treetext, $treeraw, -  $text, - $xmltext, phraseraw, - phrasetext, Special - Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/$start.htm b/Help/helps/$start.htm deleted file mode 100644 index 98bf0d4..0000000 --- a/Help/helps/$start.htm +++ /dev/null @@ -1,103 +0,0 @@ - - - - - - - - - - -$start - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$start

- -

Purpose

- -

Check if the referenced node has a left sibling in the parse - tree.

- -

Syntax

- -

returnedBoolean = variableType("$start"[, number])

- -

returnedBoolean - type: bool
-variableType - type: N or X
-number - type: int

- -

Returns

- -

Evaluates to 0 if the node has - a left sibling and 1 if the referenced - node does not have a left sibling.

- -

Remarks

- -

Example

- -

@RULES

- -

@@RULES

- -

See Also

- -

$end, Special - Variables

- -

 

- - - - diff --git a/Help/helps/$text.htm b/Help/helps/$text.htm deleted file mode 100644 index 5be0869..0000000 --- a/Help/helps/$text.htm +++ /dev/null @@ -1,105 +0,0 @@ - - - - - - - - - - -$text - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$text

- -

Purpose

- -

Fetch the 'cleaned up' text covered by the node. Leading and trailing - spaces are removed and separators are converted to single space.

- -

Syntax

- -

returnedString = variableType("$text"[, number])

- -

returnedString - type: str
-variableType - type: N or X
-number - type: int

- -

Returns

- -

Remarks

- -

Uses the original text buffer, rather than the subtree under the node, - in order to gather text.

- -

Example

- -

@RULES

- -

@@RULES

- -

See Also

- -

$treetext, $raw, - $treeraw, , $xmltext, - phraseraw, phrasetext, - Special Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/$treeraw.htm b/Help/helps/$treeraw.htm deleted file mode 100644 index 93f7484..0000000 --- a/Help/helps/$treeraw.htm +++ /dev/null @@ -1,102 +0,0 @@ - - - - - - - - - - -$treeraw - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$treeraw

- -

Purpose

- -

Fetch the text covered by the node from the parse tree. Leading and - trailing spaces are not removed and separators are not converted to single - space.

- -

Syntax

- -

returnedString = variableType("$treeraw"[, number])

- -

returnedString - type: str
-variableType - type: N or X
-number - type: int

- -

Returns

- -

Remarks

- -

 Analogous to $treetext.

- -

Example

- -

@RULES

- -

@@RULES

- -

See Also

- -

$treetext, $text, - $raw, , $xmltext, phraseraw, phrasetext, - Special Variables

- -

 

- - - - diff --git a/Help/helps/$treetext.htm b/Help/helps/$treetext.htm deleted file mode 100644 index b023b59..0000000 --- a/Help/helps/$treetext.htm +++ /dev/null @@ -1,122 +0,0 @@ - - - - - - - - - - -$treetext - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$treetext

- -

Purpose

- -

Fetch 'cleaned up' text gleaned from a node's subtree. Leading and trailing - spaces are removed and separators are converted to single space.

- -

Syntax

- -

returnedString = variableType("$treetext"[, number])

- -

returnedString - type: str
-variableType - type: N or X
-number - type: int

- -

Returns

- -

Remarks

- -

Uses the parse tree as input, not text buffer. $treetext is analogous - to $text. The difference between $treetext and $text is $text returns - text from the input text buffer while $treetext returns text from the - parse tree. If you have removed some nodes from a tree with an excise(), - $treetext returns the text minus the excised elements, while $text returns - everything in the input text buffer, including the text under excised - nodes.

- -

Example

- -

@POST
-
-
-@@POST

- -

@RULES

- -

@@RULES

- -

See Also

- -

$text, $treeraw, - $raw, , $xmltext, phraseraw, phrasetext, - Special Variables

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/$unknown.htm b/Help/helps/$unknown.htm deleted file mode 100644 index c01a181..0000000 --- a/Help/helps/$unknown.htm +++ /dev/null @@ -1,139 +0,0 @@ - - - - - - - - - - -$unknown - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$unknown

- -

Purpose

- -

Check if the token underlying a node is an unknown word. -

- -

Syntax

- -

returnedBoolean = variableType("$unknown"[, number])

- -

returnedBoolean - type: bool
-variableType - type: N
-number - type: int

- -

Returns

- -

Returns 1 if the token is an - unknown word and 0 if the token - is a known word.

- -

Remarks

- -

Requires a lookup( ) CODE Action - prior to any use of this special variable.

- - - - - - -
-

WARNING: - This special variable is not identical to the @PRE - action, unknown(). $unknown - will traverse down to the leafiest node it can find, in order to perform - its check. The @PRE action only chains down if the [s] flag is present - in the associated phrase element(s). Also, $unknown is not stopped by - a [base] attribute in a node.

- -

Example

- -

@RULES

- -

@@RULES

- -

See Also

- -

$exists, Special - Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/$uppercase.htm b/Help/helps/$uppercase.htm deleted file mode 100644 index 2ce7361..0000000 --- a/Help/helps/$uppercase.htm +++ /dev/null @@ -1,114 +0,0 @@ - - - - - - - - - - -$uppercase - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$uppercase

- -

Purpose

- -

Check if the token underlying the node is all uppercase.

- -

Syntax

- -

returnedBoolean = variableType("$uppercase")

- -

returnedBoolean - type: int
-variableType - type: N

- -

Returns

- -

Returns 1 if the token is all - uppercase and 0 if not. If multiple - words (even if all words are capitalized) returns 0.

- -

Remarks

- - - - - - -
-

WARNING: $uppercase - is not identical to the @PRE action uppercase( - ). $uppercase will traverse down to the leafiest node it can find, in - order to perform its check. The @PRE action only chains down if the [s] - flag is present in the associated phrase element(s). Also, $uppercase - is not stopped by a [base] attribute in a node.

- -

Example

- -

See Also

- -

$allcaps, $cap, $lowercase, $mixcap, - Special Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/$xmltext.htm b/Help/helps/$xmltext.htm deleted file mode 100644 index 3055893..0000000 --- a/Help/helps/$xmltext.htm +++ /dev/null @@ -1,104 +0,0 @@ - - - - - - - - - - -$xmltext - - - - - - - - - - - - - - - - - - - - - - - - - - - -

$xmltext

- -

Purpose

- -

Fetch 'cleaned up' text covered by the node from the text buffer, converting - special characters to HTML and XML. Leading and trailing spaces are removed - and separators are converted to single space.

- -

Syntax

- -

returnedString = variableType("$xmltext"[, number])

- -

returnedString - type: str
-variableType - type: N or X
-number - type: int

- -

Returns

- -

Remarks

- -

Example

- -

 "<" is converted to "&lt;".

- -

@RULES

- -

@@RULES

- -

See Also

- -

$treetext, $treeraw, -  $text, - Special Variables

- -

 

- -

 

- - - - diff --git a/Help/helps/@@.htm b/Help/helps/@@.htm deleted file mode 100644 index 674b2d1..0000000 --- a/Help/helps/@@.htm +++ /dev/null @@ -1,49 +0,0 @@ - - - - - - - - - - -@@ - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@@

- -

A pass file marker used to indicate the end of a rule.  

- -

See Also

- -

About Pass Files

- - - - diff --git a/Help/helps/@@CHECK.htm b/Help/helps/@@CHECK.htm deleted file mode 100644 index 74d3ad0..0000000 --- a/Help/helps/@@CHECK.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -@@CHECK - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@@CHECK

- -

A region marker that indicates - the end of a CHECK Region.

- -

 

- -

See Also

- -

CHECK - Region

- -

CHECK - Action

- - - - diff --git a/Help/helps/@@CODE.htm b/Help/helps/@@CODE.htm deleted file mode 100644 index 25ef1a3..0000000 --- a/Help/helps/@@CODE.htm +++ /dev/null @@ -1,58 +0,0 @@ - - - - - - - - - - -@@CODE - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@@CODE

- -

A region marker that indicates - the end of the CODE Region.

- -

 

- -

See Also

- -

CODE - Region

- -

CODE - Action

- -

 

- - - - diff --git a/Help/helps/@@MULTI.htm b/Help/helps/@@MULTI.htm deleted file mode 100644 index a2dfd61..0000000 --- a/Help/helps/@@MULTI.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -@@MULTI - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@@MULTI

- -

An optional end marker that indicates the termination of the @MULTI - Selector specification.

- -

See Also

- -

MULTI - Selector

- -

SELECT - Region

- -

 

- - - - diff --git a/Help/helps/@@NODES.htm b/Help/helps/@@NODES.htm deleted file mode 100644 index 9a711b7..0000000 --- a/Help/helps/@@NODES.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -@@NODES - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@@NODES

- -

An optional end marker that indicates the termination of the @NODES - Selector specification.

- -

See Also

- -

NODES - Selector

- -

SELECT - Region

- -

 

- - - - diff --git a/Help/helps/@@PATH.htm b/Help/helps/@@PATH.htm deleted file mode 100644 index 8c8b20f..0000000 --- a/Help/helps/@@PATH.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -@@PATH - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@@PATH

- -

An optional end marker that indicates the termination of the @PATH Selector - specification.

- -

See Also

- -

PATH - Selector

- -

SELECT - Region

- -

 

- - - - diff --git a/Help/helps/@@POST.htm b/Help/helps/@@POST.htm deleted file mode 100644 index a01c778..0000000 --- a/Help/helps/@@POST.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -@@POST - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@@POST

- -

A region marker that indicates - the end of a POST Region.

- -

 

- -

See Also

- -

POST - Region

- -

POST - Action

- - - - diff --git a/Help/helps/@@PRE.htm b/Help/helps/@@PRE.htm deleted file mode 100644 index de84acd..0000000 --- a/Help/helps/@@PRE.htm +++ /dev/null @@ -1,53 +0,0 @@ - - - - - - - - - - -@@PRE - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@@PRE

- -

A region marker that indicates - the end of a PRE Region.

- -

See Also

- -

PRE Region

- -

PRE - Action

- - - - diff --git a/Help/helps/@@RECURSE.htm b/Help/helps/@@RECURSE.htm deleted file mode 100644 index 5c22c94..0000000 --- a/Help/helps/@@RECURSE.htm +++ /dev/null @@ -1,49 +0,0 @@ - - - - - - - - - - -@@RECURSE - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@@RECURSE

- -

A region marker that indicates the end of a RECURSE Region.

- -

See Also

- -

RECURSE Region

- - - - diff --git a/Help/helps/@@RULES.htm b/Help/helps/@@RULES.htm deleted file mode 100644 index 4195767..0000000 --- a/Help/helps/@@RULES.htm +++ /dev/null @@ -1,49 +0,0 @@ - - - - - - - - - - -@@RULES - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@@RULES

- -

A region marker that indicates the end of a RULES Region.

- -

See Also

- -

RULES Region

- - - - diff --git a/Help/helps/@@SELECT.htm b/Help/helps/@@SELECT.htm deleted file mode 100644 index 6663727..0000000 --- a/Help/helps/@@SELECT.htm +++ /dev/null @@ -1,49 +0,0 @@ - - - - - - - - - - -@@SELECT - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@@SELECT

- -

A region marker that indicates the end of a SELECT Region.

- -

See Also

- -

SELECT Region

- - - - diff --git a/Help/helps/@CHECK.htm b/Help/helps/@CHECK.htm deleted file mode 100644 index 25dd08f..0000000 --- a/Help/helps/@CHECK.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -@CHECK - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@CHECK

- -

A region marker that indicates - the start of a CHECK Region.

- -

 

- -

See Also

- -

CHECK - Region

- -

CHECK - Action

- - - - diff --git a/Help/helps/@CODE.htm b/Help/helps/@CODE.htm deleted file mode 100644 index 495c8e3..0000000 --- a/Help/helps/@CODE.htm +++ /dev/null @@ -1,58 +0,0 @@ - - - - - - - - - - -@CODE - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@CODE

- -

A region marker that indicates - the start of the CODE Region.

- -

 

- -

See Also

- -

CODE - Region

- -

CODE - Action

- -

 

- - - - diff --git a/Help/helps/@DECL.htm b/Help/helps/@DECL.htm deleted file mode 100644 index 0096818..0000000 --- a/Help/helps/@DECL.htm +++ /dev/null @@ -1,58 +0,0 @@ - - - - - - - - - - -@DECL - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@DECL

- -

A region marker that indicates - the start of the DECL Region.

- -

 

- -

See Also

- -

DECL - Region

- -

Declare - Zone

- -

 

- - - - diff --git a/Help/helps/@MULTI.htm b/Help/helps/@MULTI.htm deleted file mode 100644 index 9308168..0000000 --- a/Help/helps/@MULTI.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -@MULTI - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@MULTI

- -

A selector used in the Context Zone of the pass file.  The - Context Zone specifies methods for selecting nodes in the parse tree.

- -

See Also

- -

@MULTI - Selector

- -

SELECT - Region

- -

 

- - - - diff --git a/Help/helps/@NODES.htm b/Help/helps/@NODES.htm deleted file mode 100644 index 079a4aa..0000000 --- a/Help/helps/@NODES.htm +++ /dev/null @@ -1,54 +0,0 @@ - - - - - - - - - - -@NODES - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@NODES

- -

A selector used in the Context Zone of the pass file.  The - Context Zone specifies methods for selecting nodes in the parse tree.

- -

See Also

- -

@NODES - Selector

- -

SELECT - Region

- - - - diff --git a/Help/helps/@PATH.htm b/Help/helps/@PATH.htm deleted file mode 100644 index 19ae751..0000000 --- a/Help/helps/@PATH.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -@PATH - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@PATH

- -

A selector used in the Context Zone of the pass file.  The - Context Zone specifies methods for selecting nodes in the parse tree.

- -

See Also

- -

@PATH - Selector

- -

SELECT - Region

- -

 

- - - - diff --git a/Help/helps/@POST.htm b/Help/helps/@POST.htm deleted file mode 100644 index aef7fd1..0000000 --- a/Help/helps/@POST.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -@POST - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@POST

- -

A region marker that indicates - the start of a POST Region.

- -

 

- -

See Also

- -

POST - Region

- -

POST - Action

- - - - diff --git a/Help/helps/@PRE.htm b/Help/helps/@PRE.htm deleted file mode 100644 index 770ebb1..0000000 --- a/Help/helps/@PRE.htm +++ /dev/null @@ -1,53 +0,0 @@ - - - - - - - - - - -@PRE - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@PRE

- -

A region marker that indicates - the start of a PRE Region.

- -

See Also

- -

PRE Region

- -

PRE - Action

- - - - diff --git a/Help/helps/@RECURSE.htm b/Help/helps/@RECURSE.htm deleted file mode 100644 index 86e1a60..0000000 --- a/Help/helps/@RECURSE.htm +++ /dev/null @@ -1,49 +0,0 @@ - - - - - - - - - - -@RECURSE - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@RECURSE

- -

A region marker that indicates the start of a RECURSE Region.

- -

See Also

- -

RECURSE Region

- - - - diff --git a/Help/helps/@RULES.htm b/Help/helps/@RULES.htm deleted file mode 100644 index 850a073..0000000 --- a/Help/helps/@RULES.htm +++ /dev/null @@ -1,49 +0,0 @@ - - - - - - - - - - -@RULES - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@RULES

- -

A region marker that indicates the start of a RULES Region.

- -

See Also

- -

RULES Region

- - - - diff --git a/Help/helps/@SELECT.htm b/Help/helps/@SELECT.htm deleted file mode 100644 index 2c86337..0000000 --- a/Help/helps/@SELECT.htm +++ /dev/null @@ -1,49 +0,0 @@ - - - - - - - - - - -@SELECT - - - - - - - - - - - - - - - - - - - - - - - - - - - -

@SELECT

- -

A region marker that indicates the start of a SELECT Region.

- -

See Also

- -

SELECT Region

- - - - diff --git a/Help/helps/Ana_Tab_Popup.htm b/Help/helps/Ana_Tab_Popup.htm deleted file mode 100644 index 25eae27..0000000 --- a/Help/helps/Ana_Tab_Popup.htm +++ /dev/null @@ -1,1085 +0,0 @@ - - - - - - - - - - -Ana Tab Popup - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Ana Tab Popup

- -

The Ana Tab Popup menu is launched by right clicking in the Ana Tab - Window.

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu - Item

 -

Description

-

Add

 -

Submenu for adding a pass to the analyzer - sequence. (See below.)

-

Attributes

 -

Launches the Attribute - Editor dialog and displays the attributes belonging to the selected - pass' concept in the KB. (Note that the entire analyzer sequence is in - the KB.) User has option to delete, add or change names of attributes - and values. It is recommended not to alter system attributes such as algo, - data, active and type.

-

Delete

 -

Deletes the selected pass from the analyzer - sequence.

-

Duplicate

 -

Creates a new pass immediately following and - identical to a selected pass in the analyzer sequence.

-

Edit Pass - File

 -

Opens the pass file of the selected pass in - the Workspace. (This is the same as double clicking on a pass in the analyzer - sequence.)  You - should not try to edit automatically generated rules in a pass file.  The - edits will not be saved when you regenerate the rules.  

-

Highlight - Matches

 -

Highlights words or phrases in a text file - matching selected pass in the Ana Tab or selected concept in the Gram - Tab. Must be selected before an analyzer is run over text.

-

View

 -

Submenu for viewing parse trees and other - windows relating to the current pass. (See - below.)

-

Ana Archive

 -

Submenu for archiving the analyzer sequence - and pass files. (See - below.)

-

Properties

 -

Launches the Pass - Properties dialog for non-system passes. (See below.)

-

Help

 -

Launches the VisualText Help documentation.

- -

 

- -

Add Submenu

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Folder

 -

Creates a folder for organizing passes. (Currently unavailable.)

-

Library - Pass

 -

Submenu for selecting a pass file from an - existing library of passes. (See - below.)

-

Main

 -

Creates a pass that will be executed first. (Currently - unavailable.)

-

New Pass

 -

Creates a new pass and empty pass file.

-

Pass from - File

 -

Launches an Open dialog to navigate to and - select existing pass file. Creates a new pass in the analyzer sequence - associating the existing pass file with it.

-

Stub

 -

Creates a stub region, simultaneously adding - a stub concept in the Gram - Tab.

-

Top Level

 -

Submenu for adding passes. (See - below.)

- -

 

- -

Add Library Pass Submenu

- -

- -

The Library Pass Submenu allows users to add a new pass and to associate - it with a pre-built pass file (or "library pass"). The contents - of this submenu adjust dynamically, depending on pass files that have - been placed in the release folder TextAI\VisualText\data\spec. - Users may add their own favorite or generic pass files to this folder. -  The number - next to the library pass indicates the number of passes in the library - pass sequence.  For - more information on the default library passes provided with VisualText, - see Library Passes.

- -

 

- -

Add Top Level Submenu

- -

- -

The available menu items in Add > - Top Level function identically with those of the Add - submenu. This menu provides - "hooks" for embedding sequences of passes within folders.

- -

Add Top Level Library Pass Submenu

- -

- -

This is identical with the Add > - Library Pass submenu. It provides a hook for placing a library - pass within a folder. This will be in an upcoming release of VisualText. -  The library - passes in this submenu are discussed in Library - Passes.

- -

View Submenu

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Concept Tree

 -

Displays the concept - tree of a parsed document in the Workspace. (Same as the View Concept - Tree button.)

-

Final Tree

 -

Displays the parse tree after the final pass - of an analyzer sequence.

-

Entire Tree

 -

Displays the full parse tree, starting at - the root, assuming that the current selection in the Text Tab has been - analyzed.

-

Log

 -

Displays the text version of Entire Tree, - the parse tree for the currently selected pass.

-

Tree Matches

 -

Displays portion of the parse tree matching - selected pass in the analyzer sequence.

- -

Ana Archive Submenu

- -

Ana Archive is used to view - and create archives of the analyzer sequence either on remote servers - or on local machines.  Archiving - is used to create quick backups of your work and to facilitate communication - between developers working on the same analyzers.  To - set archiving preferences, select Preferences - under the File Menu and click on the Archiving - tab.

- -

- - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Archive Local

 -

Launches a dialog box to create a local archive - of the pass files in a zip file. The name of the archive defaults to the - name of the current analyzer, suffixed with the current date and time. -

-

Archive Remote

 -

Launches a dialog box to create a remote archive - of the current analyzer sequence in the Ana Tab.  The - name of the archive defaults to the name of the current analyzer suffixed - with the current date and time.  Archive - is created on the remote server.

-

View Local

 -

Displays local archives in the Ana Archive - > View Local dialog. Presents options to delete, rename, upload (send - to server) or load into current analyzer Workspace.  Listings - in the archive can be sorted by clicking on column headers.

-

View Remote

 -

Displays server archives in the Ana Archive - > View Remote dialog. Presents options to delete, rename, download - (send to local archive) or load into current analyzer Workspace. Listings - in archive can be sorted by clicking on column headers.

- -

Pass Properties Dialog

- -

- - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Item

 -

Description

-

Name

 -

Name of the pass file associated with the - pass in the analyzer sequence.

-

Type

 -

Type of algorithm for the pass. Type can be - either Pattern (for PAT algorithm) or Recursive (for REC algorithm)

-

Active

 -

By default, a pass is active indicating that - it will not be skipped in the analyzer sequence. Unselecting Active will - cause the pass to be skipped during analysis.

-

To End

 -

Makes passes after the current pass either - active or inactive.

-

All

 -

Makes all passes either active or inactive.

-

From X to - X

 -

Specifies a range of passes to make active - or inactive.

-

OK

 -

Confirms changes to the Pass Properties dialog.

-

Cancel

 -

Closes the Pass Properties dialog.

- - - - diff --git a/Help/helps/Analyzer_Templates.htm b/Help/helps/Analyzer_Templates.htm deleted file mode 100644 index 96de496..0000000 --- a/Help/helps/Analyzer_Templates.htm +++ /dev/null @@ -1,302 +0,0 @@ - - - - - - - - - - -Analyzer Templates - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Analyzer Templates

- -

VisualText analyzers are built from initial templates. -  A template - is a starter or skeletal analyzer that can be copied and modified.  

- -

The templates contain a sequence of passes designed to help you perform - certain tasks.  The - Database template for example, is an analyzer with passes that will allow - you to import an existing database into the knowledge base.  The - Bare template is used to create the basic or minimal analyzer.  It - contains only the tokenize pass.  All - templates start with the tokenize pass.

- -

Template Types

- -

The templates currently available are outlined below:

- - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Name

 -

Description

-

Bare

 -

Minimal or basic analyzer.  Contains - default system pass, tokenize only.

-

Database

 -

Analyzer with sequence of passes to import - records from an existing database into the VisualText knowledge base. -  (Passes - also available as library passes.)

-

DatesPlusNumbers

 -

Analyzer with sequence of passes to recognize - simple and compound numeric words, numeric sequences and lists, various - character formatting such as punctuation and contractions, telephone numbers, - and dates. (Passes also available as library passes.)

-

EnglishLexicon

 -

Analyzer with a built-in English lexicon of - 180,000 words. Lexical entries are coded with part of speech information. -  Lexicon - is stored in the knowledge base in the dict subhierarchy.

-

HTML

 -

Analyzer with a sequence of passes to recognize - HTML formatting elements, text tagging, special HTML characters, and tagging - structures. (Passes also available as library passes.)

-

XML

-

XML analyzer. (Passes also available as library - passes.)

-

XML DTD

 -

Analyzer with a sequence of passes to recognize - an XML document type definition (DTD) file.  (Passes - also available as library passes.)

- -

 

- - - - - - -
-

Note: - Passes in the Database, Date Plus Numerics, HTML, XML and XML DTD templates - are also available as Library Passes.  For - example, you can create an analyzer with the Bare template and choose - to add the sequence of passes from the Add > Library Pass menu in the - Ana Tab.  For - general information on library passes, see Library - Passes.  To - learn how to add these passes as library passes, see Adding - Library Passes.

- -

Selecting Template Type

- -

You set the type of analyzer you want to - build when you create a new analyzer.

- -

For analyzer templates that contain a sequence - of passes, note that the relative ordering of passes in the sequence should - not be changed.  If - the sequence of passes is changed, the analyzer may not work as designed.

- -

See Also

- -

Creating - Analyzers

- - - - diff --git a/Help/helps/Buttons/attribEditorDel.gif b/Help/helps/Buttons/attribEditorDel.gif deleted file mode 100644 index d47f329..0000000 Binary files a/Help/helps/Buttons/attribEditorDel.gif and /dev/null differ diff --git a/Help/helps/Buttons/phrase_tool_button_down.gif b/Help/helps/Buttons/phrase_tool_button_down.gif deleted file mode 100644 index 2611035..0000000 Binary files a/Help/helps/Buttons/phrase_tool_button_down.gif and /dev/null differ diff --git a/Help/helps/Buttons/phrase_tool_button_up.gif b/Help/helps/Buttons/phrase_tool_button_up.gif deleted file mode 100644 index 56816c5..0000000 Binary files a/Help/helps/Buttons/phrase_tool_button_up.gif and /dev/null differ diff --git a/Help/helps/CHECK.htm b/Help/helps/CHECK.htm deleted file mode 100644 index b058b82..0000000 --- a/Help/helps/CHECK.htm +++ /dev/null @@ -1,79 +0,0 @@ - - - - - - - - - - -@CHECK region - - - - - - - - - - - - - - - - - - - - - - - - - - - -

CHECK Region

- -

NLP++ code in the CHECK Region applies - after the rule matcher has succeeded in matching a rule. These represent - the final opportunity to reject the rule match based on inconsistencies - spanning one or more rule elements. For example, before accepting a noun-phrase - rule, we could check for agreement between the determiner and noun (e.g., - accept "this company" but reject "these company").

- -

Code in the CHECK Region precedes - the execution of element actions (e.g., the recurse action by which an element invokes a minipass). - Thus, the rule match can still be rejected before any side effects have - been executed.

- -

The special predefined function fail() - immediately terminates execution of code in the CHECK Region and flags - that the rule match has failed. The analogous function succeed() also terminates CHECK code immediately, - but flags success and goes on to execute the element actions and code - in the POST Region.

- -

See Also

- -

Grammar Region

- -

CHECK - Action

- -

PRE Region

- -

POST - Region

- -

RULES - Region

- - - - diff --git a/Help/helps/CODE.htm b/Help/helps/CODE.htm deleted file mode 100644 index c85ffa8..0000000 --- a/Help/helps/CODE.htm +++ /dev/null @@ -1,67 +0,0 @@ - - - - - - - - - - -@CODE Region - - - - - - - - - - - - - - - - - - - - - - - - - - - -

CODE Region

- -

The Code Zone may optionally contain a single CODE Region, which begins - with the @CODE marker and optionally terminates with the @@CODE marker.

- -

NLP++ code in the CODE Region is executed prior to any rule matching - for the current pass.

- -

The CODE Region may, in addition to NLP++ code syntax, incorporate CODE - Actions, i.e., specialized functions that are constrained to the CODE - Region.

- -

The exitpass() function is useful - for conditionally and immediately terminating the current pass, without - executing any rules.

- -

See Also

- -

CODE - Zone

- -

CODE - Action

- -

 

- - - - diff --git a/Help/helps/DECL.htm b/Help/helps/DECL.htm deleted file mode 100644 index 13cfdbf..0000000 --- a/Help/helps/DECL.htm +++ /dev/null @@ -1,121 +0,0 @@ - - - - - - - - - - -DECL Region - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

DECL Region

- -

The DECL Region is where you define your own functions in NLP++.  It - is delimited by @DECL and optionally by @@DECL.  When - present, it must be the first region in a pass file.

- -

With VisualText 1.2 and later, functions can take parameters with the - L local variable, - as well as return values and call themselves recursively.  In - addition, you may use the special NLP++ variable references to pass information - to and from functions.  Depending - on where a function is called from, it can appropriately handle the S, N, - and X variable - references in addition to the global, or G, - variable reference.  The - return type of functions can vary dynamically -- in fact, the same function - could return different types at different times.  Multi-valued - variables (i.e., arrays) can be used to have a function return multiple - values.

- -

Recursive functions work as expected.  Also, - you may call a function in an earlier pass than the pass in which it is - defined (i.e., forward reference is ok).  NLP++ - functions have syntax similar to C/C++ functions.

- -

<function definition> ::= <id> \( <local vars>  \) -   \{ - <statements> \}
-<local vars> ::= <local vars> \, <local var>  | - <local var>  | -  (empty)
-<local var> ::= 'L' \( <string> \)
-<statement> ::= <return statement>
-<return statement> ::= 'return' <expr> \;   | -   'return' - \;
 

- -

An example follows.

- -

@DECL
-
-myfunction(L("var"))    # Define your - function.
-{
-"output.txt" << "Entering - myfunction." << "\n";
-G("tmp") = L("var"); -   # - Assign value to global variable.
-L("tmp") = L("var"); -   # - Assign value to local variable.
-return L("tmp") * 2;   # Return a value.
-}

- -

@CODE

- -

"output.txt" << - myfunction(3) << "\n"; -   # - Call your function.
-
-@@CODE

- -

This function call returns and outputs the number "6" to the - file "output.txt".

- -

See Also

- -

Declare Zone

- - - - diff --git a/Help/helps/Declare.htm b/Help/helps/Declare.htm deleted file mode 100644 index 42856a7..0000000 --- a/Help/helps/Declare.htm +++ /dev/null @@ -1,50 +0,0 @@ - - - - - - - - - - -Declare Zone - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Declare Zone

- -

The Declare Zone consists of a single region, the Declare Region (or - Decl Region), in which you can define your own NLP++ functions.

- -

See Also

- -

DECL Region

- - - - diff --git a/Help/helps/Deployment.htm b/Help/helps/Deployment.htm deleted file mode 100644 index b76fcaa..0000000 --- a/Help/helps/Deployment.htm +++ /dev/null @@ -1,88 +0,0 @@ - - - - - - - - - - -Deployment - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Deployment

- -

Analyzers can run inside VisualText, as standalone - programs, or embedded within other programs.  See - the Compiled - and Standalone Analyzers topic for technical details.

- -

Analyzers built with VisualText are now readily - available.  See - the Apps tab of our website above for TAIParse, - our advanced analyzer for general English text.  TAIParse - currently performs accurate part-of-speech tagging (93+% in internal blind - tests) and also supports chunking, parsing, and semantic analysis of text. -  Passes - that use HTML and/or strip it out are available and can easily be extended.

- -

A Resume Analyzer - prototype at about 80% accuracy (internal blind testing) is also available - on the web site.  TAI - has constructed numerous other analyzers that will be made available periodically - at the Apps tab.

- -

A bridge enabling .NET - programs to call analyzers is also available.  See - the VisualText\apps\DotNetTest and DotNetWeb folders in the product installation - for sample deployments.  These - use the libraries DotNetAPI and DotNetBridge, both in VisualText\bin.

- -

VisualText now supports analyzers that process - Unicode, addressing all the world's - languages (and mixed-language texts).

- -

A Java - bridge is not part of the release, but examples of Java programs calling - C++ libraries are readily available on the Web.  If - you need help with that, please contact us.

- -

A Gnu/C++/Linux version of the VisualText - runtime libraries, enabling analyzers to run on Linux - platforms, is available.

- -

Questions?  Contact - us at support@textanalysis.com, 1-949-376-8507.  Business - line at 1-877-235-6259.

- -

 

- - - - diff --git a/Help/helps/Dialog_boxes/Pass_Properties_Dialog.htm b/Help/helps/Dialog_boxes/Pass_Properties_Dialog.htm deleted file mode 100644 index edbaaa9..0000000 --- a/Help/helps/Dialog_boxes/Pass_Properties_Dialog.htm +++ /dev/null @@ -1,255 +0,0 @@ - - - - - - - - - - -Pass Properties Dialog - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Pass Properties Dialog

- -

The Pass Properties Dialog is used to specify what kind of pass you - would like to create in the analyzer sequence.  This - dialog is also used to provide a name for the pass and to control whether - the pass is active or skipped over during the analysis process.

- -

The Pass Properties dialog is launched automatically when you add a - new pass to the analyzer sequence.  You - can also access the Pass Properties dialog by selecting an existing pass - in the Ana Tab, right-clicking and selecting Properties.

- -

- - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Item

 -

Description

-

Name

 -

Name of the pass file associated with the - pass in the analyzer sequence.

-

Type

 -

Type of algorithm for the pass. Type can either - be Pattern (for PAT algorithm) or Recursive (for REC algorithm)

-

Active

 -

By default, a pass is active indicating that - it will not be skipped in the analyzer sequence. Unselecting Active will - cause the pass to be skipped during analysis. See Inactivating - Passes.

-

To End

 -

Makes passes after the current pass either - active or inactive.

-

All

 -

Makes all passes either active or inactive.

-

From X to - X

 -

Specifies a range of passes to make active - or inactive.

-

OK

 -

Confirms changes to the Pass Properties dialog.

-

Cancel

 -

Closes the Pass Properties dialog.

- -

 

- - - - diff --git a/Help/helps/Dialog_boxes/archive_analyzer_dialog.gif b/Help/helps/Dialog_boxes/archive_analyzer_dialog.gif deleted file mode 100644 index 76a33a3..0000000 Binary files a/Help/helps/Dialog_boxes/archive_analyzer_dialog.gif and /dev/null differ diff --git a/Help/helps/Dialog_boxes/pass_properties_dialog.gif b/Help/helps/Dialog_boxes/pass_properties_dialog.gif deleted file mode 100644 index 816b06d..0000000 Binary files a/Help/helps/Dialog_boxes/pass_properties_dialog.gif and /dev/null differ diff --git a/Help/helps/DictTok_Sample.htm b/Help/helps/DictTok_Sample.htm deleted file mode 100644 index 4b8e1b4..0000000 --- a/Help/helps/DictTok_Sample.htm +++ /dev/null @@ -1,67 +0,0 @@ - - - - - - - - - - -DictTok_Sample - - - - - - - - - - - - - - - - - - - - - - - - - - - -

DictTok_Sample

- -

 [All - sample analyzers are in Program Files\TexAI\VisualText\apps.]

- -

Shows how to install and use the new DICTTOK analyzer pass.

- -

Basically you just edit the analyzer.seq file for your project by changing - "tokenize" to "dicttok".

- -

Then you need to make a knowledge base with POS information.  Or - swap in the KB files from the TAIPARSE\BIN folder to the BIN folder of - your analyzer project.

- -

WARNING:  We - assume you are using an empty or "don't care" knowledge base - that comes with the analyzer project by default.  If - you've built your own KB, you can rebuild it starting with the TAIParse - knowledge base.  TAI - support can help you start building your own KBs, just send us an email - at support@textanalysis.com .

- -

 

- -

 

- - - - diff --git a/Help/helps/DotNetTest.htm b/Help/helps/DotNetTest.htm deleted file mode 100644 index 0baa3f2..0000000 --- a/Help/helps/DotNetTest.htm +++ /dev/null @@ -1,51 +0,0 @@ - - - - - - - - - - -DotNetTest - - - - - - - - - - - - - - - - - - - - - - - - - - - -

DotNetTest

- -

The DotNetTest analyzer is available in the VisualText release at VisualText/apps/DotNetTest. -  This shows - how to use our DotNetAPI and DotNetBridge libraries in order to call a - VisualText analyzer from a .NET - program.

- -

 

- - - - diff --git a/Help/helps/Find_Window.htm b/Help/helps/Find_Window.htm deleted file mode 100644 index 8633577..0000000 --- a/Help/helps/Find_Window.htm +++ /dev/null @@ -1,73 +0,0 @@ - - - - - - - - - - -Find Window - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Find Window

- -

The Find Window displays search - results when a search is performed over multiple files. Clicking on a - line in the Find Window displays the file where the search target was - found in the Workspace.

- -

- -

Searches can be initiated by selecting - Find in Files from the Edit Menu - or by clicking on the Find in - Files button on - the Main Toolbar.

- -

Either action launches the Find in Files dialog box which - contains a number of search control parameters.

- - - - diff --git a/Help/helps/Functions/Miscellaneous_Functions/fail.htm b/Help/helps/Functions/Miscellaneous_Functions/fail.htm deleted file mode 100644 index 2be2961..0000000 --- a/Help/helps/Functions/Miscellaneous_Functions/fail.htm +++ /dev/null @@ -1,78 +0,0 @@ - - - - - - - - - - -fail - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

fail

- -

Purpose

- -

In CHECK Region, abort the rule that has just matched. Noop elsewhere.

- -

Syntax

- -

fail()

- -

Returns

- -

Remarks

- -

Constrained to appear directly in the CHECK region, not within functions - defined in the DECL region.

- -

Described both here and in Special Functions.

- -

Example

- -

See Also

- -

or, - succeed, Special - Functions, CHECK - Actions

- -

 

- - - - diff --git a/Help/helps/Functions/Regions_and_Rule_Actions/Pre_Actions/debug.htm b/Help/helps/Functions/Regions_and_Rule_Actions/Pre_Actions/debug.htm deleted file mode 100644 index a87056f..0000000 --- a/Help/helps/Functions/Regions_and_Rule_Actions/Pre_Actions/debug.htm +++ /dev/null @@ -1,118 +0,0 @@ - - - - - - - - - - -debug - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

debug

- -

Purpose

- -

Pre action that succeeds unconditionally. Used to place a C++ breakpoint - at a particular rule.

- -

Syntax

- -

<fromRuleEltNumber,toRuleEltNumber>debug;

- -

fromRuleEltNumber - type: int
-toRuleEltNumber - type: int

- -

Returns

- -

Remarks

- -

debug can appear in the PRE, CHECK and POST regions.

- -

Example

- -

@PRE
-<1,1> debug();

- -

@RULES

- -

_xNIL <- _xALPHA @@

- -

See Also

- -

Special Functions, - PRE - Actions

- -

 

- -

 

- - - - diff --git a/Help/helps/G.htm b/Help/helps/G.htm deleted file mode 100644 index 53895f3..0000000 --- a/Help/helps/G.htm +++ /dev/null @@ -1,46 +0,0 @@ - - - - - - - - - - -G - - - - - - - - - - - - - - - - - - - - - - - - - - - -

G

- -

Reference to a global variable.  See - Variables.

- - - - diff --git a/Help/helps/GenTxt.htm b/Help/helps/GenTxt.htm deleted file mode 100644 index 6e43b02..0000000 --- a/Help/helps/GenTxt.htm +++ /dev/null @@ -1,58 +0,0 @@ - - - - - - - - - - -GenTxt - - - - - - - - - - - - - - - - - - - - - - - - - - - -

The GenTxt General Analyzer

- -

NOTE: - GenTxt is no longer supported and is unavailable for download.  It - is superseded by TAIParse, a general English analyzer available at

- -

http://www.textanalysis.com/Apps/POS_Tagger/pos_tagger.html

- -

An older version that includes a semantic framework (yet to be incorporated - into the above) is available at

- -

http://www.textanalysis.com/Apps/Natural_Language_Processing/natural_language_processing.html

- -

A knowledge base that embeds WordNet is also available from TAI.

- -

 

- - - - diff --git a/Help/helps/Getting_Help/Customer_Support.htm b/Help/helps/Getting_Help/Customer_Support.htm deleted file mode 100644 index ab714b3..0000000 --- a/Help/helps/Getting_Help/Customer_Support.htm +++ /dev/null @@ -1,66 +0,0 @@ - - - - - - - - - - -Customer Support - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Customer Support

- -

Assistance from TAI

- -

TAI provides pre-built analyzers and knowledge bases to help jump start - your analyzer development project. For example, we provide a generic analyzer - called TAIParse that can be used as a starting point for many analyzers, - in both restricted and broad domains.  TAIParse - incorporates and utilizes a large knowledge base with dictionary information. -  A knowledge - base that embeds WordNet is available on request, as well.

- -

We also offer training and professional services to accelerate the development - of your analyzer.

- -

For more information about pre-built analyzers and other assistance - from Text Analysis International, please contact our customer support - line at 408-746-9932.  You - may also contact us via email at info@textanalysis.com.

- -

 

- -

 

- - - - diff --git a/Help/helps/Getting_Help/Tours_and_Tutorials.htm b/Help/helps/Getting_Help/Tours_and_Tutorials.htm deleted file mode 100644 index e4ff0b0..0000000 --- a/Help/helps/Getting_Help/Tours_and_Tutorials.htm +++ /dev/null @@ -1,76 +0,0 @@ - - - - - - - - - - -Tours and Tutorials - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tours and Tutorials

- -

We have provided - several tours and tutorials to help you get started with VisualText.  

- -

VisualText - Tours

- -

There are two - tours:  the - VisualText Quick Tour and the - Corporate Analyzer tour.  The - Quick Tour introduces you to the major components of VisualText. The Corporate - Analyzer walks you through the analyzer-building process.  We - suggest you take the VisualText - Quick Tour, study the tutorials, and then continue with the Corporate - Analyzer tour.

- -

VisualText - Tutorials

- -

Exercises in - the form of tutorials have been provided to help you learn VisualText. -  The tutorials - in the Tutorials section give you a chance to create an analyzer using - VisualText. A brief description of the tutorials and sample code is provided - in the Tutorial - Introduction. Various features and functions are explained as you - go, with no attempt to be systematic about the meaning of all the menus, - buttons and rules.

- - - - diff --git a/Help/helps/Getting_Help/VisualText_Help.htm b/Help/helps/Getting_Help/VisualText_Help.htm deleted file mode 100644 index b370368..0000000 --- a/Help/helps/Getting_Help/VisualText_Help.htm +++ /dev/null @@ -1,247 +0,0 @@ - - - - - - - - - - -VisualText Help - - - - - - - - - - - - - - - - - - - - - - - - - - - -

VisualText Help

- -

The present Help documentation can be accessed by selecting VisualText Help from the Help Menu.

- -

Help Stand-Alone Application

- -

The Help documentation is also available as a stand-alone application, - from the Windows Start button > Programs > VisualText > Help. - The executable is located by default in the installation directory:

- -

C:\Program Files\TextAI\VisualText\Help\Help.chm

- -

VisualText Help System

- -

The VisualText Help documentation is built using RoboHelp, which is - layered on top of HTML Help.  The - Help system appears in a separate Help Viewer window and is used to navigate - the help topics.  The - Help Viewer tabs include:  Contents, - Index, Search, - and Glossary.  Each - of these tabs is described below:

- -

Contents: The - Contents tab displays the sections and pages of the Help System.  To - view the contents of a section or page, click on the item and the related - topic is displayed in the Help Viewer.  The - major divisions are:

- - - -

Index:  The - Index tab contains a list of keywords and subkeywords that are linked - to topics in the Help system.  To - view topics linked to index items, click on the keyword and select Display. You can also hit the return - key and the related topic will appear in the Help Viewer.  When - more than one topic is linked to a keyword, a Topics - Found dialog opens.  From - the list of topics, you can select the topic you want to view.

- -

Search:  The - Search tab allows you to search for instances of specified words in the - Help system. When a word is found, the topics containing the word are - listed.  To - view the results, select the topic from the list and click Display.

- -

Glossary:  The - Glossary tab displays an alphabetical listing of technical terms related - to VisualText.  Definitions - for selected terms are displayed in the lower pane of the Glossary Tab.

- -

Using Help

- - - -

Documentation Feedback

- -

Please send suggestions about our Help documentation to help@textanalysis.com. -  This address - is for documentation issues only.  For - technical questions or comments, please send email to support@textanalysis.com.

- - - - - - -
-

Note: - To see product information and version - number, select Help > About - VisualText.

- -

 

- - - - diff --git a/Help/helps/Getting_Started/Caveats_and_Limitations.htm b/Help/helps/Getting_Started/Caveats_and_Limitations.htm deleted file mode 100644 index 36496f7..0000000 --- a/Help/helps/Getting_Started/Caveats_and_Limitations.htm +++ /dev/null @@ -1,1030 +0,0 @@ - - - - - - - - - - - - -Programming Notes and Issues - - - - - - - - - - - - - - - - - - - -
- -

Programming Notes and Issues

- -

This document details issues with -VisualText and NLP++. Additional information about the current release can be -found in the New in Current Version -page and in the Release Report, located by default at

- -

         C:\Program -Files\TextAI\VisualText\Docs\ReleaseReport.txt.

- -

DOCUMENTATION ISSUES

- -

         -Zipped -tutorial analyzers may be missing in some versions, so you'll have to start -working them from Tutorial number 1.

- -

         -Documentation -for Linux is not provided in the present Help.  Workaround: -contact our customer support for assistance.

- -

INSTALLATION

- -

         -On -rare occasions, the installation program may hang. Please try another computer, and contact our -customer support for workarounds.

- -

RUNNING ANALYZERS

- -

         -Analyzers -cannot be run on files with the .log extension.

- -

         -VisualText -is optimized for processing large numbers of relatively small text files. Input -files larger than 1MB may substantially -slow processing, and it is inadvisable to process individual files larger than 4 MB.  Workaround: -a program may split large texts into buffers and submit them to an analyzer one -at a time, using the knowledge base to manage information across segments of a -document.

- -

         -Scale -Up issues.  A variety -of scale-up issues have been addressed in VisualText 2.x. If you are experiencing problems, contact us.

- -

         -Knowledge -base DLL file.  For -huge knowledge bases, the KB.DLL compiled knowledge base can become extremely -large (100s of MB).  VisualText -QDBM addresses this, but is too slow for commercial-grade use at this time.

- -

GUI ISSUES

- -

         -It -is safest to edit the gram subhierarchy using the Gram Tab only, rather -than the KB Editor. Where feasible, use the Gram Concept Properties dialog -(available from Gram Tab popup > Properties) to edit attributes of gram -concepts. Use the Attribute Editor only to edit attributes that are not -accessible from the Gram Concept Properties dialog.

- -

         -Editing -the sys hierarchy can easily corrupt the Knowledge Base.  Before -editing the sys hierarchy (either manually or programmatically), you are -advised to create a backup copy of your project.  A -KB Safe Mode is provided in the VisualText preferences for those who want to -disable inadvertent manual editing of the sys hierarchy.

- -

         -Size -Dynamically in VisualText display preferences is currently working for the KB -Editor window only.

- -

         -Currently, -folders are not part of the Ana Tab. Descriptions for Ana Tab folders in the -Ana Tab Popup Menu are for future enhancements.

- -

WRITING CODE AND RULES

- -

         -NLP++ -is an evolving language.  If -constructs that you would like to see in the language are absent, you may often -create them yourself by writing a user-defined NLP++ function or a C++ version -within the User Project for an analyzer.  We -appreciate your feedback about capabilities to incorporate in the language, -both to provide you with faster execution of the functions and to enhance NLP++ -and VisualText as a development environment.

- -

         -One -way to implement a new type of pass algorithm is to define a new NLP++ function -in the user project and invoke that function from the code region of a -pass. The user function may traverse the parse tree, interface to third party -code, and so on.  (Now -the same can be done with an NLP++ function defined in a @DECL region).

- -

         -Some -data types are not yet implemented, including bool, and char. -Nonzero integer values are used for Boolean true and zero is used for -Boolean false, analogous to the C Programming Language.  We -recommend NOT explicitly comparing the returns of functions documented as bool -with 1 and 0.  For -example, write if (batchstart()) rather than if (batchstart() == 1).

- -

         -Numeric -overflow and underflow are not flagged.  For -example, the maximum int is 2147483647 in Windows and must be checked by -the programmer as needed.  For -large numbers, the float type should be used.

- -

         -POST -Actions should operate on rule elements from right to left (or bottom to top, -if written one element per line).

- -

When the POST -region includes group and excise actions, the numbering of rule elements is -affected. For example, if element 3 is to be excised, then referring to -elements 4, 5, etc. subsequently is invalid. Therefore, operations on rule -elements should proceed from right to left, that is from high-numbered rule -elements to lower-numbered elements. For example, the following POST actions -will work as expected:

- -

N("status",4) -= "good";
-excise(2,3);

- -

The node matching -element 4 gets a variable named "status" with value "good", -after which the nodes matching elements 2 and 3 are excised. But if written -this way:

- -

excise(2,3);
-N("status",4) = "good";

- -

there may no -longer be an element 4 after the excise action, or a different rule element may -be referred to than expected. Conditionally excising rule elements would -further complicate the tracking of rule element numbers.

- -

         -Stacking -POST Actions should be avoided.

- -

While group actions -may be repeated in a POST Region (but see below), the user should avoid -writing multiple actions (e.g., single, excise, singler, singlex) that modify -the parse tree within a single POST Region. After an action such as single, -the elements of the current rule may no longer available.

- -

         -GROUP -action differs in compiled versus interpreted versions of an analyzer
-
-In interpreted NLP++, a group action such as group(5,7,"_list") -collapses elements 5 through 7 under a new node called _list, which is numbered -5, with the subsequent element (which was numbered 8) now numbered as 6.  But -in compiled code, 5 through 7 are collapsed as before, but now elements 6 and 7 -are retained with zero nodes, and numbering is unchanged.  That -is, element 8 is still number 8.  This -discrepancy will be fixed in a subsequent version.

- -

         -Some -NLP++ actions will not accept general NLP++ expressions as arguments. For -example, singler(1,3) is valid, but -singler(G("lo"),G("hi")) will result in a syntax error.

- -

         -Partly -constrained wildcards dont work in some cases. -The following example

- -

@PRE
-<2,2> var(lower);
-@RULES
-_NP <-
- _The
- _xWILD [plus]
- _Bears
- @@

- -

will match nodes to -the wildcard regardless of whether they have a nonzero variable called lower. A workaround is to write the rule as

- -

@PRE
-<2,2> var(lower);
-@RULES
-_NP <-
- _The
- _xWILD [plus fail=(_Bears)]
- _Bears
- @@

- -

This will work as -specified. Only nodes with a nonzero variable called lower will match the -_xWILD element.

- -

         -Infinite -loops.  As in -most grammar systems, it is possible to write "grammar loops" in -NLP++.  Also, a -lone rule such as

- -

@POST

- -

  singler(1,1);

- -

@RULES

- -

_np <- noun [opt] -prep [lookahead] @@

- -

where everything is -optional except the lookahead elements, can cause an infinite loop.

- -

 

- -

         -Modifying -the lookahead region. Reducing or -grouping a node at or past the [lookahead] keyword can cause a pass to silently -stop matching rules, as below:

- -

 

- -

@POST

- -

  single(); # -Correct would be singler(1,4);

- -

@RULES

- -

_np <- _det _quan -_adj _noun _prep [lookahead] @@

- -

         -In -recursive analyzer passes, an infinite loop occurs with rules having one -right-hand-side element with singlet mode:

- -

_adjlist <- -_adj [s] @@

- -

 Here, -_adj is found "inside" repeatedly.  Two -ways to fix this are shown below.

- -

_adjlist [base] <- _adj [s] @@

- -

_adjlist <- _adj   @@

- -

 In -the first, case, _adj is found "inside" the current parse tree node only -once, since the base attribute placed -on the _adjlist keeps the internal _adj from being found inside it.  In -the second case, _adj only matches when it labels the current "top -level" parse tree node being examined.  Once -it is reduced to _adjlist, the internal _adj is no longer visible.

- -

         -KB -Phrase.  A -knowledge base PHRASE is implemented merely as the first element in the phrase. - Thus, fetching -the phrase and then adding a new first element requires re-fetching the phrase.

- -

G("phrase") -= findphrase(G("some concept"));  # -Fetch the "phrase", i.e., the first node.

- -

addnode(G("phrase"),"nodename", -1);           # -Add a new first node of phrase.

- -

In the above, -G("phrase") no longer points to the first node in the phrase.  A -workaround is to track the first node in a phrase, or to re-fetch the phrase as -needed using findphrase, as below

- -

 

- -

G("phrase") -= findphrase(G("some concept"));  # -Fetch the "phrase", i.e., the first node.

- -

addnode(G("phrase"),"nodename", -1);           # -Add a new first node of phrase.

- -

G("phrase") -= findphrase(G("some concept"));  # -Re-fetch the first node of phrase.

- -

 

- -

 

- -

 

- -
- - - - diff --git a/Help/helps/Getting_Started/Installation.htm b/Help/helps/Getting_Started/Installation.htm deleted file mode 100644 index a9ec81c..0000000 --- a/Help/helps/Getting_Started/Installation.htm +++ /dev/null @@ -1,337 +0,0 @@ - - - - - - - - - - -Installation - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Obtaining and Installing VisualText

- -

VisualText can be downloaded over the internet. In order to do this, - contact Text - Analysis International, Inc. for the appropriate password and file - name.

- -

Downloading VisualText

- -

In most cases, you will download VisualText directly from the Text Analysis - International website, at

- -

http://www.textanalysis.com

- -

You will be asked to type in your user name (usually your email address) - and a password supplied when you purchased the product.  You - will then select the download file and browse to a location on your machine - for downloading the file.  Remember - where you placed the downloaded file, as you will need to run it (e.g., - by double-clicking on it) in order to install the product on you machine. -  Follow - the instructions in the popup screens of the installation program.

- -

Some materials (e.g., sample tutorials) will be installed in the form - of compressed .zip files. If you do not have decompression software to - access these files, you can download WinZip from http://www.zdnet.com/downloads.

- -

Downloading via FTP

- -

In some cases, you may be given a password and instructions for downloading - material from Text Analysis International via an FTP address.

- -

Installing VisualText

- -

Before starting the installation process, make sure the information - on your computer is backed up.

- - - - - - -
-

Note: - If you already have a version of VisualText on your computer, you must - first remove the older version before you can install the newer one, as - described below. Once you have done this, restart the install program - to install the newer version.

- -

Start the installation process by double clicking on the VisualText - program icon in the directory where you downloaded the VisualText software. - This will launch the VisualText InstallShield Wizard.

- -

To Remove a Previous Installation of - VisualText:

- - - - - - - - -
-

Note: - It is not necessary nor will you be prompted to reboot your machine after - an earlier version of the software has been removed.

- -

To Install VisualText:

- - - -

 

- -

If you are installing VisualText and there was an older version on your - machine, make sure that any shortcuts you've created point to the right - one. Also, note that any subdirectories and files you may have created - under the VisualText path will not be removed when you remove an existing - installation. Therefore, you may wish to move or delete those old files. - In addition, you can delete the compressed .zip download file to save - space.

- -

Installation Files

- - - -

C:\Program Files\TextAI\VisualText\

- -

Note: This default location is the assumed - installation location in the tutorials - and discussions in this documentation.

- - - -

C:\apps

- - - -

C:\Program Files\TextAI\VisualText\bin\VisualText.exe.

- - - -

C:\Program Files\TextAI\VisualText\Help\Help.chm.

- - - -

C:\Program Files\TextAI\VisualText\docs

- -

The VisualText installation program creates a program folder containing - the VisualText icon, Help documentation, and Release Notes in the Start - menu under Programs > VisualText.

- -

Release Report

- -

A Release Report containing known issues and notes - on features can be found in a file called ReleaseNotes.txt in the docs - folder:

- -

C:\Program Files\TextAI\VisualText\docs\ReleaseNotes.txt

- -

To see a related list in the Help, refer to Programming - Notes and Issues.

- -

 

- - - - diff --git a/Help/helps/Gram_Tab_Popup.htm b/Help/helps/Gram_Tab_Popup.htm deleted file mode 100644 index e5e79c6..0000000 --- a/Help/helps/Gram_Tab_Popup.htm +++ /dev/null @@ -1,745 +0,0 @@ - - - - - - - - - - -Gram Tab Popup - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Gram Tab Popup

- -

The Gram Tab Popup menu is launched by right clicking in the Gram Tab - Window.

- -

- -

The Gram Tab is used to construct a concept (or sample) hierarchy and - to sample data from input files to the concept hierarchy.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Add

 -

Submenu for adding concepts to the sample - hierarchy. (See below.)

-

Delete

 -

Deletes selected concept - from the sample hierarchy.

-

Delete Children

 -

Deletes the children - of a selected concept but keeps the concept itself.

-

Find

 -

Finds selected concept across multiple files. - Results are displayed in the Find Window.

-

Delete Samples

 -

Deletes the sample concepts - associated with a rule concept or label concept.

-

Sample File

 -

Submenu for importing and exporting sample - files. (See below.)

-

Attributes

 -

Launches the Attribute - Editor dialog and displays the attributes belonging to the selected - item. User has option to delete, add or change names of attributes and - values.

-

Highlight - Matches

 -

Highlights words or phrases in a text file - matching selected concept in the sample hierarchy.

-

Mark - for Generation

 -

Manually sets an attribute called Dirty on - a concept so that Generate - Dirty will apply to that concept and its descendants.

-

Pass File

 -

Displays content of the pass file associated - with selected concept. Same effect as double clicking the concept.

-

Properties

 -

Displays Gram - Concept Properties dialog showing rule generation properties for selected - concept in the sample hierarchy. When an attribute is set for a particular - concept, the attribute is inherited by all the descendants of the concept - given the attribute. For example, when the Dirty attribute is set for - a folder concept, all concepts under the folder concept inherit the Dirty - attribute.

-

View Parse - Tree

 -

Displays the parse tree of the currently selected - text in the Workspace. Same effect as selecting Display - Entire Tree from the Parse - Tree Menu and View - Parse Tree button from the Debug Tool.

-

Gram Archive

 -

Submenu for archiving the sample hierarchy. - (See below.)

-

Help

 -

Launches VisualText Help documentation.

- -

Add Submenu

- -

- - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Stub

 -

Adds a stub at the top (i.e., leftmost) level - of the sample hierarchy. A stub concept is associated with a region of - automatically generated passes in the analyzer sequence.

-

Folder

 -

Adds a folder concept to the current stub - concept or folder concept. Folders are used to help organize the sample - hierarchy.

-

Rule

 -

Adds a rule concept to the current folder - concept or stub concept.

-

Label

 -

Adds a label concept to current rule concept. - Labels are used to organize subsamples of a given sample.

- -

Sample File Submenu

- -

- - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Export

 -

Creates a file containing all the samples - of the currently selected rule concept. Samples are added one per line.

-

Import

 -

Reads a file containing samples into the currently - selected rule concept. Each line in the file should contain only one sample - text.

- -

 

- -

Gram Archive Submenu

- -

Gram Archive is used to view - and create archives of the Gram Hierarchy either on remote servers or - on local machines.  Archiving - is used to create quick backups of your work and to facilitate communication - between developers working on the same analyzers.  To - set archiving preferences, select Preferences - under the File Menu and click on the Archiving - tab.

- -

- - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Archive Local

 -

Launches a dialog box to create a local archive - of the current sample hierarchy, including rule concepts and samples. - The name of the archive file defaults to the name of the current analyzer - suffixed with the current date and time.

-

Archive Remote

 -

Launches a dialog box to create a remote archive - of the current sample hierarchy, including rule concepts and samples. - The name of the archive file defaults to the name of the current analyzer - suffixed with the current date and time.  Archive - is created on the remote server.  

-

View Local

 -

Displays local archives in the Gram Archive - > View Local dialog. Presents options to delete, rename, upload (send - to server) or load into current analyzer Workspace.  Listings - in the archive can be sorted by clicking on column headers.

-

View Remote

 -

Displays server archives in the Gram Archive - > View Remote dialog. Presents options to delete, rename, download - (send to local archive) or load into current analyzer Workspace.  Listings - in the archive can be sorted by clicking on column headers.

- -

 

- - - - diff --git a/Help/helps/Help.brs b/Help/helps/Help.brs deleted file mode 100644 index 8b13789..0000000 --- a/Help/helps/Help.brs +++ /dev/null @@ -1 +0,0 @@ - diff --git a/Help/helps/Help.glo b/Help/helps/Help.glo deleted file mode 100644 index b92b466..0000000 --- a/Help/helps/Help.glo +++ /dev/null @@ -1,344 +0,0 @@ -NAME=.ana file -Analyzer project definition file. Used by the VisualText GUI to load an analyzer. -NAME=.pat file -A file associated with a user pass in the analyzer sequence which holds NLP++ rules and code. -NAME=action -An NLP++ function specialized for a particular region in a pass file. For example, the single action only works in the @POST region, after a rule has matched. -NAME=action region -A region related to rule matching. Includes the @PRE, @CHECK, and @POST regions. -NAME=algorithm -In general programming, a method or procedure for accomplishing a task. In VisualText, sometimes short for a pass algorithm. -NAME=Ana Tab -The part of the VisualText interface in which you control the text analyzer sequence, passes and pass files. -NAME=analyzer -A text analyzer, i.e., a program that takes text as input and processes it in some way. Text analyzers typically transform, critique, or extract information from text. -NAME=analyzer project -The full set of folders and files associated with a text analyzer. Includes files that define the analyzer, input files, and other data files used by the analyzer. -NAME=analyzer project file -The file that is used by the VisualText GUI to load an analyzer. Also called a .ana file. -NAME=ASCII Table -A table that displays the decimal and hexadecimal numbers and their corresponding ASCII characters. -NAME=attr -(1) Abbreviation for attribute. (2) a rule element modifier. -NAME=attribute -A knowledge base object attached to a concept and representing a property of the concept. Consists of a key (or name) and one or more values. A concept may have multiple attributes. -NAME=Attribute Editor -A VisualText tool that allows you to change, add and edit the attributes and/or values of a concept (or node) in the knowledge base. -NAME=backup aware -Refers to rule elements that don't overrun an adjacent rule element. Only an _xWILD with no lists (match, fail, except) is "backup aware". -NAME=Bare -Refers to the absolute basic or minimal analyzer. The Bare template comes with the system passes tokenize and lines. -NAME=batch -Refers to running a group of input files non-interactively. -NAME=Bison -Gnu freeware for writing grammars with associated code. Similar to YACC (Yet Another Compiler Compiler). -NAME=blank space -A single whitespace character. -NAME=bleeding -When a matched rule keeps another rule from matching. -NAME=C file -A C Programming Language code file. -NAME=CG -Abbreviation for Conceptual Grammar. -NAME=Character Viewer -a VisualText tool that shows a count of characters in a line of text and the ASCII characters for each line of text. -NAME=CHECK Action -A function restricted to operate in the @CHECK region, i.e., after a rule's right-hand-side phrase has matched. -NAME=CHECK Region -An action region of a pass file. NLP++ code in this region apply after the matcher has succeeded in matching a rule. -NAME=child concept -A concept that is the child of another concept. -NAME=CODE Action -A function that operates in the @CODE region. Some code actions are retained because they are still useful, but they may be overhauled in future VisualText releases. -NAME=CODE Region -The region delimited by @CODE and optionally @@CODE. Executes NLP++ code prior to any rule matching in the pass file. -NAME=Code Zone -The part of a pass file where NLP++ code that is independent of rule matching is written. -NAME=command file -A batch file containing commands that edit or add to the knowledge base. -NAME=concept -A knowledge base object. The VisualText knowledge base consists of a hierarchy of concepts. -NAME=Concept Oriented Programming -A method of programming that deals with the concepts underlying a task. -NAME=concept tree -A parse tree containing only nodes that have been built in the selected analyzer pass. -NAME=Conceptual Grammar -A knowledge representation framework consisting of concepts, attributes, and phrases combined into knowledge hierarchies and graphs -NAME=context node - -NAME=Context Zone -The part of a pass file where methods for selecting nodes of the parse tree are specified. -NAME=COP -Abbreviation for Concept Oriented Programming -NAME=DECL Region -Region for user-defined NLP++ functions; delimited by @DECL and (optionally) @@DECL. -NAME=dictionary concept -A concept in the dict hierarchy of the knowledge base. Also called word concept. -NAME=Dictionary Editor -A VisualText tool that allows you to make, add and edit a dictionary database. [Currently unavailable.] -NAME=dissolve -Remove a top-level node and more (or splice) its children to the top-level. -NAME=element action -An element modifier that specifies an action to be performed when the current rule has matched. For example, "rename=noun" specifies that the node matching the current element is to be renamed "noun." -NAME=element modifier -A keyword or keyword and value pair that affects the matching or follow-on actions of a rule element. For example, "plus" specifies that one or more nodes must match the current element. -NAME=escape -(1) to prefix a character with an escape character. (2) a character used as an escape character. -NAME=escape character -A character that indicates the succeeding characters should be taken literally rather than interpreted as a special character -NAME=except list -A list of elements that are exceptions to a match or fail list. For example, _xWILD [fail=( A B C ) except=( D )] will fail on A, B, or C except in the case that it also matches D. For example, could be used to fail on nouns except for humanNouns. -NAME=excise -Remove a sequence of nodes from a parse tree. -NAME=fail list -A list of elements that will cause a match to fail. For example, _xWILD [fail=( A B C )] will match nodes until it encounters one named A, B, or C. -NAME=feeding -When a rule match causes another rule to succeed. -NAME=file path - -NAME=folder concept -A concept to help organize related sets of samples in the Gram Tab. folder concepts contain other folder concepts and/or rule concepts. -NAME=Gram hierarchy -The hierarchy of concepts visible in the Gram Tab, and used to manage concepts for stubs, folders, rules, labels, and samples. -NAME=Gram Tab -The part of the VisualText interface in which you manage samples taken from input texts. -NAME=Grammar Region -The part of the pass file where rules and actions are written. -NAME=Grammar Zone -The part of a pass file where the main rules of the pass file are written. -NAME=Hex Viewer -Views text files as hexadecimal characters. -NAME=IDE -Abbreviation for integrated development environment. Refers to a user interface (GUI) program with tools that work in concert to support developers. -NAME=index concept - -NAME=integrated development environment -A user interface (GUI) program with tools that work in concert to support developers. -NAME=intermediate parse tree -A parse tree as modified by a selected pass in the Ana Tab. -NAME=internal node -A node that is not a leaf of the parse tree. See nonterminal node. -NAME=KB -Abbreviation for knowledge base -NAME=KB Editor -A VisualText tool that allows you to edit the knowledge base. -NAME=KB.DLL - -NAME=KBMS -Abbreviation for knowledge base management system -NAME=key-value pair - -NAME=keyword density - -NAME=knowledge base -(1) A hierarchical database. (2) A repository for concepts, relationships, and other knowledge, typically organized in a meaningful way. -NAME=knowledge base child concept - -NAME=knowledge base concept -A basic unit of knowledge in the Knowledge Base. -NAME=knowledge base management system -A software system for managing a knowledge base. Analogous to a database management system. -NAME=knowledge base node -A member of a list of nodes called a phrase. Each concept in the knowledge base may own one phrase. A node is very similar to a concept, except that it is not attached to the hierarchy. A node is often treated as an instance of a concept in the hierarchy. -NAME=label concept -A concept in the Gram Tab hierarchy that holds subsamples of larger samples. Label concepts can occur only under rule concepts. For example, a label concept called "AreaCode" might be placed under a rule concept named "PhoneNumer". -NAME=leaf -Same as leaf node. -NAME=leaf node -A token node of the parse tree. A leaf token has no children. Also called a terminal node. -NAME=leaf token -Same as leaf node. -NAME=lhs -(1) The lefthand side element of a rule, also called the suggested element. (2) The suggested node built for the lefthand side of a rule. -NAME=library pass -A prebuilt pass that may be copied into the current analyzer sequence. -NAME=list node -A parse tree node that collects a list of related nodes. -NAME=literal -A node or token that represents a literal string in the input text. -NAME=match list -A list of elements that must be matched. For example, _xWILD [match=( A B C )] will match nodes as long as they are named A, B, or C. -NAME=Minipass Zone -The part of a pass file where nested minipasses can be specified. -NAME=MULTI Selector -A selector used in the SELECT Region of a pass file which specifies a list of node names for the rule matcher. The whole subtree under the specified node is searched. -NAME=natural language engineering -Principled method for constructing NLP systems. -NAME=NLE -Abbreviation for natural language engineering -NAME=NLP -Abbreviation for natural language processing, a subfield of Artificial Intelligence concerned with human languages. -NAME=NLP++ -Proprietary programming language of Text Analysis International. A general programming language with specializations for natural language processing. -NAME=node -(1) A parse tree node. (2) A knowledge base node. -NAME=node variable -See parse tree node variable -NAME=NODES Selector -A selector used in the SELECT Region of a pass file which specifies a list of node names for the rule matcher. The phrase immediately under the specified node is searched. -NAME=nonliteral -A parse tree node that may represent more than one token. The name of a nonliteral always starts with an underscore character. -NAME=nonterminal node -A node that dominates another node in a parse tree. That is, it has children. -NAME=noop -A reduce action in the @POST region that does nothing. Used to override the default action of an empty @POST region. (When a @POST region is non-empty, noop becomes the default action.) -NAME=offset -A number that indicates character position of a node in text. -NAME=ontology -A hierarchy of concepts, typically used to categorize the world. -NAME=operator -Part of an NLP++ expression that performs a mathematical, logical, or other function. For example, "+" is an operator used to add two numbers or catenate two strings. -NAME=opt -Element modifier denoting an optional element. -NAME=optional -Element modifier denoting an optional rule element. -NAME=output action -An action that causes output to be written to a file. For example, ndump writes all the variables and values for a node out to a file. -NAME=parse -(1) To ascribe structure to a linear sequence of words or symbols according to the rules of a grammar. (2) To create a parse tree representing a text. (3) A parse tree or interpretation of a text. -NAME=parse tree -A data structure that tracks patterns matched in an input text. -NAME=parse tree node -A unit or data structure representing a piece of text or an idea that the text represents. Nodes are combined to form a parse tree. -NAME=parse tree node variable - -NAME=parser -A program that takes a text as input and produces a parse tree as output. -NAME=pass -A discrete step in the text analyzer with its own pass algorithm. -NAME=pass algorithm -The method to be executed for a pass in the analyzer. Examples are: tokenize, pattern, and recursive algorithms. -NAME=pass file -A file of NLP++ code and rules associated with a pass of the text analyzer. Pattern and Recursive pass algorithms make use of pass files. -NAME=pat algorithm -See Pattern algorithm. -NAME=PATH Selector -A selector used in the SELECT Region of a pass file which specifies a path of node names for the rule matcher. -NAME=Pattern algorithm -A pass algorithm that executes the rules in a pass file by traversing the parse tree once. -NAME=PNODE -Name for the parse tree node data type. -NAME=POST Action -An action that occurs in the @POST region. For example, single tells the rule matcher to build a new node for a matched rule. -NAME=POST Region -An action region of a pass file. NLP++ code in this region apply after a rule match has been accepted. -NAME=PRE Action -Action that occurs in the @PRE region. Further constrains the matching of a rule element. -NAME=PRE Region -An action region of a pass file. Actions in this region represent additional conditions on the matching of individual rule elements. -NAME=print action -An action that prints to a file. -NAME=rec algorithm -See Recursive algorithm. -NAME=recurse region -A named region that is marked by an initial @RECURSE and specifies a minipass within a pass file. The minipass is invoked by a recurse element modifier in a rule element, e.g., in the main Grammar Zone. -NAME=recursion -An algorithm whereby the same action is taken repeatedly until some condition or goal is achieved. -NAME=recursive algorithm -A pass algorithm that executes the rules in a pass file by traversing the parse tree multiple times, till no rules match. -NAME=recursive grammar -A set of rules that is executed repeatedly, till no rules match. -NAME=recursive pass -See Recursive algorithm. -NAME=reduce -(1) Place a sequence of one or more nodes under a new node. Often refers to placing the nodes that match the righthand side of a rule under a new node named with the lefthand side of a rule. (2) A reduce action. -NAME=reduce action -An action in the @POST region that builds a new node for a matched list of nodes. -NAME=reduction -A reduce action. -NAME=rewrite grammar -Same as recursive grammar. -NAME=RFA -Rule File Analyzer. A VisualText analyzer that reads the pass files for a text analyzer. A bootstrapping parser that reads a simplified dialect of NLP++. -NAME=RFB -Similar to the RFA, but parses the full NLP++ language. -NAME=rhs -(1) The righthand side phrase of a rule. (2) The nodes that matched the righthand side. -NAME=root -The top-level node of the parse tree. A node with no parent. -NAME=RUG -Abbreviation for the automated Rule Generation machinery. -NAME=rule -An NLP++ construct consisting of a phrase of elements and a suggested element. Actions for a matched rule are governed by preceding @PRE, @CHECK, and @POST regions, if present. -NAME=rule action -A specialized NLP++ function that knows about particular contexts. -NAME=rule concept -A Gram hierarchy concept that owns a set of related samples, from which the rule generator will generalize, merge, and generate rules. A rule concept may have label concepts under it and may own a set of samples. -NAME=rule element -A literal or nonliteral unit in the phrase of a rule. -NAME=rule file -Synonym for pass file. -NAME=rule matcher - -NAME=Rules File Analyzer -See RFA and RFB. -NAME=rules region -A region delimited by @RULES and optionally @@RULES, for holding NLP++ rules. -NAME=sample -A piece of a larger text, representing a discrete idea. For example, a "310-555-1212" is a sample telephone number. -NAME=sample concept -A concept that stores a single sample or subsample in the Gram Tab. Sample concepts are placed under rule concepts and label concepts. -NAME=sample hierarchy -Same as Gram hierarchy. -NAME=select region -A region optionally delimited by @SELECT and @@SELECT that specifies how context nodes are to be selected in the parse tree. -NAME=selector -A marker such as @NODES which specifies which nodes will be subjected to rule matching. -NAME=single-tier reduction -The standard reduce action in the @POST region. Specifies that a new node is to be built, with the matched phrase of nodes to be placed underneath the new node. -NAME=singlet -A phrase element modifier. -NAME=singlet chain - -NAME=spider -A program that automatically searches the World Wide Web and retrieves documents and links to those documents. -NAME=spidering -The process of searching the World Wide Web and automatically retrieving documents and links to those documents. -NAME=string - -NAME=structural description -Same as parse tree. -NAME=stub -A placeholder for a sequence of passes that will be generated automatically. -NAME=stub concept -A Gram Tab concept that is associated with a region of automatically generated passes in the analyzer sequence. -NAME=stub region -A sequence of automatically generated passes within the overall analyzer sequence. -NAME=subhierarchy - -NAME=subtree - -NAME=suggested element -A node to which matched elements of a rule are reduced. -NAME=suggested node - -NAME=system attribute -An attribute assigned by the system to a concept in the knowledge base. System attributes include algo, type, active, passnum etc. Altering system attributes of a concept is not recommended. -NAME=system pass -A pass in the analyzer sequence that is defined by the system not the user. The tokenize pass is an example of a system pass. -NAME=template -A starter or skeletal analyzer that can be copied and modified. -NAME=terminal node -A node that does not dominate another node in a parse tree. Also called a leaf node. -NAME=Text Tab -the part of the VisualText interface where you manage input and output files. -NAME=token -A node representing a literal text string. -NAME=tokenize -(1) Pass algorithm for converting an input text to an initial parse tree. (2) The default system and first pass in an analyzer sequence. -NAME=top-level node -A node being traversed by the rule matcher. -NAME=trigger -An element modifier that causes a rule to be matched by first matching the associated rule element. -NAME=uninterned - -NAME=variable -See node variable. -NAME=variable action -An action that operates on a node variable. -NAME=white space -also whitespace; any of space, tab, carriage return or newline -NAME=word concept -See dictionary concept. -NAME=YACC -Yet Another Compiler Compiler. See Bison. diff --git a/Help/helps/Help.hhc b/Help/helps/Help.hhc deleted file mode 100644 index d534ab9..0000000 --- a/Help/helps/Help.hhc +++ /dev/null @@ -1,2643 +0,0 @@ - - - - - - - - - - \ No newline at end of file diff --git a/Help/helps/Help.hhk b/Help/helps/Help.hhk deleted file mode 100644 index b67c0fa..0000000 --- a/Help/helps/Help.hhk +++ /dev/null @@ -1,6611 +0,0 @@ - - - - - - - \ No newline at end of file diff --git a/Help/helps/HowToBuild/Corporate/Anaphora/Anaphora.htm b/Help/helps/HowToBuild/Corporate/Anaphora/Anaphora.htm deleted file mode 100644 index 3b5fa4b..0000000 --- a/Help/helps/HowToBuild/Corporate/Anaphora/Anaphora.htm +++ /dev/null @@ -1,249 +0,0 @@ - - - - - - - - - - -Anaphora - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Companies

 -

CORPORATE ANALYZER
-Anaphora

 -

Company Conj

-
- - -
- - -

Ana Tab Window: Passes 22 & 23

- -

This section describes the analyzer passes "anaphora" and - "anaphoraCompany".

- -

- -

Anaphora Pass

- -

Anaphora is a fancy linguistic word for words and phrases that refer - to other words and phrases in a text. One common anaphora relates to pronouns - such as "they" and "its". Other phrases include words - that reference actions such as "the sale", "this last purchase", - etc. Below are highlights from the "anaphora" pass:

- -

- -

AnaphoraCompany Pass

- -

First, we simply match an _anaphora node. If we find a suitable company - candidate in previous sentences (shown further below), then we change - our _anaphora to be a _company and point to the company found previously - in the text.

- -

- -

For debugging purposes, we write the results to a dump file called "anaphora.txt". - Here is the final output of this pass:

- -

- -

Searching Through Sentences

- -

Once the rule finds an _anaphora node, we can now use our "parse" - concept in the KB to search backward for the company being referred to. - This finally shows the power behind our KB-based analyzer design.

- -

Let us remember where we are. We are matching this rule in the context - of the @PATH, which is set to match in sentences and therefore we can - directly access the sentence object we have created under the "parse" - concept of the KB. In our @CHECK area, we will loop through the previous - sentences in the KB, and through the objects we created under those sentences, - in order to find the first object that matches the desired type of anaphora.

- -

For example, if the anaphora is "it", then we will look backward - for the first "company" type. If the anaphora is an "action" - type such as "this last purchase", then we will back up to find - the first "action" type. Although this example is simplified, - it can be designed to deal with much more complicated and subtle linguistic - constructions.

- -

We loop through the sentences and through the objects within each sentence. - We know by design that there will only be two types of objects: "company" - and "action" so we search for a match accordingly. When we find - a match, we set S("exit") to 1 and that exits our while loops. - If we find no object, we call fail, - a special function for the @CHECK area that aborts the rule match altogether.

- -

- - - -
- - -

Next Section: - Company Conj

- - - - diff --git a/Help/helps/HowToBuild/Corporate/CompConj/CompConj.htm b/Help/helps/HowToBuild/Corporate/CompConj/CompConj.htm deleted file mode 100644 index 1682f91..0000000 --- a/Help/helps/HowToBuild/Corporate/CompConj/CompConj.htm +++ /dev/null @@ -1,194 +0,0 @@ - - - - - - - - - - -CompConj - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

Anaphora

 -

CORPORATE ANALYZER
-Company Conjunction

 -

Events

-
- - -
- - -

Ana Tab Window: Pass 24

- -

This section describes the analyzer pass "conjCompany".

- -

- -

ConjCompany Pass

- -

This pass is another example of conjunctions (see "Money" - section) but this time we are dealing with companies. It is constructed - in the same manner as the conjMoney pass:

- -

- -

The conjunction rule was built primarily to handle the following sentence. - A more general treatment might use a recursive pass to gather conjunctions - one at a time.

- -

- - - -
- - -

Next Section: - Events

- - - - diff --git a/Help/helps/HowToBuild/Corporate/Companies/Companies.htm b/Help/helps/HowToBuild/Corporate/Companies/Companies.htm deleted file mode 100644 index eb56582..0000000 --- a/Help/helps/HowToBuild/Corporate/Companies/Companies.htm +++ /dev/null @@ -1,315 +0,0 @@ - - - - - - - - - - -Companies - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Money

 -

CORPORATE ANALYZER
-Companies

 -

Anaphora

-
- - -
- - -

Ana Tab Window: Passes 19 - 21

- -

This section describes the analyzer passes "companies", "companyAttributes", - and "companyDet".

- -

- -

Seeing Green

- -

There are currently two colors that appear when highlighting rule matches - in the input text: blue and green. Usually, the text will highlight in - blue, representing newly built nodes. When a rule matches without building - new nodes, the text corresponding to the rule match is highlighted in - green.

- -

Companies Pass

- -

Below is the highlighting for the company pass in our corporate analyzer, - pass 19:

- -

- -

The rule in this pass finds company nodes, but does not alter the parse - tree. Rather, it updates the KB "parse" area's tracking of company - instances found in the text, within each sentence of the text. The "parse" - area was introduced in the initKB pass.

- -

- -

Also, the "sentences" pass creates a numbered sentence concept - (e.g., "sentence1") for every sentence that it detects in the - parse tree, as well as having the "_sentence" node in the parse - tree refer to its corresponding name in the "parse" area. We - can see this using our handy "View>Tree of Selected" and - displaying any portion of the text as seen below. The sentence object - is written as ("name" "sentence1") ("object" - concept "<name>"):

- -

- -

With this in mind, we now want to create "company objects" - and attach them to the sentence object in the "parse" area of - the KB. Notice below the @PATH command. This directs the rule matcher - to search for matches only in "_sentence" nodes. In addition, - we can access the "_sentence" context node from a matched rule - with the special NLP++ X - function. X(varname) and X(varname, 3) refer to a variable in the last - and third component of the @PATH, respectively. Below, we create a new - company concept with the makeconcept - function and attach it to the sentence "object". We also add - the attributes "type" and "normal" to the new company - concept:

- -

- -

Green Highlighting

- -

Now that we understand this pass better, let us look at why the company - rule displays its matches in green.

- -

In the above rule, note two things: the use of _xNIL and the absence - of tree building functions in the @POST area.

- -

(1) _xNIL is not really a special nonliteral, but it is a convention - that documents when a rule is intended NOT to create a new, or suggested, - node.

- -

(2) In NLP++, an empty @POST region uses the single - reduction by default. But as soon as any code is placed in a @POST region, - that default is overridden. To get a rule to create a new node, an action - such as single must be specified - explicitly in that case.

- -

With a non-empty @POST area that lacks a reduce action, this rule fires, - but does not change the structure of the parse tree. Because the rule - fired without modifying the tree, the highlighting appears in green.

- -

CompanyAttributes Pass

- -

Remembering the money pass discussion involving noun phrases, we use - the same technique to gather attributes in the text surrounding companies. - Below are the rules that matched company attribute phrases such as "TAI - (Text Analysis Int'l)", "Founded in 1998, TAI", "California - company", etc. We leave it to you to investigate the companyAttributes - pass:

- -

- -

CompanyDet Pass

- -

The companyDet pass stands for company "determiner" pass. - Determiner is a linguistic term for those function words we put in front - of nouns and their adjectives, e.g., "the", "a", "this", - "that". In keeping with our idea of grabbing the entire company - phrase, we add this pass to simply attach the determiner to the company - phrase we are building in our passes 19-21.

- -

One interesting function you may have noticed in the previous pass (pass - 20) is pncopyvars. This function - is a handy function that copies all the variables from a matched node - to the newly created node. In this case, we can think of it as passing - all the meaning or semantic information from the _company item in the - rule below, to the new _company tree node being constructed by this rule. - Also notice the presence of the single - action. This builds the new node for the rule. We also include the partial - parse tree for the phrase "the Los Angeles-based AMRP". The - tree shows the semantic information passed up from the _company node below - to the _company node above:

- -

- - - -
- - -

Next Section: - Anaphora

- - - - diff --git a/Help/helps/HowToBuild/Corporate/Events/Events.htm b/Help/helps/HowToBuild/Corporate/Events/Events.htm deleted file mode 100644 index 09828c0..0000000 --- a/Help/helps/HowToBuild/Corporate/Events/Events.htm +++ /dev/null @@ -1,404 +0,0 @@ - - - - - - - - - - -Events - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Company - Conj

 -

CORPORATE ANALYZER
-Events

 -

Meta Events

-
- - -
- - -

Ana Tab Window: Pass 25

- -

This section describes the analyzer pass "events".

- -

- -

Important:

- -

This is by far the most important section of our corporate analyzer - tour. Here, everything we have done so far will come together.

- -

Events

- -

In language, it is usual to have some kind of action as the center of - meaning within a sentence. John kicked - the ball. The ball was kicked - by John. Actions are associated with verbs. The corporate world deals - with sales, mergers, acquisitions, and statements of fact about a company, - e.g., "Founded in 1972". Below, we can see the events that our - analyzer recognizes in pass 25:

- -

- -

Event Rule 1

- -

Our "events" pass 25 contains two rules. One is for a stock - "event", the second for "buy" and "acquire" - events. The second rule is :

- -

- -

There are several remarkable things going on here. First, if you look - at the graphic above containing the highlighting, you will see that many - events are highlighted, yet there are only two rules in this pass. The - second rule is generic enough to match three quite distinct sentences.

- -

This follows from the way in which the analyzer is constructed up to - the current pass. Remember that we gathered not only company names and - money values, but also the attributes in the text surrounding these constructs. - That is, we gathered complex noun phrases to distill the domain-critical - objects from the text. Once we gather all the subphrases into one neat - concept such as _company or _money, we can build abstract rules for events, - as in the second rule above. The idea that "companies buy companies - for money" is clearly expressed by rules at the event level of abstraction.

- -

In this way, an information extraction system can become a powerful - tool for accurately and deeply extracting critical content from a text.

- -

Generate Rule

- -

The "Generate Rule" menu item is a super-accelerator for building - rules by hand. You invoke it from the right-click popup menu of a text - window. It automatically generates rules from text in much the same way - that RUG generates rules.

- -

NOTE: - Generate Rule only works if you have previously run the analyzer on the - text with the "Generate Logs" toggle turned on:

- -

- -

Using our other handy function "View>Tree of Selected", - we can take a look at the tree for pass 25 for sentence 5:

- -

- -

Examining the top-level nodes of the parse tree above, you will see - your rule! I.e., "_company have _acquire _company for _money". - This is what Generate Rule does for you automatically, as follows.

- -

Below, choose the same text and instead of viewing the tree, select - "Generate Rule".  (Note - that the period is not included in the selected text.)  Generate - Rule goes to the events pass, opens it up, and places a new rule at the - bottom of the file. You can see the new rule on the left, in the image - below:

- -

- -

We can also use Generate Rule to generate another rule for a second - sentence, with an interesting result in this case:

- -

- -

Note that the two generated rules differ only by the concept "_have", - even though the sentences appear to be very different. Also, notice that - we have left in the "real" rule above the generated rule to - show you the final form. The final rule shows that "_have" is - "optional" and that the verb or action for the event is a list - that matches either _buy or _acquire concepts. In fact, this rule matches - three sentences that on the surface appear to be very different, but linguistically, - turn out to be very similar in structure. And the corporate analyzer has - helped uncover the similarity!

- -

Action Concept

- -

When the event rules fire, we want to create an action object in the - sentence under the "parse" concept. In the first part of our - @POST area for the above rule, we add the object in the same way we added - our object for company in a previous section. The following code in the - @POST area creates the object in the KB "parse" area:

- -

- -

You can see the resulting KB structure below after the @POST action - code has executed:

- -

- -

Conjunctions

- -

The second part of the @POST area for the second rule is where we recover - our company and money conjunctions. Again, conjunctions are groupings - of similar items in a list and are normally tedious to deal with in text - analysis systems. Remembering that we gathered up _money and _company - concepts into arrays, we can now easily loop through those arrays to create - the concepts in our "parse" area of the Knowledge Base. Below, - you can see the two "while" loops where we loop through the - arrays for company and money. The functions addstrval - and addnumval add multiple values - to a concept's attribute. The first adds a string value and the second - an integer value.

- -

- -

One important note here is the use of the special NLP++ function N. Notice - that in the first conjunction area for companies, we are using N("conj - count",4) and in the second, for money, we are using N("conj - count",6). The reason for this is that the _company concept is the - 4th matching element of the rule, and the _money concept is the 6th. We - repeat the rule here for convenience:

- -

- -

Stock Event

- -

The stock event is found as the first rule in the "events" - pass file and keys off the word "stock" and the verb "to - be". It works just like the "buy" and "acquire" - rule described above, but does not process conjunctions. It creates a - "stock" event in the sentence.

- -

- -

This rule constructs a slightly different structure in the KB than the - "buy" or "acquire" event. Whereas the second rule - constructs an event concept under its sentence, the first rule constructs - a "stock" concept under the associated company and adds related - information as attributes under the "stock" concept:

- -

- - - -
- - -

Next Section: - Meta Events

- - - - diff --git a/Help/helps/HowToBuild/Corporate/Format/Format.htm b/Help/helps/HowToBuild/Corporate/Format/Format.htm deleted file mode 100644 index 7246bcc..0000000 --- a/Help/helps/HowToBuild/Corporate/Format/Format.htm +++ /dev/null @@ -1,246 +0,0 @@ - - - - - - - - - - -Format - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Join

 -

CORPORATE ANALYZER
-Formatting

 -

Rule Generation

-
- - -
- - -

Ana Tab Window: Passes 5, 6, 7

- -

This section describes the analyzer passes "sentences", "paragraphs", - and "removeWhite".

- -

- -

Formatting

- -

It is often useful for an analyzer to track the formatting of the input - text as part of characterizing it. Pass 5 groups words into sentences. - It assumes that prior passes, e.g., pass 4 (join), have removed or hidden - periods (and semicolons, question marks, etc.) that are not likely to - be end-of-sentence markers in the current text. The resulting parse tree - is shown below:

- -

- -

Pass 6 groups the sentences into paragraphs:

- -

- -

Another convenient step, which helps simplify rules in subsequent passes, - is to excise white space from the parse tree. In this way, subsequent - rules need not consider white space. Below is a display of a tree before - and after spaces are removed. Before:

- -

- -

The parse tree below results from running Pass 7, called "removeWhite":

- -

- -

Precision Processing

- -

An advantage of grouping the text into sentences and paragraphs is that - it makes for a more precise and efficient text analyzer. Subsequent passes - can be instructed to process only within particular contexts, such as - sentences or paragraphs. In this way, spurious matches in unintended areas - of the parse tree are minimized.

- -

The @PATH selector, as shown in the pass file for pass 7 below, instructs - the rule matcher to start at the root of the parse tree (named "_ROOT"), - then find children named "_paragraph", then find their children - named "_sentence". Only when such a "_sentence" node - is found does the rule matcher attempt to match the rules of the current - pass.

- -

- - - -
- - -

Next Section: - Rule Generation

- - - - diff --git a/Help/helps/HowToBuild/Corporate/GramTab/GramTab.htm b/Help/helps/HowToBuild/Corporate/GramTab/GramTab.htm deleted file mode 100644 index 53d069c..0000000 --- a/Help/helps/HowToBuild/Corporate/GramTab/GramTab.htm +++ /dev/null @@ -1,297 +0,0 @@ - - - - - - - - - - -GramTab - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Formatting

 -

CORPORATE ANALYZER
-Rule Generation

 -

KB & Dictionary -

-
- - -
- - -

Ana Tab Window: Passes 8 - 12

- -

This section describes the analyzer passes "city", "company", - and "state" in the "words" stub region.

- -

- -

Sample Files

- -

We have included a few automatically generated (or "RUG") - passes in the corporate analyzer, for the purpose of "dictionary" - lookup. We describe how we set up this capability in the corporate analyzer.

- -

First, we added a folder called "Samples" under which we have - created three files: "cities.txt", "companies.txt", - and "states.txt". Below is "companies.txt". Note that - each text sample (company name in this case) appears on a separate line. - Only multi-word phrases are added to the companies.txt file, for reasons - that will become apparent later.

- -

- -

Stubs

- -

Next, we created a "stub", or placeholder in the analyzer - sequence, for automatically generated passes. This allows us to control - where in the sequence of passes these automatically generated passes will - occur. Since we are using RUG for a dictionary-like function, the stub - comes early on in the analyzer flow, right after formatting:

- -

- -

Gram Tab

- -

The Gram Tab is where we set up a hierarchy of text samples. We created - a stub concept called "words" and under it, we created rule - concepts for city, company, and state. The Gram Tab allows us to quickly - add text samples to the rule concepts, e.g., by importing the sample files - built above. You can also add samples to the hierarchy manually by highlighting - a text region and selecting the rule concept in which to place it.

- -

- -

Setting Auto-Generation Flags

- -

Before using RUG to generate our rules for us, we may need to provide - constraints and instructions that control the rule-construction process. - In the case of literal constructs such as company names, we use the Gram - Concept Properties window to specify that "literal" rules are - to be generated.

- -

- -

Invoking RUG

- -

To automatically generate the passes, we now choose "All" - under "Generate Rules" under the "Analyzer" menu:

- -

- -

Automatically Generated Rules

- -

Now we can double-click on the rules in the Gram Tab to see the corresponding - passes that have been built. Note, in pass file 10, the warning about - editing automatically generated files. Currently, any edits made to such - a pass file will be overwritten the next time RUG is invoked.

- -

Second, notice the form of the rules. RUG creates rules that match the - phrase and rewrites them to "_company". The rules also include - a "@PRE" area that restricts the words to be capitalized or - not. The form of the rules in this pass file is determined by the flags - set in the "Properties" window above.

- -

- -

Highlighting Matches

- -

These pass files are like any other pass files in that you can highlight - pass matches. In the Gram Tab, you can click on any rule icon and display - the highlights as show below:

- -

- - - -
- - -

Next Section: - KB & Dictionary -

- - - - diff --git a/Help/helps/HowToBuild/Corporate/HandDict/HandDict.htm b/Help/helps/HowToBuild/Corporate/HandDict/HandDict.htm deleted file mode 100644 index eda7ee6..0000000 --- a/Help/helps/HowToBuild/Corporate/HandDict/HandDict.htm +++ /dev/null @@ -1,276 +0,0 @@ - - - - - - - - - - -HandDict - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Rule - Generation

 -

CORPORATE ANALYZER
-KB & Dictionary

 -

Money

-
- - -
- - -

Ana Tab Window: Passes 13 - 15

- -

This section describes the analyzer passes "lookup", "vocabulary", - and "phrases".

- -

- -

Lookup Pass

- -

In the last section, we saw how we can automatically generate a dictionary. - In this section, we show how to use the Knowledge Base as a dictionary. - The reason for doing this is that, unlike the current version of the RUG - passes, we cannot attach additional information to the automatically generated - rules.

- -

First, look at the @RULES area below. You will notice that we are looking - for phrases of length one through four. We do this simply by writing 4 - rules, attempting to match the longest phrase first:

- -

- -

Next, we examine the @CHECK area, which allows us to accept or reject - a match based on criteria we specify. The objective of this pass is to - look up the matched phrase in the "companies" area of our KB, - which we assume was set up before we built our analyzer. To perform the - lookup, we first fetch the text of the phrase using the phrasetext - function, assigning the text to a variable called "phrase". - We search for the phrase within the "companies" KB concept (defined - in the initKB pass), using the - findhierconcept function. If no - such phrase was found in the KB, we reject the rule match with the special - fail function that exits the current - rule match.

- -

- -

In the @POST area, we "normalize" the phrase whose concept - was found in the KB. Recall that we set up our KB such that we placed - alternative wordings for our companies under a "normalized" - form. So, we use the findattr - function to look for the "synonym" attribute, as above. If it - is there, we need to set the concept's parent as the normalized form of - the phrase. We repeat our graphic from KB initialization here for clarification:

- -

- -

Debug Printouts

- -

To help debug the corporate analyzer, we created a dump file called - "lookup.txt". The C++ -like "<<" operator conveniently - pipes values to an output file. Below is a screen shot of the lookup.txt - file.

- -

- -

Vocabulary Pass

- -

The "vocabulary" pass illustrates how we can quickly build - vocabulary lookups by hand.

- -

- -

Phrases Pass

- -

We can also build a pass for phrases by hand. Note that separating the - "vocabulary" and "phrases" passes enables rules in - the later pass to use nonliterals built in the earlier pass. For example, - the first rule below uses the nonliteral "_conj", or conjunction - token, to handle things like "CEO and president" or "CEO - & president".

- -

- - - -
- - -

Next Section: - Money

- - - - diff --git a/Help/helps/HowToBuild/Corporate/InitKB/InitKB.htm b/Help/helps/HowToBuild/Corporate/InitKB/InitKB.htm deleted file mode 100644 index 8391e48..0000000 --- a/Help/helps/HowToBuild/Corporate/InitKB/InitKB.htm +++ /dev/null @@ -1,224 +0,0 @@ - - - - - - - - - - -InitKB - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Intro

 -

CORPORATE ANALYZER
-Initial KB

 -

Join

-
- - -
- - -

Ana Tab Window: Pass 3

- -

This section describes the analyzer pass "initKB".

- -

- -

Constructing the KB

- -

Because the corporate analyzer is expected to analyze more than one - text, we need to make sure the "parse" area is cleaned out each - time. More generally, we need to initialize or reset any parts of the - KB that the analyzer manages dynamically. If the KB happens to not have - a "corporate" concept, the analyzer can recover by creating - that area of the KB from scratch.

- -

- -

NLP++ Knowledge Base Functions

- -

To initialize the KB, we'll use predefined knowledge base functions - supplied with NLP++. The initKB pass consists purely of code, with no - rules, as shown in the screen shot below.

- -

The function findroot returns - the root node of the KB hierarchy. The first call to findconcept - looks for a concept named "corporate" directly under the root. - A global variable named "corporate" is assigned the corporate - concept, if found (or 0, otherwise). The special function G refers - to a global variable.

- -

If the corporate concept was not found, then makeconcept - is used to create it as a child of the root of the hierarchy. (Since the - corporate concept is absent, that means there is no "companies" - area with synonyms for company names. Even so, the analyzer can still - execute, though it may not correlate things like "TAI" and "Text - Analysis International".)

- -

If the "parse" concept is found in the knowledge base, then - its descendants are removed with rmchildren. - This clears the parse area so that the new article can be processed with - a clean slate.

- -

When initKB is executed, the global variables "corporate", - "parse", and "companies" point to their respective - concepts in the KB. A global variable called "section number" - is also initialized to 0 here.

- -

- - - -
- - -

Next Section: - Join

- - - - diff --git a/Help/helps/HowToBuild/Corporate/Intro/Intro.htm b/Help/helps/HowToBuild/Corporate/Intro/Intro.htm deleted file mode 100644 index 57e102a..0000000 --- a/Help/helps/HowToBuild/Corporate/Intro/Intro.htm +++ /dev/null @@ -1,429 +0,0 @@ - - - - - - - - - - -Intro - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 TOC

 -

CORPORATE ANALYZER
-Intro

 -

Initial KB

-
- - -
- - -

Intro

- -

Although VisualText may be the most powerful text processing tool around, - it may not be obvious to the first time user how to build an analyzer. - You may know what you want to do, but how do you go about doing it? This - demonstration analyzer gives a first-time user of VisualText an idea of - exactly how an analyzer is built.

- -

Notes: - This demonstration analyzer is meant to stimulate ideas in the VisualText - user. It does not represent in any way a complete or exemplary example - of analyzer building. We expect that you've at a minimum studied the Quick Tour and the Tutorials - in order to follow the Corporate tour.

- -

Now is a good time to open the corporate analyzer, as - described in the Quick Tour, and examine the input file named "Sold.txt" - in the Dev folder of the Text Tab.

- -

Our Analyzer

- -

Before we can build an analyzer, we must have a task in mind. In our - corporate analyzer, we want to pull out information about companies from - newswires. We can pull out information like "who bought whom for - how much", "who's stock changed and by how much", and other - information about companies such as current CEO, founding date, and contact - information.

- -

Our corporate analyzer will build records that could be used to feed - a database. This type of analyzer is called an "information extraction" - system. Here is a peek at the corporate analyzer's first output record - for our sample text:

- -

- -

VisualText Features and Tips

- -

Key features that make VisualText powerful, flexible, and easy to use - and debug are tabulated in the Quick Tour Introduction. - Some areas we'll focus on in this deeper tour of the Corporate analyzer - are:

- -
    - -

     Multi-Pass Processing

    - -

     Knowledge Base

    - -

     Automatic Rule Generation

    -
- -

Some useful guidelines gleaned from building information extraction - systems are:

- -
    - -

     Do high confidence work first

    - -

     Group text into chunks to process

    - -

     Don't process everything

    - -

     Don't worry about mistakes if they - don't hurt you

    -
- -

We will see examples of these ideas at work in our corporate analyzer - example.

- -

Plan of Attack

- -

The next step is to sketch a way to build this analyzer. Building a - text analyzer is art, not science, and every developer will write a unique - analyzer.

- -

An initial decision to make is whether your analyzer will gather localized - information only (i.e., do "shallow" processing), or correlate - information from various parts of a text (or multiple texts). The deeper - and more accurate the analyzer, the more correlation you will need to - do, and the more you will rely on the knowledge base (KB) to store information - about the domain (e.g., "corporate events") and the current - text.

- -

The corporate analyzer will utilize the KB in order to demonstrate "going - deep."

- -

Knowledge Based Analyzers

- -

The first step in building our analyzer is to create a dedicated area - for it within the KB. We start by making a concept named "corporate", - as below. Under that, we create two concepts: "companies" and - "parse". The "companies" concept holds the companies - we are interested in and their alternative wordings, as reference knowledge - for the analyzer. The "parse" concept stores information that - the analyzer creates and uses dynamically.

- -

- -

KB Lookup and Normalization

- -

The knowledge base can support looking up and normalizing words and - phrases. For example, to store the fact that "Text Analysis International, - Inc." and "TAI" are synonyms, you can construct a hierarchy - with "TAI" as parent and place its alternative phrasings below - it. Under the "companies" concept, we'll place a normalized - form of a company name, e.g., its abbreviation when it's unique. Under - the normalized form, we place major textual variants. To explicitly specify - the relationship between concepts, we add a "synonym" attribute - to each variant. The screen shot of the KB and Attribute Editor below - depicts this scheme.

- -

- -

The "companies" area under our "corporate" concept - is one we construct beforehand, either by hand or by writing another VisualText - analyzer to populate the knowledge base for us from a file. The "parse" - area, on the other hand, is managed dynamically by the corporate analyzer - while it is processing an input article. We assume, for our demo system, - that TAI and AMRP are the only companies the analyzer needs to know about!

- -

Parse Concept

- -

In the corporate analyzer, we need to gather information across sentences, - because information about a given company may be spread throughout the - article. For instance, "TAI buys AMRP" occurs in the first sentence, - while the phrase "The sale" in the fourth sentence" refers - back to the first sentence. If the fourth sentence adds useful information, - we need to relate the two. Therefore, we use the knowledge base to build - a model of the article as we process it, enabling correlations across - sentences. Below is a screen shot of the parse area built by the final - version of the analyzer for our sample input article.

- -

- - - -
- - -

Next Section: - Initial KB

- - - - diff --git a/Help/helps/HowToBuild/Corporate/Join/Join.htm b/Help/helps/HowToBuild/Corporate/Join/Join.htm deleted file mode 100644 index 50bf9e6..0000000 --- a/Help/helps/HowToBuild/Corporate/Join/Join.htm +++ /dev/null @@ -1,229 +0,0 @@ - - - - - - - - - - -Join - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Initial - KB

 -

CORPORATE ANALYZER
-Join

 -

Formatting

-
- - -
- - -

Ana Tab Window: Pass 4

- -

This section describes the analyzer pass "Join".

- -

- -

Tokenization

- -

The tokenize pass, which is - the first pass of every VisualText analyzer, performs an initial segmentation - of characters in the input text to alphabetic, numeric, punctuation, and - white space tokens. However, in practice, we may want to treat constructs - such as "$9.99" and "No. 1" as tokens. VisualText - makes it convenient to "join", or "retokenize", a - phrase as a single token.

- -

The example rule below illustrates how we might handle phrases such - as "No. 1". The rule itself appears after the @RULES marker. - It matches a phrases that starts with "No", continues with ".", - then any amount of white space (denoted by "star"), and finally - a numeric token. When such a phrase is matched, a new node named "_rank" - is built to dominate the matched phrase. The @POST action specifies what - will be done when the rule has been matched. In this case, the suggested - node, named "_rank", will get a variable named "number" - with the numeric token's text as its value. That is, "No. 33" - will lead to creation of a variable named "number" with value - "33", attached to the "_rank" node. The single - action specifies that a default reduction of the matched phrase is to - be executed. The rule below can be used to bleed, or hide, periods that - we don't want treated as ends of sentences.

- -

- -

Seeing the Matches

- -

VisualText allows you to quickly see the matches for any given pass. - The blue text corresponds to nodes built by rule matches in the "join" - pass.

- -

- -

Note that the apostrophe-s in possessive constructs is not highlighted - in blue. That is because the rule that identifies possessives, shown below, - excises the apostrophe and "s" from the parse tree.

- -

- - - -
- - -

Next Section: - Formatting

- - - - diff --git a/Help/helps/HowToBuild/Corporate/MetaEvents/MetaEvents.htm b/Help/helps/HowToBuild/Corporate/MetaEvents/MetaEvents.htm deleted file mode 100644 index 46e4041..0000000 --- a/Help/helps/HowToBuild/Corporate/MetaEvents/MetaEvents.htm +++ /dev/null @@ -1,325 +0,0 @@ - - - - - - - - - - -MetaEvents - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Events

 -

CORPORATE ANALYZER
-MetaEvents

 -

Output

-
- - -
- - -

Ana Tab Window: Pass 26-28

- -

This section describes the three analyzer passes: eventAttributes, anaphoraEvent, - and events2.

- -

- -

EventAttributes Pass

- -

Some phrases describe entire events. Below is one of them. The phrase - "In 2006 alone," describes the "acquisition" event - that follows. Notice that it is in green. We will explain below.

- -

- -

This pass takes care of attaching the date to the "acquire" - event via an attribute. If we look at the resulting information in the - KB, we can see the final format:

- -

- -

Notice in the rule itself that the _event item is already matched. This - pass (26) must come after pass 25 because the rules in pass 26 expect - that events have already been processed. We create a "date" - concept below the action concept and add a "year" attribute - and value to the new date concept. But we do not gather these items under - a new node. Remember that if we have a @POST area associated with a @RULES - area, we need to specifically call an NLP++ action in order to "reduce" - or create a new node under which all the items are gathered. There is - no "single()" function and the suggested node name is "_xNIL", - which is a convention telling us that we are not interested in building - a new node for this rule match. That is why VisualText displays the match - for this rule in "green" (see text above).

- -

- -

AnaphoraEvent Pass

- -

Pass 27 is much like pass 23, the anaphoraCompany pass. The purpose - of this pass is to find all the _anaphora concepts and loop back through - the sentences under the "parse" area of our KB to find a match. - Only this time, we are looking for "actions" to match, not "companies". - Below, you can see that phrases that reference other events in the text - are highlighted:

- -

- -

Below you can see the rule and the @POST area for this rule. We are - matching _anaphora and creating a new parent node _eventAnaphora. Notice - that since we have a @POST area, we must call the function "single()" - to create the _eventAnaphora node:

- -

- -

In the above @POST region, we also print out some information to the - dump file "anaphoraEvent.txt" for debugging purposes. This way - we can easily monitor what this pass is choosing as its reference to the - anaphora:

- -

- -

Looping Through the Sentences & - Objects 

- -

As in pass 23, we will loop back through the sentence concepts we built - under "parse" in the KB to find the "action" objects - we constructed earlier. In layman's terms, we are looking for "action" - events that match phrases like "The sale", "this last purchase", - etc. The code below is almost identical with pass 23 with the exception - of our check for "action" instead of "company":

- -

- -

Events2 Pass (28)

- -

This pass uses the previous pass in two ways. One, it looks for the - presence of an _eventAnaphora concept in a particular pattern. Two, once - it finds it, it makes a "comment" concept under the object that - _eventAnaphora is pointing to. Remember that in the previous pass (27), - we looped back through the "parse" area until we found an "action". - In this case, the phrase "The sale" in the 4th sentence ends - up attaching a "comment" concept under the "buy" action - in the second sentence:

- -

- -

Below is the rule that does this:

- -

- -

The second rule in pass 28 does something similar:

- -

- - - -
- - -

Next Section: - Output

- - - - diff --git a/Help/helps/HowToBuild/Corporate/Money/Money.htm b/Help/helps/HowToBuild/Corporate/Money/Money.htm deleted file mode 100644 index f794b64..0000000 --- a/Help/helps/HowToBuild/Corporate/Money/Money.htm +++ /dev/null @@ -1,265 +0,0 @@ - - - - - - - - - - -Money - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 KB - & Dictionary

 -

CORPORATE ANALYZER
-Money

 -

Companies

-
- - -
- - -

Ana Tab Window: Passes 16 - 18

- -

This section describes the analyzer passes "money", "moneyAttributes", - and "conjMoney".

- -

- -

Lessons in Phrases

- -

In diagramming sentences in English class, you may have come across - noun phrases (NP). These come - in many many forms and basically represent nouns (persons, places, things) - and attributes about those nouns (color, amount, type). Knowing this, - we can write a more organized text analyzer.

- -

In passes 16 - 18, we process money phrases. First we gather the root - money noun (money), then process its attributes (moneyAttributes), and - then gather lists of money constructs (conjMoney).

- -

Money

- -

In the money phrase, we key off the word "dollars". Since - this is a demonstration analyzer, we will only write rules for those patterns - we are matching in our "Sold.txt" article. The rules below match - constructs such as "$30 million dollars" and "30 million - dollars".

- -

- -

In the @POST area, we perform our "normalization", multiplying - the numeric value in the text by 1,000,000 to compute the number of dollars.

- -

Money Attributes:

- -

Next, we gather the attributes in the text surrounding the money phrases. - Study the moneyAttributes pass for yourself. Below is the highlighting - for this pass. You can easily see the attribute phrases: "more than - X", "X in cash", "X in assumed debt", "a - total of X", etc.:

- -

- -

 

- -

Conjunctions

- -

In analyzing text, conjunctions can be very painful to handle. Not so - in VisualText. NLP++ includes an array construct that can simplify the - problem of conjoining similar phrases. To illustrate, we will elaborate - one case.

- -

In the conjMoney pass, we take the money and value and put it in the - arrays S("conj value") and S("conj type") using the - counter S("conj count"). These will be used in future passes - to loop through and recover the money values.

- -

- -

Tree of Selected

- -

Below is the matched conjunction phrase and the tree for that phrase. - Two useful features of VisualText should be pointed out here. One is that - the semantic information is displayed in the trees. Below, you can see - the result of firing the above conjunction rule: ("conj count" - 2) ("conj value" 120000000140000000) ("conj type" - "cash" "assumed debt").

- -

The second convenient function is found in the right-click menu of the - processed text window. Select the pass you are interested in (usually - it is already selected), then select any portion of text you want. Using - the right-click menu, choose "View>Tree of Selected" and - the partial parse tree for that text will appear in the context of the - chosen pass. This is a handy function to remember.

- -

- - - -
- - -

Next Section: - Companies

- - - - diff --git a/Help/helps/HowToBuild/Corporate/Output/Output.htm b/Help/helps/HowToBuild/Corporate/Output/Output.htm deleted file mode 100644 index 13c320f..0000000 --- a/Help/helps/HowToBuild/Corporate/Output/Output.htm +++ /dev/null @@ -1,187 +0,0 @@ - - - - - - - - - - -Output - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Meta - Events

 -

CORPORATE ANALYZER
-Output

 -

 

-
- - -
- - -

Ana Tab Window: Pass 29

- -

This section describes the last analyzer pass, "output". (The - complete Corporate Analyzer sequence is shown below.)

- -

- -

Output

- -

One of the beauties of using a KB-based text analyzer is that once you - have everything processed, the output is a simple matter of looping through - the "sentence" concepts in the "parse" area in the - KB and printing out the results. If we open up the sentences under the - "parse" concept, we can easily see what we want to output to - the file. The rule is, if there is a child concept of the object under - the sentences, print it out. If not, don't output anything. Below, this - is clearly illustrated by looking at the concept for "TAI". - Those with a "+" have something to output, those without do - not:

- -

- -

We leave it to you to look at the details of the code for this pass. - Please note, however, that there is no rule for this pass, just a @CODE - area. The last rule-matching pass was pass 28, and all that remains is - to output our results. Notice, when inspecting the output file, that information - from later sentences has been copied back to earlier sentences. This is - due to our analyzer's treatment of "anaphora". Namely, information - later in the text has been correlated to references earlier in the text. - Below is one of the more interesting outputs, taken from sentence 5:

- -

- - - -
- - -

Go forth and build better - analyzers!

- - - - diff --git a/Help/helps/HowToBuild/Corporate/Toc/TOC.htm b/Help/helps/HowToBuild/Corporate/Toc/TOC.htm deleted file mode 100644 index b8f244c..0000000 --- a/Help/helps/HowToBuild/Corporate/Toc/TOC.htm +++ /dev/null @@ -1,576 +0,0 @@ - - - - - - - - - - -TOC - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

CORPORATE ANALYZER
-Table of Contents

- - -
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Pass #

 -

Name

 -

Description

-

 

 -

Intro

 -

Purpose and strategy for building the corporate analyzer

-

3

 -

Initial KB

 -

Knowledge Base initialization
-(Pass: InitKB)

-

4

 -

Join

 -

Tokenization and retokenization
-(Pass: Join)

-

5-7

 -

Formatting

 -

Formatting text into chunks
-(Passes: Sentences, Paragraphs, RemoveWhite)

-

8-12

 -

Rule Generation

 -

Working with automatic rule generation in the Gram Tab
-(Passes: City, Company, State)

-

13-15

 -

KB & Dictionary

 -

Dictionaries built by hand
-(Passes: Lookup, Vocabulary, Phrases)

-

16-18

 -

Money

 -

Money phrases and lists
-(Passes: Money, MoneyAttributes, ConjMoney)

-

19-21

 -

Companies

 -

Company phrases
-(Passes: Companies, CompanyAttributes, CompanyDet)

-

22-23

 -

Anaphora

 -

Words that reference previous locations in a text
-(Passes: Anaphora, AnaphoraCompany)

-

24

 -

Company Conj

 -

List of companies in text
-(Pass: ConjCompany)

-

25

 -

Events

 -

Buy, Acquire, and Stock events
-(Pass: Events)

-

26-28

 -

Meta Events

 -

Event attributes and phrases that refer back to other - events
-(Passes: EventAttributes, AnaphoraEvent, Events2)

-

29

 -

Output

 -

Outputting the data to a file
-(Pass: Output)

-
- - - - diff --git a/Help/helps/HowToBuild/DictLookupURL/DictionaryLookup.htm b/Help/helps/HowToBuild/DictLookupURL/DictionaryLookup.htm deleted file mode 100644 index f0fea92..0000000 --- a/Help/helps/HowToBuild/DictLookupURL/DictionaryLookup.htm +++ /dev/null @@ -1,583 +0,0 @@ - - - - - - - - - - -Constructing URLs - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

URLs for Dictionary - Lookup

- -

The Dictionary Lookup Tool is a very handy - tool for building and refining lexicons.  This - tool allows users to easily access dictionary sites on the Web and to - display search results within VisualText.  

- -

To access online dictionaries, the paths to - these sites must be set in VisualText - preferences located under the File Menu.  

- -

List of Commonly Used Online Dictionary Sites

- -

A list of commonly used online dictionary sites and the paths to their - search pages is given below.  To - get access to one of these dictionaries, create an entry in the Dictionary - Lookup preferences under Name and copy the URL path into URL input panel. -   To - create URL paths for other dictionaries, follow the steps outlined below.

- - - --- - - - - - - - - - - - - - - - - - - - -
-

Online Site

 -

URL Path

-

Wordnet

 -

http://www.cogsci.princeton.edu/cgi-bin/webwn/?stage=1&word=

-

Merriam Webster

 -

http://www.m-w.com/cgi-bin/dictionary?book=Dictionary&va=

-

Dictionary.com

 -

http://www.dictionary.com/cgi-bin/dict.pl?term=

-

Google

 -

http://www.google.com/search?btnG=Google+Search&lr=lang_en&as_q=

- - - - - - -
-

Note: The Google path returns pages for English - only.

- -

Once set, the dictionaries can be accessed - from the main Tools - Menu, and from the Tools menu in both the Text - Tab Popup Menu and the Text - File Popup Menu.  Dictionaries - can also be accessed from the Attribute - Editor.

- -

Constructing - Simple URLs

- -

Constructing the URL for an online dictionary - requires some knowledge of CGIs, HTML and FORMS.  If - you are familiar with these topics, constructing the URL path for an online - source is fairly routine.  You - may want to familiarize yourself with these topics before trying to construct - a URL path.

- -

To construct the URL path that can be used - by Dictionary Lookup, you need to do three things:

- -
    - -
  1. Discover - the CGI path of the online source;
    2. - -
  2. Discover - the names of the variables that need to accompany the CGI to do the lookups;
    3. - -
  3. Put the - word variable at the end of the URL you are constructing.
    4. -
- -

These steps are outlined below using the Google - website (http://www.google.com) as an example.

- -

Discover the CGI Path

- -

The simplest way to discover the CGI path is - to do a search for a word on the desired site, and then look at the resulting - page's URL path found at the top of your browser.

- - - -

- -

In the Address portion of the browser, the - desired URL is displayed.

- -

- -

The syntax for a CGI path specifies the name - for a CGI program just before the ? sign.  In - this case, we find the CGI name to be "search".

- -

Discover Variable Names - to Accompany CGI

- -

Variables for word lookup are specified after - the question mark "?".   Key-value - pairs can be found separated by the ampersand character "&". - Each key-value pair is further separated by an equal sign "=". -

- -

The first key-value pair (q=baas) is our search - term "baas" and the second key-value pair (btnG=Google+Search) - is the Google convention used to indicate an unknown key-value pair.  

- -

From the URL in the Address portion of the - browser above, we have determined the following:

- - - -

The plus sign "+" in the last key-value - pair is a CGI convention for a "space" or "blank" - character.

- -

Put Word Variable at End - of URL

- -

To construct the URL in a format that can be - used by Dictionary Lookup, we need to rearrange the order of the key-value - pairs.  The - value for the word key-value pair must be removed and it must be placed - at the end of path.

- -

In our Google example, we discovered that the - word key-value pair is q=bass.  The - "bass" portion of the key-value pair must be removed, and "q=" - should be placed at the end of our URL.  The - unknown key-value pair along with the CGI  name - can stay in the same position.  The - resulting URL looks like this:

- -

http://www.google.com/search?btnG=Google+Search&q=

- - - - - - -
-

Note: - Remember to put the question mark after - the CGI Path Name and an ampersand "&" between the key-value - pairs.

- -

When you copy the URL into the Dictionary Lookup - preference and give it the name Google, Google becomes one of the dictionaries - listed in the list of dictionary choices, allowing you access Google directly - from within VisualText.

- -

Constructing Advanced URLs

- -

To construct URLs for advanced or more complicated - searches such as returning only pages in English, you will have to add - additional information to the URL path.  This - information can be extracted from the HTML and FORMS used to create the - webpage.  Often - it is necessary to use HTML and FORMS since it is difficult to extract - key-value pair information as we did above for simple searches.

- -

To illustrate - how you can extract this information, we will use the same Google example - and create a URL for searches that are restricted to English pages only. -  (The Google - search page is located at http://www.google.com)

- -

Google Advanced Search and CGI Path

- -
    - -

  1. From the Google search - page, click on "Advanced Search".  

    2. - -

  2. Type in the word baas and restrict the search to return - only pages written in English. -  (We have - also restricted the search to 10 results.)

    3. -
- -

- -
    - -

  1. Submit the query and - inspect the Address portion of the browser.  

    2. -
- -

The CGI path is quite long and trying to discover - what key-value pairs are needed to restrict the search to English pages - only will be very difficult.  To - find out what part of the CGI path is used to specify only English pages, - you can use the View Source option on the browser to inspect the HTML - source code.

- -

Using View Source

- -

The HTML source code can be viewed for the - webpage by right-clicking on the webpage and selecting View Source from - the menu options.  You - can search for the specific FORM items you are trying to restrict.  

- - - -

First, we need to verify the key name for the - search word. Sometimes, it will be the same as for the simple search. - However, depending on the CGI, the name may be different.  

- - - -

- -

The key name for the word field is as_q. -  Recall, - that the key name used for the simple search is simply q. -  

- -

Next, we need to find the SELECT form item - for the pulldown menu that will specify a particular language.

- - - -

- -

The key name for the language SELECT form is - lr.

- -

To find out what key is used to specify English - as the desired language we must search for OPTION.  

- - - -

- -

This gives us the needed value for restricting - our search to English: lang_en. - We put this together with the SELECT key name "lr" and we get: - lr=lang_en.

- -

With these three pieces of information, advanced - search value, select language value, language option value, you can specify - a URL in Dictionary Lookup which specifies pages in English for Google. -  

- -

http://www.google.com/search?btnG=Google+Search&lr=lang_en&as_q=

- - - - diff --git a/Help/helps/HowToBuild/Tour15/AnalyzerTab/Tour_AnalyzerTab.htm b/Help/helps/HowToBuild/Tour15/AnalyzerTab/Tour_AnalyzerTab.htm deleted file mode 100644 index d616c39..0000000 --- a/Help/helps/HowToBuild/Tour15/AnalyzerTab/Tour_AnalyzerTab.htm +++ /dev/null @@ -1,250 +0,0 @@ - - - - - - - - - - -Tour AnalyzerTab - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Parse Tree

 -

Quick Tour
-Ana Tab

 -

Pass Files

-
- - -
- - -

The Analyzer

- -

Click on the Ana Tab button to display the Ana Tab Window.

- -

This window maintains the ordered sequence of passes that define the - current text analyzer. System or built-in pass types (e.g., tokenize) - are represented by flat "DNA" icons. Passes with associated - pass files have twisted DNA icons. Our corporate analyzer has 29 passes:

- -

- -

Stubs

- -

You may have noticed that there are two passes with red circles. These - represent "stubs" or placeholders in which automatically generated - passes are deposited. We'll talk about automatic rule generation and the - Gram Tab in a later section:

- -

Double Click

- -

The most common action in the Ana Tab is to open up a pass file by double-clicking - on the desired pass:

- -

- -

Right-Click and Toolbar

- -

The Ana tab has an extensive right-click menu as well as arrows in the - Tab Toolbar that serve for deleting or moving analyzer passes:

- -

- -

 

- -

- - - -
- - -

Next Section: - Pass Files -

- - - - diff --git a/Help/helps/HowToBuild/Tour15/Debug/Tour_Debug.htm b/Help/helps/HowToBuild/Tour15/Debug/Tour_Debug.htm deleted file mode 100644 index f485693..0000000 --- a/Help/helps/HowToBuild/Tour15/Debug/Tour_Debug.htm +++ /dev/null @@ -1,264 +0,0 @@ - - - - - - - - - - -Tour Debug - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Gram - Tab

 -

Quick Tour
-Debugging

 -

KB

-
- - -
- - -

Examining the Analyzer Under Construction

- -

There are numbers of tools and methods for debugging analyzers, including - highlighting and what is called "Page Mode".

- -

Highlighting

- -

Highlighting is a simple and intuitive tool for debugging - analyzers. Turn on the highlighting toggle in the toolbar:

- -

- -

Then open up Sold.txt and select the Ana Tab window. - Single click on various analyzer passes. The matched text for each pass - will be highlighted in blue and green, letting you know which words and - phrases matched in that pass:

- -

- -

Page Mode

- -

Another important debugging tool is called "Page Mode":

- -

- -

Among other things, it allows you to examine the change in parse trees - for each pass.

- -

With the "Page Mode" on, choose the phrase - "$120 million in cash and $14.3 million in assumed debt" and - view its tree. Then in the Ana Tab, click on pass 4, "join", - and travel down the analyzer passes observing the changes in the parse - tree:

- -

- - - -
- - -

Next Section: - KB

- - - - diff --git a/Help/helps/HowToBuild/Tour15/Find/Tour_Find.htm b/Help/helps/HowToBuild/Tour15/Find/Tour_Find.htm deleted file mode 100644 index a6cfcdb..0000000 --- a/Help/helps/HowToBuild/Tour15/Find/Tour_Find.htm +++ /dev/null @@ -1,195 +0,0 @@ - - - - - - - - - - -Tour Find - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

 KB

 -

Quick Tour
-Find

 -

 

- - - -
- - -

Find Window

- -

One other important window in the VisualText interface is the Find window. - You can search for a word or phrase in specified files and in other analyzers. - Below, we search for the concept "_company" using Find in Files - :

- -

- -

Double-Click

- -

Once you have searched on a word, the find window allows you to go directly - to an instance of the word or phrase by double-clicking on the line of - interest:

- -

- -

This will take you directly to the window and highlight the instance:

- -

- - - -
- - -

Hope you enjoyed the - tour of VisualText!

- - - - diff --git a/Help/helps/HowToBuild/Tour15/GramTab/Tour_GramTab.htm b/Help/helps/HowToBuild/Tour15/GramTab/Tour_GramTab.htm deleted file mode 100644 index b5eb946..0000000 --- a/Help/helps/HowToBuild/Tour15/GramTab/Tour_GramTab.htm +++ /dev/null @@ -1,236 +0,0 @@ - - - - - - - - - - -Tour GramTab - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Pass Files

 -

Quick Tour
-Gram Tab

 -

Debugging

-
- - -
- - -

Automatic Rule Generation (RUG)

- -

In VisualText, rules can be automatically generated from text samples. - In our corporate analyzer, we use the Gram Tab much like a "dictionary". - More complex samples with subsamples can also be managed in the Gram Tab.

- -

- -

Stubs

- -

Stubs are place holders where the automatically generated rules will - appear in the Analyzer Sequence. Notice that the icon for words above - matches the "stub_words" pass below, taken from the Analyzer - Tab:

- -

- -

Double-Click

- -

If you double-click on a sample such as "New Hampshire Medical", - the text file of the sample will pop up in the Workspace. In our corporate - analyzer, we are using sample files for our dictionaries. One can imagine - using large lists of words and phrases in files that can be converted - into rules automatically:

- -

- -

If you double click on a rule concept (one with a right-pointing arrow) - such as "company", you can see its corresponding pass file containing - automatically generated rules:

- -

- - - -
- - -

Next Section: - Debugging

- - - - diff --git a/Help/helps/HowToBuild/Tour15/Intro/Tour_Intro.htm b/Help/helps/HowToBuild/Tour15/Intro/Tour_Intro.htm deleted file mode 100644 index 84d03d8..0000000 --- a/Help/helps/HowToBuild/Tour15/Intro/Tour_Intro.htm +++ /dev/null @@ -1,699 +0,0 @@ - - - - - - - - - - -Tour Intro - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 TOC

 -

Quick Tour
-Intro

 -

Screen Layout

-
- - -
- - -

VisualText Features and Benefits

- -

VisualText is an Integrated Development Environment for building deep - text analysis applications. It is like Visual C++(R), but with specializations - for Natural Language Processing. Some features and benefits of VisualText:

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FEATURE

 -

BENEFIT

-

Integrated development environment (IDE)

 -

Reduces resources needed to build text analyzers

-

Rich, integrated GUI tool set and data views

 -

Accelerates development of deep, accurate - analyzers

-

NLP++ Programming Language

 -

Anything thinkable can be programmed! Users - focus on the task and heuristics rather than programmatic detail

-

Hierarchical Knowledge - Base Management System

 -

Easy to build ontologies, dictionaries, semantics, - and any other desired schemas.

-

Automatic Rule Generation (RUG)

 -

Easy and quick to maintain and enhance analyzers - merely by highlighting text.

-

NLP++ integrates code, rules, parse trees, - and KB

 -

Easy to access/modify knowledge.

-

Interpreted NLP++ execution

 -

Blazing fast modify-and-test development cycle

-

Compilation of analyzer and KB

 -

Fast execution in runtime environment

-

Excellent GUI support for debugging analyzers

 -

Reduced time and effort to debug analyzers

-

Multi-pass analyzers

 -

Modular, maintainable, readable, extensible

-

Single parse tree

 -

No combinatorial explosion; results in efficient - analyzers.

-

Paradigm independent, integrates multiple - paradigms and styles of text analysis

 -

Enables developer to use and combine the most - appropriate methods for the task at hand

-

User project for adding C++ code

 -

Open architecture for easy integration with - 3rd party software, extension of the core capability by the customer.

-

Rules operate in specific contexts

 -

Higher precision, faster execution.

-

Built-in confidence operator

 -

Easy to compute and assess confidence

-

 

 -

 

- -

Sample Applications

- - - -

GUI

- -

VisualText's comprehensive interface allows you to focus on the main - goal: quickly building an accurate and complete text analyzer.

- -

- -

NLP++ Language

- -

NLP++ is a new C++ -like programming language specially designed for - building deep text analyzers:

- -

- -

Knowledge Base

- -

VisualText integrates a powerful and flexible Knowledge Base that allows - analyzers to be as smart as you want them to be:

- -

- -

Debugging Tools

- -

A suite of debugging tools lets you highlight rule matches, examine - parse trees, and study your analyzer one step at a time:

- -

- -

Flexible, Easy to Maintain, and Fast

- -

VisualText is not tied to one style or paradigm of text analysis. If - you can imagine a text analyzer, then VisualText can build it!

- -

Being "visual" means easy to build and maintain. VisualText - displays text analyzers in such a way as to make it easy for anyone to - modify or extend existing analyzers.

- -

And finally, VisualText is fast. It can substantially cut the cost and - time needed to build text analyzers. Text analyzers under development - are run in interpreted mode, so that you can quickly modify and test your - analyzer. Once a text analyzer is built, you can compile it to run faster.

- - - -
- - -

Next Section: - Screen Layout

- - - - diff --git a/Help/helps/HowToBuild/Tour15/Kb/Tour_KB.htm b/Help/helps/HowToBuild/Tour15/Kb/Tour_KB.htm deleted file mode 100644 index 416105f..0000000 --- a/Help/helps/HowToBuild/Tour15/Kb/Tour_KB.htm +++ /dev/null @@ -1,214 +0,0 @@ - - - - - - - - - - -Tour KB - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Debugging

 -

Quick Tour
-Knowledge Base

 -

Find

-
- - -
- - -

Knowledge Base

- -

A Knowledge Base or "KB" is like a database but differs in - several respects: it is hierarchical and is managed dynamically - by the analyzer. Databases are designed first, then used by programs. - Knowledge Bases are used dynamically with text analyzers and change as - the analyzer processes a text.

- -

Open the KB by selecting the KB Editor toolbar button. - Open concepts by clicking the + to their left:

- -

- -

Double-Click

- -

The corporate analyzer makes heavy use of the KB. Above, there are two - areas under the "corporate" concept: "companies", - and "parse". The companies area is static and is used in looking - up synonyms of company names. The "parse" area is used by the - analyzer to reconstruct the "corporate" world by creating concepts - under each sentence concept, in tandem with processing an input text.

- -

Below, we show the result of double-clicking on the "acquire" - concept under sentence 5, then double-clicking on the "comment" - concept. Note that the analyzer has created attributes that describe the - "acquire" event:

- -

- - - -
- - -

Next Section: - Find

- - - - diff --git a/Help/helps/HowToBuild/Tour15/Load/Tour_Load.htm b/Help/helps/HowToBuild/Tour15/Load/Tour_Load.htm deleted file mode 100644 index 86048aa..0000000 --- a/Help/helps/HowToBuild/Tour15/Load/Tour_Load.htm +++ /dev/null @@ -1,296 +0,0 @@ - - - - - - - - - - -Tour Load - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Screen - Layout

 -

Quick Tour
-Loading the Analyzer

 -

Log Window

-
- - -
- - -

Open the "Corporate" Analyzer

- -

(Actions for you to take are flagged by a yellow arrow - at the start of a line.) Now it's time for you to load the Corporate Analyzer, - which can be done in two ways. The first starts with double-clicking the - desktop icon for VisualText:

- -

- -

The second is by double-clicking the icon - for a particular VisualText analyzer. The icon for the Corporate analyzer - looks like:

- -

 

- -

Method One: Loading an analyzer within - VisualText

- -

Open VisualText. Next, choose "Open Analyzer" - under the main menu "File":

- -

- -

You are then prompted to choose an analyzer. Browse to - c:\Program Files\TextAI\VisualText\apps\corporate and open the file named - corporate.ana (it may be listed as corporate):

- -

- -

Method Two

- -

Locate the corporate analyzer in your Windows File Explorer - window. Then double-click on the VisualText analyzer file (.ana). This - will invoke VisualText and load the corporate analyzer:

- -

- -

You are Now Ready to Start!

- -

Once you have VisualText and the Corporate Analyzer loaded, we can continue - our tour.

- - - -
- - -

Next Section: - Log Window

- - - - diff --git a/Help/helps/HowToBuild/Tour15/Log/Tour_Log.htm b/Help/helps/HowToBuild/Tour15/Log/Tour_Log.htm deleted file mode 100644 index fb7b068..0000000 --- a/Help/helps/HowToBuild/Tour15/Log/Tour_Log.htm +++ /dev/null @@ -1,173 +0,0 @@ - - - - - - - - - - -Tour Log - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Load - Analyzer

 -

Quick Tour
-Log Window

 -

Text Tab

-
- - -
- - -

Log Window

- -

In the bottom left-hand corner resides the Log Window, which monitors - the status of the currently loaded analyzer. Analyzer load and run errors - are shown in the Log Window.

- -

- - - -
- - -

Next Section: - Text Tab

- - - - diff --git a/Help/helps/HowToBuild/Tour15/Output/Tour_Output.htm b/Help/helps/HowToBuild/Tour15/Output/Tour_Output.htm deleted file mode 100644 index 00faccf..0000000 --- a/Help/helps/HowToBuild/Tour15/Output/Tour_Output.htm +++ /dev/null @@ -1,228 +0,0 @@ - - - - - - - - - - -Tour Output - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Running

 -

Quick Tour
-Output

 -

Parse Tree -

-
- - -
- - -

Examining the Output

- -

After running the corporate analyzer on the input text, - you can now examine the output in either of two ways. One is to click - on the "Output" button in the Debug Toolbar:

- -

- -

The second is to click the button next to the Output button and choose - the "output" file from the available dump files:

- -

- -

The Output File

- -

The output below shows the purpose of the Corporate analyzer, which - is to extract and correlate key business information from the Sold.txt - file. The name "output.txt" is a special file name in VisualText, - that has its own button on the toolbar. If the analyzer does not output - to a file called "output.txt", the button will be greyed out. - Below is a portion of the output file from our Corporate analyzer. Take - some time to examine the entire content of the output file in VisualText.

- -

- - - -
- - -

Next Section: - Parse Tree -

- - - - diff --git a/Help/helps/HowToBuild/Tour15/ParseTrees/Tour_ParseTrees.htm b/Help/helps/HowToBuild/Tour15/ParseTrees/Tour_ParseTrees.htm deleted file mode 100644 index 107c3fa..0000000 --- a/Help/helps/HowToBuild/Tour15/ParseTrees/Tour_ParseTrees.htm +++ /dev/null @@ -1,227 +0,0 @@ - - - - - - - - - - -Tour ParseTrees - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Output

 -

Quick Tour
-Parse Trees

 -

Ana Tab -

-
- - -
- - -

Parse Trees

- -

VisualText builds multiple pass analyzers in which each pass uses the - parse tree as modified by the - previous pass. A parse tree is a data structure that records and groups - patterns discovered in the input text. The elements of a parse tree are - called nodes. Each line in the parse tree display below represents one - node, or phrase, within the parse tree. Nodes may carry information in - the form of node variables and values.

- -

Tree Buttons

- -

Select the "events2" pass in the Ana Tab, then - click the "View Parse Tree" button in the Debug Toolbar, as - shown below.

- -

In this way, you can view the parse tree for the pass that is currently - selected in the Ana Tab. (Note: If the text has been analyzed with Generate - Logs off, then the final parse tree is displayed.) The display shows patterns - and groupings identified in the input text, as well as the node variables - and their values as parenthesized lists of the form ("variable_name" - "value1" "value2" ...).

- -

- -

Viewing Partial Trees:

- -

VisualText also allows users to display partial parse trees for selected - parts of the text. The phrase "$14.3 million in assumed debt" - is highlighted in the text window and displayed in a parse tree window, - below.

- -

- - - -
- - -

Next Section: - Ana Tab -

- - - - diff --git a/Help/helps/HowToBuild/Tour15/PassFiles/Tour_PassFiles.htm b/Help/helps/HowToBuild/Tour15/PassFiles/Tour_PassFiles.htm deleted file mode 100644 index 14e8315..0000000 --- a/Help/helps/HowToBuild/Tour15/PassFiles/Tour_PassFiles.htm +++ /dev/null @@ -1,356 +0,0 @@ - - - - - - - - - - -Tour PassFiles - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Ana Tab

 -

Quick Tour
-Pass Files

 -

Gram Tab

-
- - -
- - -

Heart of the Analyzer

- -

The heart of the text analyzer lies in the pass files, which contain - NLP++ code and rules.

- -

Pass File Zones, Regions, and Selectors

- -

Pass file areas are marked by words that are blue, all-caps, and start - with an "@". Three commonly used regions are detailed below.

- - - - - - - - - - - - - - - - - - - - - - - -
-

AREA

 -

DESCRIPTION

 -

EXAMPLE BELOW

-

@PATH

 -

A Selector that defines the context in which - rules will be tried.

 -

In the corporate analyzer, most paths are - defined as applying to sentences under paragraphs,under the root node - of the parse tree.

-

@RULES

 -

Marks the start of a Rule Region. An NLP++ - rule consists of a pattern (called a phrase) that is matched against the - parse tree.

 -

Below, the rules match patterns for constructs such as $120 million - or 160 million dollars.

-

NOTE: words with underscores (_money) are nonliterals, or abstract - tokens, while those without match raw text ("dollars");

-

@POST

 -

When a rule matches, NLP++ code in the Post - Region tells the rule matcher what actions to take. See the _money rules - below.

 -

Below, the numeric value for the money phrase - is calculated and stored on the suggested node.

- -

 

- -

- -

Code Region

- -

An NLP++ pass file includes a CODE Region that is independent of any - rule matching. One example of this is in the third pass "initKB", - where the analyzer constructs needed concepts in the Knowledge Base before - processing the text:

- -

- - - -
- - -

Next Section: - Gram Tab

- - - - diff --git a/Help/helps/HowToBuild/Tour15/Run/Tour_Run.htm b/Help/helps/HowToBuild/Tour15/Run/Tour_Run.htm deleted file mode 100644 index 311d5ef..0000000 --- a/Help/helps/HowToBuild/Tour15/Run/Tour_Run.htm +++ /dev/null @@ -1,285 +0,0 @@ - - - - - - - - - - -Tour Run - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Text - Tab

 -

Quick Tour
-Running the Analyzer

 -

Output

-
- - -
- - -

Running the Analyzer on Text

- -

First, select the text you want to process, in the Text - Tab:

- -

- -

To view the text, either double-click on the text icon - (above) in the Text Tab, or go to the Debug Toolbar and click on the "paper" - button:

- -

- -

The text window pops up in the workspace:

- -

- -

The Run Button

- -

To run the analyzer on the selected input text, choose - the running-man button in the Workspace Toolbar:

- -

- -

The cursor becomes an hour glass and the stop button is activated:

- -

- -

The Log Window will display stats about the run, including any syntax - errors it discovered:

- -

- -

You have successfully run the corporate analyzer on our "Sold.txt".

- - - -
- - -

Next Section: - Output

- - - - diff --git a/Help/helps/HowToBuild/Tour15/Screen/Tour_Screen.htm b/Help/helps/HowToBuild/Tour15/Screen/Tour_Screen.htm deleted file mode 100644 index f200d5a..0000000 --- a/Help/helps/HowToBuild/Tour15/Screen/Tour_Screen.htm +++ /dev/null @@ -1,492 +0,0 @@ - - - - - - - - - - -Tour Screen - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Intro

 -

Quick Tour
-Screen Layout

 -

Loading

-
- - -
- - -

Screen Layout

- -

The general screen layout of VisualText includes Toolbars, the Main - Menu, the Tab Window, the Workspace, and the Log and Find Windows:

- -

- -

The principal controls are the Main Menu, the Toolbars, and the Tabs.

- -

Main Menu

- -

The main menu is a superset of the buttons available in the Toolbars:

- -

- -

Toolbars

- -

VisualText has 5 toolbars: Main, Browser, Tab, Knowledge Base, and Debug:

- -

- -

The table below describes each Toolbar:

- - -
- - - - - - - - - - - - - - - - - - - - - - - - - -
-

TOOLBAR

 -

DESCRIPTION

-

Main

 -

Standard functions found in all GUI applications, - including open, close, cut, and paste

-

Browser

 -

Functions for managing web browser windows

-

Tab

 -

Functions for managing the Gram, Ana, and - Text Tabs

-

Workspace

 -

Functions for the Analyzer and Knowledge Base

-

Debug

 -

Tools for examining the input text file, trees, - pass file, etc.

-
-

 

- -

Tab Window

- -

The Tab Window area accesses three key windows: the Gram Tab, Ana Tab, - and Text Tab.

- -

- -

The table below describes each:

- - -
- - - - - - - - - - - - - - - - - -
-

TAB - WINDOW

 -

DESCRIPTION

-

Gram

 -

Manage user-added text samples for automatic - rule generation

-

Ana

 -

Manage the analyzer sequence and pass files

-

Text

 -

Manage input files, output files, and other - folders and files

-
-

Right Click Menus

- -

Right-click popup menus play an important role in VisualText. The Gram - Tab, Ana Tab, Text Tab, Knowledge Base, Pass File Windows, Parse Tree - Windows, Text Windows all employ right-click menus to provide context-sensitive - functions quickly. Below is the right-click menu for the Text Tab:

- -

- - - -
- - -

Next Section: - Loading

- - - - diff --git a/Help/helps/HowToBuild/Tour15/TextTab/Tour_TextTab.htm b/Help/helps/HowToBuild/Tour15/TextTab/Tour_TextTab.htm deleted file mode 100644 index 7e67ec8..0000000 --- a/Help/helps/HowToBuild/Tour15/TextTab/Tour_TextTab.htm +++ /dev/null @@ -1,243 +0,0 @@ - - - - - - - - - - -Tour TextTab - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - -
-

 Log

 -

Quick Tour
-Text Tab

 -

Running

-
- - -
- - -

Choosing the Text to Process

- -

Click the Text Tab and open the Dev folder within there.

- -

If Sold.txt has no red checkmark, then click Generate - Logs in the Analyzer menu and select Run from the Analyzer menu.

- -

The Text Tab is a repository for texts to be processed by your analyzer. - You can organize text files in folders and run the analyzer on individual - files or on entire subtrees of folders and files. Inside the "Dev" - folder, we've selected the text file "Sold.txt" for processing:

- -

- -

Choosing Previously Processed Text - Files

- -

In the Tab Toolbar,a recent-text pulldown provides quick access to the - most recently selected texts. This is useful when the Text Tab holds many - files:

- -

- -

Processed Text

- -

When a text has been processed, a "+" will appear in front - of the icon and the icon will have a checkmark. You can open the text - file icon and see the log files produced during the analysis (or "run") - of the text:

- -

- - - -
- - -

Next Section: - Running

- - - - diff --git a/Help/helps/HowToBuild/Tour15/Toc/Tour_TOC.htm b/Help/helps/HowToBuild/Tour15/Toc/Tour_TOC.htm deleted file mode 100644 index 453b67d..0000000 --- a/Help/helps/HowToBuild/Tour15/Toc/Tour_TOC.htm +++ /dev/null @@ -1,405 +0,0 @@ - - - - - - - - - - -Tour TOC - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

The VisualText Quick Tour
-Table of Contents

- - -
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Title

 -

Description

-

Intro

 -

Introduction to VisualText's Power

-

Screen - Layout

 -

General layout of the Graphic User Interface

-

Loading Analyzer

 -

Load the demonstration analyzer "Corporate"

-

Log

 -

Monitoring the Log File

-

Text - Tab

 -

Text for developing text analyzers

-

Running the - Analyzer

 -

Run the analyzer on the text

-

Output

 -

Examining the output

-

Parse - Trees

 -

Structures built by the analyzer

-

Ana - Tab

 -

Building the analyzer sequence

-

Pass - Files

 -

Rules and Code in pass files

-

Gram - Tab

 -

Automatic Rule Generation

-

Debugging

 -

Examining the details of the analyzer at work

-

Knowledge Base

 -

Exploring the Knowledge Base

-

Find

 -

Finding words and concepts in files

-
- - - - diff --git a/Help/helps/Image Files/mainToolbar.gif b/Help/helps/Image Files/mainToolbar.gif deleted file mode 100644 index b42b5ea..0000000 Binary files a/Help/helps/Image Files/mainToolbar.gif and /dev/null differ diff --git a/Help/helps/Introduction/Intellectual_Property_Notices.htm b/Help/helps/Introduction/Intellectual_Property_Notices.htm deleted file mode 100644 index 5793774..0000000 --- a/Help/helps/Introduction/Intellectual_Property_Notices.htm +++ /dev/null @@ -1,263 +0,0 @@ - - - - - - - - - - -Intellectual Property Notices - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Intellectual Property Notices

- - - -
- - -

VisualText Help Copyright 2007 Text Analysis International, Inc. - All rights reserved.

- -

VisualText Copyright 1999 - 2007 Text Analysis International, Inc. - All rights reserved.

- -

NLP++ Copyright 2000 - 2007 Text Analysis International, Inc. All - rights reserved.

- -

Conceptual Grammar Copyright 1999 - 2007 Text Analysis International, - Inc. All rights reserved.

- -

Patents Pending by Text Analysis - International, Inc.

- - - -
- - -

WordNet 1.6

- -

Copyright 1997 by Princeton University. All rights reserved.

- -

THIS SOFTWARE AND DATABASE IS PROVIDED "AS IS" AND PRINCETON - UNIVERSITY MAKES NO REPRESENTATIONS OR WARRANTIES, EXPRESS OR IMPLIED. - BY WAY OF EXAMPLE, BUT NOT LIMITATION, PRINCETON UNIVERSITY MAKES NO REPRESENTATIONS - OR WARRANTIES OF MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE - OR THAT THE USE OF THE LICENSED SOFTWARE, DATABASE OR DOCUMENTATION WILL - NOT INFRINGE ANY THIRD PARTY PATENTS, COPYRIGHTS, TRADEMARKS OR OTHER - RIGHTS.

- - - -
- - -

Microsoft Visual C++ , .NET, Microsoft and Windows are registered trademarks - of Microsoft Corporation. Other brands and their products are trademarks - or registered trademarks of their respective holders and should be noted - as such.

- - - -
- - -

ICU License - ICU 1.8.1 and later

- -

COPYRIGHT AND PERMISSION NOTICE
- 

- -

Copyright (c) 1995-2005 International - Business Machines Corporation and others

- -

All rights reserved.

- -

 

- -

Permission is hereby granted, free of - charge, to any person obtaining a

- -

copy of this software and associated documentation - files (the

- -

"Software"), to deal in the - Software without restriction, including

- -

without limitation the rights to use, - copy, modify, merge, publish,

- -

distribute, and/or sell copies of the - Software, and to permit persons

- -

to whom the Software is furnished to do - so, provided that the above

- -

copyright notice(s) and this permission - notice appear in all copies of

- -

the Software and that both the above copyright - notice(s) and this

- -

permission notice appear in supporting - documentation.

- -

 

- -

THE SOFTWARE IS PROVIDED "AS IS", - WITHOUT WARRANTY OF ANY KIND, EXPRESS

- -

OR IMPLIED, INCLUDING BUT NOT LIMITED - TO THE WARRANTIES OF

- -

MERCHANTABILITY, FITNESS FOR A PARTICULAR - PURPOSE AND NONINFRINGEMENT

- -

OF THIRD PARTY RIGHTS. IN NO EVENT SHALL - THE COPYRIGHT HOLDER OR

- -

HOLDERS INCLUDED IN THIS NOTICE BE LIABLE - FOR ANY CLAIM, OR ANY SPECIAL

- -

INDIRECT OR CONSEQUENTIAL DAMAGES, OR - ANY DAMAGES WHATSOEVER RESULTING

- -

FROM LOSS OF USE, DATA OR PROFITS, WHETHER - IN AN ACTION OF CONTRACT,

- -

NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING - OUT OF OR IN CONNECTION

- -

WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.

- -

 

- -

Except as contained in this notice, the - name of a copyright holder

- -

shall not be used in advertising or otherwise - to promote the sale, use

- -

or other dealings in this Software without - prior written authorization

- -

of the copyright holder.

- -

-----

- -

All trademarks and registered trademarks - mentioned herein are the property of their respective owners.

- - - -
- - -

 

- - - - diff --git a/Help/helps/Introduction/System_Requirements.htm b/Help/helps/Introduction/System_Requirements.htm deleted file mode 100644 index 921a1d3..0000000 --- a/Help/helps/Introduction/System_Requirements.htm +++ /dev/null @@ -1,59 +0,0 @@ - - - - - - - - - - -System Requirements - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

System Requirements

- -

VisualText runs on PCs with Windows: NT, Windows 2000, Win98, and WinME.

- -

Analyzers built with VisualText can run stand-alone or embedded in larger - applications. The runtime system has also been ported to a Redhat Linux - 8.0 environment with a Gnu C++ compiler. Support for Linux and other environments - that compile C++ can be arranged with Text Analysis International, Inc.

- -

For VisualText, we recommend a machine running at - 500MHz or greater, with 256MB RAM (128MB minimum), and 250MB available - hard disk. Minimum requirements for monitor are 15" with a resolution - of 1024x768 or better.

- -

 

- - - - diff --git a/Help/helps/Introduction/VTWorkFlow.gif b/Help/helps/Introduction/VTWorkFlow.gif deleted file mode 100644 index a466118..0000000 Binary files a/Help/helps/Introduction/VTWorkFlow.gif and /dev/null differ diff --git a/Help/helps/Introduction/VisualText_Quick_Tour.htm b/Help/helps/Introduction/VisualText_Quick_Tour.htm deleted file mode 100644 index 6e0ae1b..0000000 --- a/Help/helps/Introduction/VisualText_Quick_Tour.htm +++ /dev/null @@ -1,84 +0,0 @@ - - - - - - - - - - -VisualText Quick Tour - - - - - - - - - - - - - - - - - - - - - - - - - - - -

VisualText Quick Tour

- -

Welcome to the VisualText Quick Tour, which introduces the major components - of VisualText and shows how these components work together to create, - run, and debug a text analyzer.

- -

The Quick Tour walks you through a working analyzer called the Corporate - Analyzer. The Corporate Analyzer demonstrates the features of VisualText - and introduces you to building analyzers using VisualText. It is intended - as a demonstration analyzer only and is meant to be used with the sample - input files provided.

- -

In addition to the Quick Tour, we have provided a more detailed tour - of the Corporate Analyzer in the Sample Analyzers "section" (or - "folder" in the Help). The full tour takes you through the step-by-step - process used in building the Corporate Analyzer.

- -

We recommend that you run the Corporate Analyzer application as you - go through both the Quick Tour and the more detailed Corporate Analyzer - tour. The Corporate Analyzer application is located in the APPS folder - under your VisualText installation directory, in a folder named corporate.

- -

To run the analyzer, follow the steps outlined below:

- -
    - -

  1. Launch VisualText.

    2. - -

  2. From the File - menu, select Open Analyzer.

    3. - -

  3. In the Open - dialog, navigate to the top level corporate folder (c:\Program Files\TextAI\VisualText\apps\corporate, - by default.)

    4. - -

  4. Select the corporate.ana - file and click Open.

    5. -
- -

Start - the Quick Tour.

- - - - diff --git a/Help/helps/Introduction/What_is_VisualText.htm b/Help/helps/Introduction/What_is_VisualText.htm deleted file mode 100644 index f77b21e..0000000 --- a/Help/helps/Introduction/What_is_VisualText.htm +++ /dev/null @@ -1,81 +0,0 @@ - - - - - - - - - - -What is VisualText - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

What is VisualText?

- -

VisualText is an integrated development environment (IDE) for creating - text analyzers. What is a text analyzer? Any program that takes as input - text data and produces an output result. Examples include programs that - extract information to populate structured business databases, web page - categorizers, email routers and autoresponders, chat managers, analyzers - for text-from-speech, grammar checkers, and spell checkers. VisualText - is very well suited to information extraction - systems, that is, systems that find and correlate the critical information - in a text. Information extraction systems typically produce records (e.g. - XML) suitable for populating a database.

- -

The figure below shows the development process using VisualText. A developer, - using the VisualText IDE, creates a text analyzer. An input text file - is given to the text analyzer, which processes the text and outputs a - result.

- -

- -

VisualText encompass several language-processing methods as well as - mixtures of methods. VisualText is an open architecture, enabling users - to extend its core algorithms and functions. This flexibility enables - users to conveniently and quickly develop text analyzers suited to their - own purposes.

- -

 

- - - - diff --git a/Help/helps/Introduction/Why_Use_It_.htm b/Help/helps/Introduction/Why_Use_It_.htm deleted file mode 100644 index a3cfbf7..0000000 --- a/Help/helps/Introduction/Why_Use_It_.htm +++ /dev/null @@ -1,64 +0,0 @@ - - - - - - - - - - -Why Use It? - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Why Use It?

- -

A developer could use another integrated development environment (IDE) - such as Visual C++ to create text analyzers, but developers can rapidly - create accurate applications using VisualText. This is because all the - machinery needed to write a non-trivial text analyzer is included in VisualText, - while a developer would need to write that machinery first in another - programming environment. This machinery includes, for example, rule syntax - specification and interpreter, a parse tree builder, an integrated knowledge - base, and much more.

- -

In the past, text analyzers could be maintained only by the original - developers or by programmers with specialization in natural language processing - (NLP) or by computational linguistics. VisualText changes all that, allowing - non-programming staff to update text analyzers by adding text examples - with simple point-and-click commands.

- -

For further information, see http://www.textanalysis.com. -

- -

 

- - - - diff --git a/Help/helps/LJ.htm b/Help/helps/LJ.htm deleted file mode 100644 index bce8d18..0000000 --- a/Help/helps/LJ.htm +++ /dev/null @@ -1,87 +0,0 @@ - - - - - - - - - - -LJ - - - - - - - - - - - - - - - - - - - - - - - - - - - -

LJ

- -

Purpose

- -

Left-justify number in a given field size.

- -

Syntax

- -

LJ(num,fieldsize_num)

- -

Returns

- -

Remarks

- -

Example

- -

@CODE

- -

 

- -

@@CODE

- -

See Also

- -

rightjustifynum, - percentstr, Formatting - and I/O Functions

- - - - diff --git a/Help/helps/Log_Window.htm b/Help/helps/Log_Window.htm deleted file mode 100644 index 7904056..0000000 --- a/Help/helps/Log_Window.htm +++ /dev/null @@ -1,82 +0,0 @@ - - - - - - - - - - -Log Window - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Log Window

- -

The Log Window displays basic information about an analyzer when an - analyzer is opened as well as information about passes as they are processed - and applied. It can be a valuable tool when debugging your analyzer.

- -

- -

Types of Messages

- -

Log Window messages are divided into Informational - Messages and Error - Messages. Both types of messages provide a brief description of the - activity in the "Event" column. A more detailed description - of the activity is given under "Details". When an Error (or - warning) Message is displayed, the pass file number and the line containing - the error in the pass file is provided in the Details column.

- -

One feature of the Log Window is the ability to open a pass file that - has been flagged by an Error (or warning) Message. You do this by simply - clicking on the Error line in the Log Window. The pass file is opened - in the Workspace and the cursor is placed at the exact spot in the file - with the error.

- -

The Log Window display can be cleared by right-clicking in the Log Window - and selecting Clear.

- -

For more on messages displayed in the Log Window, see Log - Window Messages.

- -

 

- -

 

- - - - diff --git a/Help/helps/MULTI.htm b/Help/helps/MULTI.htm deleted file mode 100644 index b6ed13f..0000000 --- a/Help/helps/MULTI.htm +++ /dev/null @@ -1,120 +0,0 @@ - - - - - - - - - - -@MULTI Selector - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

MULTI Selector

- -

The MULTI Selector lets you specify a list - of node names to be searched for independently in the parse tree. The - entire subtree under such a context node will be traversed to look for - rule matches. (The MULTI Selector currently does not check the unsealed - flag.)

- -

It may be useful to know that the MULTI algorithm proceeds from the - most deeply buried phrases (i.e., those at the leaves of the parse tree) - to the phrase immediately below the context node.

- -

Using the MULTI Selector

- -

Below is a cautionary note on using the MULTI Selector.

- -

Common Error: MULTI with [Singlet] or [Tree] - Modifier

- -

MULTI in combination with either of the element modifiers [s] (i.e., - singlet) or [tree] can lead to unexpected behavior, such as matching the - same pattern multiple times. Normally, you will want to avoid using these - rule element modifiers when specifying a MULTI traversal of a parse subtree. - The problem surfaces in rules of the form

- -

_X <- _Y [s] @@

- -

that is, where the right-hand-side consists of a phrase of one element. - If _Y matches a buried node, then it will match the same node every time - an ancestor of that node is traversed. E.g., given the subtree

- -

_CONTEXT_NODE
-  |
-...._A ....
-  |
-  _B
-  |
-  _C
-  |
-  _Y

- -

then _Y, _C, _B, and _A will all match the above rule when MULTI is - the selector, yielding a (presumably) undesirable subtree such as

- -

_CONTEXT_NODE
-  |
-...._X ....
-  |
-  _A
-  |
-  _X
-  |
-  _B
-  |
-  _X
-  |
-  _C
-  |
-  _X
-  |
-  _Y

- -

See Also

- -

Context - Zone

- -

SELECT - Region

- -

PATH - Selector

- -

NODES - Selector

- -

 

- - - - diff --git a/Help/helps/Misc/ScreenLayout.gif b/Help/helps/Misc/ScreenLayout.gif deleted file mode 100644 index 0a73baf..0000000 Binary files a/Help/helps/Misc/ScreenLayout.gif and /dev/null differ diff --git a/Help/helps/Misc/StatusBar.gif b/Help/helps/Misc/StatusBar.gif deleted file mode 100644 index 9b36251..0000000 Binary files a/Help/helps/Misc/StatusBar.gif and /dev/null differ diff --git a/Help/helps/N.htm b/Help/helps/N.htm deleted file mode 100644 index e22e61e..0000000 --- a/Help/helps/N.htm +++ /dev/null @@ -1,46 +0,0 @@ - - - - - - - - - - -N - - - - - - - - - - - - - - - - - - - - - - - - - - - -

N

- -

Reference to a parse tree node or its variable.  See - Variables.

- - - - diff --git a/Help/helps/NLP_PP_Stuff/AT-CHECK_Actions.htm b/Help/helps/NLP_PP_Stuff/AT-CHECK_Actions.htm deleted file mode 100644 index 7348a06..0000000 --- a/Help/helps/NLP_PP_Stuff/AT-CHECK_Actions.htm +++ /dev/null @@ -1,72 +0,0 @@ - - - - - - - - - - -@CHECK Actions - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

CHECK Action

- -

CHECK Actions operate after - the rule's right-hand-side phrase has matched.

- -

The or action is the only specialized - function built for this region. General NLP++ code serves to implement - most CHECK Actions that you would want to perform.

- -

Two special NLP++ functions operate in the - CHECK Region. When a rule match is to be aborted, the function fail - is called to accomplish this. To succeed without performing further checks, - call the succeed function. Both - of these can occur in other regions, but may operate differently. If succeed is executed, then element actions - and POST Actions are performed next.

- -

See Also

- -

succeed

- -

CODE - Action

- -

PRE - Action

- -

POST - Action

- - - - diff --git a/Help/helps/NLP_PP_Stuff/AT-CODE_Actions.htm b/Help/helps/NLP_PP_Stuff/AT-CODE_Actions.htm deleted file mode 100644 index 24a2255..0000000 --- a/Help/helps/NLP_PP_Stuff/AT-CODE_Actions.htm +++ /dev/null @@ -1,468 +0,0 @@ - - - - - - - - - - -@CODE Actions - - - - - - - - - - - - - - - - - - - - - - - - - - - -

CODE Action

- -

The CODE Actions in this section predate the - NLP++ language, but are retained because they are still useful. In subsequent - releases, these actions may be modified.

- -

Types of CODE Actions

- -

CODE Actions are divided into two groups:

- - - -

Tables of CODE Actions

- -

Actions available in the CODE Region are provided in the tables below. -  For examples - on using these actions, refer to the individual action topics in the @CODE - Actions section.

- -

Variable Actions

- - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

ACTION(ARG1,ARG2...)

 -

DESCRIPTION

-

varstrs(varname)

 -

Create an empty multi-string-valued global variable with name varname. - The POST Action addstrs() adds - values to this type of variable.

-

sortvals(varname)

 -

Sort the strings in multi-string-valued global variable varname.

-

gtolower(varname)

 -

Convert the strings in multi-string valued - global variable to lower case.

-

guniq(varname)

 -

Remove redundancies in a sorted, multi-string - valued global variable.

-

lookup(var,file,flag)

 -

Specialized word lookup. Global - variable var - has multiple words as values. file - is a file of strings, one per line. flag - tells which bit-flag of the word's symbol table entry to modify. E.g., - lookup("Words", "dict.words", - "word") looks up all the values in the Words - variable in the dict.words file - and modifies the word bit-flag - (which says whether the word is a proper English word).

- -

Print Actions

- - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

CODE Action

 -

Description

-

print(str)

 -

Print str to the standard output (usually - the console or DOS window). (See startout - below).

-

printvar(var)

 -

Print the values of the global - variable named var to standard - output. (See startout - below).

-

fprintvar(file, - var)

 -

Print the values of the global - variable named var to file.

-

prlit(file, - str)

 -

Print the literal string str - to file.

-

gdump(filename)

 -

Dump the all the global variables and their - values to the given filename.

-

fileout(file)

 -

Open file - for appending. File becomes a - variable useable in print actions with a file argument, e.g. prlit().

-

startout(0)

 -

Divert standard output to the main output - file, set by the caller of the analyzer (default is output.txt).

-

stopout(0)

 -

Stop diverting standard output to the main - output file. Subsequent file-less output is to standard output.

- -

 

- -

See Also

- -

PRE - Action

- -

CHECK - Action

- -

POST - Action

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/AT-POST_Actions.htm b/Help/helps/NLP_PP_Stuff/AT-POST_Actions.htm deleted file mode 100644 index 05e5b86..0000000 --- a/Help/helps/NLP_PP_Stuff/AT-POST_Actions.htm +++ /dev/null @@ -1,1055 +0,0 @@ - - - - - - - - - - -@POST Actions - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

POST Action

- -

NLP++ code in the POST Region executes after - the rule matcher has successfully matched a rule and the match is accepted. -

- -

Types of POST Actions

- -

POST Actions are divided into four groups:

- - - -

One suite of POST Actions for the RFA (Rule-File Analyzer) is not described, - as they are not relevant to the VisualText user. Their names typically - begin with the prefixes rfa- or - rfb-.

- -

Tables of POST Actions

- -

Actions available in the POST Region are provided - in the table below.  For - examples on using these actions, refer to the individual action topics - in the included in the @POST Actions sections.

- -

Node Actions

- - - - - - - - - - - - - - - - - - - - - - - -
-

POST Action

 -

Description

-

pncopyvars(aPnode)

-

pncopyvars(num)

-

pncopyvars(0)

 -

Copy a node's variables to the suggested node - of a rule match. Must be called from the POST Region and can be considered - a new-style NLP++ action.  See - pncopyvars.

-

setbase(num, bool)

 -

Set the BASE attribute of the numth - node to "true" or "false".

-

setunsealed(num, bool)

 -

Set the UNSEALED attribute of the numth - node to "true" or "false".

-

xrename(name [, num])

 -

Rename the numth - context node to name. - If num is absent or 0, rename - last context node.

- -

Reduce Actions

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

POST Action

 -

Description

-

setlookahead(num1)

 -

Set the node that rule matcher will continue - with, after the current rule match is done. New - in 1.7.5.2

-

single()

 -

Single-tier reduce. Reduce the entire set - of nodes that matched a rule phrase.

-

singler(num1,num2)

 -

Single-tier reduce of a range of rule elements. - eg, if finding a period to be end-of-sentence in context, we only want - to reduce the period to end-of-sentence, not the whole context.

-

singlex(num1,num2)

 -

Single-tier reduce of a range of rule elements. - All nodes not in the range are EXCISED. eg, if matching a keyword html - tag, we only want to reduce the keywords and to zap the rest of the tag.

-

singlezap()

 -

Single-tier reduce. Excise all the nodes in - the matched phrase.

-

merge()

 -

Single-tier reduce that DISSOLVES each top-level - node in the matched phrase.

-

merger(num1,num2)

 -

Single-tier reduce that DISSOLVES each top-level - node in the matched range.

-

listadd(olist,oitem - [, keep])

 -

Add a new node to a list node's children. - If the item occurs after the list (olist - < oitem), - it is added as the last child. If the item occurs before the list, it - is added as the first child. The optional keep arg can be "true" - or "false". If "true", it keeps the nodes between - the list and the item as children of list. If "false", it excises - all the intervening nodes.

-

excise(num1,num2)

 -

Excise the nodes matching the range of elements - from the parse tree. For multiple excise actions, operations must be ordered - such that the highest number is listed first in the operation.

-

splice(num1,num2)

 -

DISSOLVE the top level nodes of given range.

-

group(num1,num2,label_str)

 -

Reduce the range of rule elements (num1,num2) - and name the group node label. - Unlike other reduce actions, this one can be repeated. WARNING: - The rule element numbers change after each group action, so subsequent - POST Actions must account for the new numbering of the rule elements. - Reimplemented as an NLP++ function so that args can be NLP++ expressions.

-

noop(0)

 -

Perform no POST Action. This is a convenience - to disable the default single() - reduce action.

- -

Print Actions

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

POST Action

 -

Description

-

print(str)

 -

Print the literal string str - to standard output.

-

printr(num1,num2)

 -

Print the text for the rule element range - from num1 to num2 - to standard output.

-

prchild(file, - num, name)

 -

Look for name - immediately under the node matching rule element num. - Print its text to file, if found.

-

prtree(file, - num, name)

 -

Look for name - anywhere under the node matching rule element num. - Print its text to file, if found.

-

prxtree(filename, prestr,
-  ord, name, poststr)

 -

To the filename, - print the first node called name - found in the ordth element's tree, - preceded by prestr and followed - by poststr. If named node is not - found, nothing is printed out. E.g., prxtree("out.txt", "date: - ", 3, "_date", "\n") prints out a line like "date: - 3/9/99 <cr>" if a _date node is found within the subtree of - elt 3.

-

prlit(file, - str)

 -

Print the literal string str - to file.

-

fprintnvar(file, - var, ord)

 -

To the file, - print the value of the variable var - in the ordth element's node.

-

fprintxvar(file, - var, ord)

 -

To the file, - print the value of the variable var - in the ordth context node.

-

fprintgvar(file, - var)

 -

To the file, - print the value of the global variable var.

-

gdump(file)

 -

Dump all the global variables and their values - to file.

-

xdump(file, - ord)

 -

Dump all the variables in the ordth - context node and their values to file.

-

ndump(file, - ord)

 -

Dump all the variables in the ordth - phrase element's node and their values to file.

-

sdump(file)

 -

Dump all the variables in the suggested node - and their values to file.

-

prrange(file, - num1,num2)

 -

Print the text under a range of rule elements - (num1,num2) to - file.

-

pranchor(file,num1,num2)

 -

Print a web URL to file. - Treat the range (num1,num2) as - a URL. Use the global variable named Base - to resolve and print complete relative URLs. (a prior pass must find the - <base> HTML tag and set Base - appropriately. NOT yet using the file's URL, so not handling cases where - <base> itself is a relative URL.)

- -

Variable Actions

- - - - - - - - - - - -
-

POST Action

 -

Description

-

addstrs(var,num)

 -

Add a string to the multi-string global - variable var. Fetch the string - from the nodes matching rule element num. - var should have been declared - using the CODE Action varstrs(). - (REMOVED. Use NLP++ array syntax - instead.)

- -

See Also

- -

CODE - Action

- -

PRE - Action

- -

CHECK - Action

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/AT-PRE_Actions.htm b/Help/helps/NLP_PP_Stuff/AT-PRE_Actions.htm deleted file mode 100644 index 087bcce..0000000 --- a/Help/helps/NLP_PP_Stuff/AT-PRE_Actions.htm +++ /dev/null @@ -1,493 +0,0 @@ - - - - - - - - - - -@PRE Actions - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

PRE Action

- -

PRE Actions operate after a - rule element has successfully matched. They provide further constraints - on the element match, and may result in the match being rejected if the - additional constraints are not met.  PRE Actions provide - the only means for specifying additional constraints on element matching. -

- -

PRE Action Syntax

- -

Unlike other actions, PRE Actions - have a mandatory range. Their form is

- -

<num,num> - pre_action_name(args....)

- -

The PRE Action applies to each rule - element in the <num, num> range individually.

- -

PRE Actions operate only on tokens, or leaf nodes of the parse tree, - at present. So, if the current node covers more than a single - leaf node, the PRE Action automatically fails. - Another thing to be careful of is that PRE Actions disregard the base attribute of a node. That is, even - if the current node or a subnode has the base - attribute set, the PRE Action will look under it for the leaf node.

- - - - - - -
-

Note: - The PRE Region does not currently accept general NLP++ functions, expressions, - and statements.

- -

Table of PRE Actions

- -

Actions available in the PRE Region - are provided in the table below.  For - examples on using these actions, refer to the individual action topics - in the @PRE Actions section.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

PRE Action

 -

Description

-

uppercase()

 -

Succeed if leaf token is all uppercase.

-

lowercase()

 -

Succeed if leaf token is all lowercase.

-

cap()

 -

Succeed if leaf token has capitalized first - letter.

-

length(num)

 -

Succeed if leaf token has length exactly equal - to num.

-

lengthr(num1,num2)

 -

Succeed if leaf token length is in the range - of lengths (num1,num2), - inclusive.

-

numrange(num1,num2)

 -

Succeed if leaf token is numeric in the given - range.

-

unknown()

 -

Succeed if leaf token is an unknown word. - Meaningful only if a prior pass has performed the CODE Action lookup().

-

regexp(str)

 -

Succeed if regular expression str matched. -  str may - include characters, with ? meaning a single char, * meaning 0 or more - chars.

-

regexpi(str)

 -

Like regexp, - but case insensitive.

-

var(str)

 -

Succeed if matched node has a variable called - str with nonzero value.

-

vareq(str,str_or_num)

 -

Succeed if matched node has a variable called - str whose value equals a string - or numeric value in str_or_num.

-

varne(str,str_or_num)

 -

Succeed if matched node has a variable called - str whose value does not equal - a string or numeric value in str_or_num.

-

varz(str)

 -

Succeed if matched node has a variable called - str with zero value (or lacks - the variable altogether).

-

debug()

 -

Succeed unconditionally. Used to place a C++ - breakpoint at a particular rule.

- -

 

- -

See Also

- -

CODE - Action

- -

CHECK - Action

- -

POST - Action

- - - - diff --git a/Help/helps/NLP_PP_Stuff/About_NLP++.htm b/Help/helps/NLP_PP_Stuff/About_NLP++.htm deleted file mode 100644 index 90c3ac7..0000000 --- a/Help/helps/NLP_PP_Stuff/About_NLP++.htm +++ /dev/null @@ -1,130 +0,0 @@ - - - - - - - - - - -About NLP++ - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

About NLP++

- -

NLP++ is a general purpose programming language that integrates extensions - for natural language processing (NLP). In addition to a C++ -like code - syntax, NLP++ embodies a syntax for pass files, contexts, rules, and actions. - Similar to LISP, NLP++ is a loosely-typed language.  In - addition to integer, string, and array data types, NLP++ includes parse - tree and knowledge base data types. In this way, the user can naturally - address the rules, parse trees, and the knowledge base, while using C++ - -like code to embed heuristics. Integration with the parse tree makes - it possible to manipulate nodes and to manage semantic information within - them. Integration with the knowledge base makes it possible to dynamically - create and manage complex data structures. Integration with rules makes - it possible to control every aspect of rule matching, to exploit context, - and to elaborate the actions that follow rule matching.

- -

NLP++ supports Concept Oriented Programming (COP), enabling you to focus - on heuristics, the domain, and the task, rather than the underlying data - structures. Because VisualText supports an interpreted environment, you - can modify and test without having to wait for code to compile, which - accelerates the development of text analyzers.

- -

NLP++ is an evolving language but is already rich and robust.

- -

Text Analyzers and NLP++

- -

As implemented in VisualText, a text analyzer consists of:

- - - -

NLP++ code and rules occur entirely within pass files. Everything contained - within a pass file is considered to be part of the NLP++ language, including - region markers (e.g., @RULES and @@RULES), specialized regions, code, - actions, rules, and comments.

- -

NLP++ Syntax

- -

NLP++ combines several types of syntax:

- - - -

One region of pass files, the PRE Region also - has an element range construct, e.g., "<1, 3>" that is - not found elsewhere. We strive to keep the syntax sets as integrated as - possible. For example, double quoted "abc" is a string in every - context, and C/C++ conventions for backslashed characters are used everywhere. - # is used for comments throughout, and a token starting with an underscore - (e.g., "_noun") denotes a nonliteral everywhere.

- -

See Also

- -

Code - Syntax

- -

Rule - Syntax

- -

Pass - File Layout

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/About_NLP++_Variables.htm b/Help/helps/NLP_PP_Stuff/About_NLP++_Variables.htm deleted file mode 100644 index 5e495d5..0000000 --- a/Help/helps/NLP_PP_Stuff/About_NLP++_Variables.htm +++ /dev/null @@ -1,398 +0,0 @@ - - - - - - - - - - -About NLP++ Variables - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Variables

- -

There are two broad classes of variables in NLP++:

- - - -

Global variables and - local variables are the only general class implemented to date. - Globals are referenced by the special G function.  Local - variables are denoted by L, - and currently specify parameter lists and appear within user-defined NLP++ - function definitions in the @DECL region.

- -

Parse tree node variables, or - node variables for short, are - attached to particular nodes in the parse tree, and serve to decorate - the tree with semantic information. The special functions N, X, and S refer to parse tree nodes in the context - of a matched rule.

- -

The special NLP++ functions that reference variables are tabulated below:

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

VARIABLE

 -

DESCRIPTION

-

G(varname)

 -

Global variable.

-

S(varname)

 -

Variable belonging to the suggested - concept of a rule.

-

X(varname - [,num])

 -

Variable belonging to the numth - context node starting at the root - of the parse tree. Usually used to refer to the numth - node of the @PATH select list. With @NODES, the preferred form is X(varname).

-

N(varname [,num])

 -

Variable belonging to - a node that matched the numth element of - a rule phrase.

-

L(varname])

 -

Local variable, specifies user-defined NLP++ - function parameters and appears in function body.

-

 

 -

 

- -

The shorthand notations N(varname) - and X(varname) refer to the last - element of the phrase and the last context node, respectively. The former - is convenient for rules that have only one element in their phrase. The - latter is convenient when using the @NODES - selector. N(varname, 0) and X(varname, 0) have the identical meaning.

- -

There is no notion of a local variable or pointer, at this time. S variables are often used as an idiom - for local variables (the difference is that they "hang around" - if a new node is suggested by the current rule).

- - - - - - -
-

Note: - Variable names must be quoted. E.g., N("unknown word", 2)

- -

WARNING: X(varname,num) does NOT refer to the numth member of a @NODES list. That list is - not a path in the parse tree. We recommend using X(varname) with @NODES, to avoid "blindly" - traversing down the parse tree. Using X(varname,2), for example, implies - that you know what the nodes immediately below the root are. Such a use - should be clearly documented.

- -

Special Variable Names

- -

To get useful information from the parse tree and other sources when - analyzing a text, some special variable names have been provided. All - start with a dollar sign ($). - No such variable names are usable with suggested (S) - variables, at present.

- -

Note the following about using special variables:

- -
    - -

  1. Special variable names must be quoted. E.g., N("$text", - 2).

    2. - -

  2. You should not assign to or create your own "dollar" - variables. NLP++ reserves all variable names that start with a dollar - sign.

    3. - -

  3. The $uppercase, $lowercase, $cap, and $unknown - differ from the analogous PRE Actions in that they all traverse down to - the "leafiest" node they can find, in order to perform their - check. The corresponding PRE Actions only chain down if the [s] flag is - present in the associated phrase element(s). Also, the $ variables are - not stopped by a [base] attribute in a node.

    4. -
- -

See - Special NLP++ Variables for a listing of the special variables.

- -

Special Node References

- -

N, - X, - and S - can, in addition to referring to a node's variables, refer to the node - itself. Thus N(2) - refers to the first node that matched element 2 of the matched rule. N() and N(0) refer - to the last rule element's (first or only) node. Similarly, X(2) refers - to the second context node in the @PATH dominating the current rule. S() refers - to the suggested concept.

- -

Variable Types

- -

NLP++ is a loosely typed language. -  The same - variable can be assigned an integer in one statement and a string in the - next.  Type - is determined by usage -- there are no type declarations for variables.

- -

G("tmp") = 123;

- -

G("tmp") = "abc";

- -

 

- -

Variables can take multiple - values, in which case they are called arrays. -  Each value - in an array can be of an arbitrary data type:

- -

 

- -

G("tmp")[0] = 123;

- -

G("tmp")[1] = "abc";

- -

 

- -

See the Data - Types topic for more details about arrays and other data types.

- -

See Also

- -

Code - Syntax

- -

Tokens

- -

Data - Types

- -

Special - Variables

- -

Functions - and Actions

- -

Operators - and Expressions

- -

Statements - and Blocks

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/About_Pass_Files.htm b/Help/helps/NLP_PP_Stuff/About_Pass_Files.htm deleted file mode 100644 index ff40188..0000000 --- a/Help/helps/NLP_PP_Stuff/About_Pass_Files.htm +++ /dev/null @@ -1,250 +0,0 @@ - - - - - - - - - - -About Pass Files - - - - - - - - - - - - - - - - - - - - - - - - - - - -

About Pass Files

- -

A pass file holds the NLP++ code and rules associated with a Pattern - or Recursive pass type. Five zones divide a pass file and are ordered - as follows:

- - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

ZONE

 -

DESCRIPTION

-

Declare

 -

Contains user-defined NLP++ function definitions.

-

Code

 -

Contains NLP++ code that is independent of - rule matching.

-

Context

 -

Selects nodes of the parse tree. Rules will - match only against the phrases immediately under the selected nodes.

-

Minipass

 -

Contains nested minipasses called Recurse - Regions. Each Recurse Region is named and contains a Grammar Region. Rule - elements in the Grammar Zone can invoke these named rule sets.

-

Grammar

 -

Contains the main Grammar Region of the current - pass file.

- -

Optional Zones

- -

A minimal pass file may consist solely of a Code Zone or a Grammar Zone.

- -

If absent, the Context Zone defaults to the selector

- -

  @NODES _ROOT

- -

which traverses the phrase immediately below the root of the parse tree, - invoking the rule matcher on each node of the phrase.

- -

Pass File Markers

- -

The '@' sign is a reserved character that indicates the start of a region - (e.g., @CODE) or a selector (e.g., @NODES).

- -

To indicate the end of a region, the name of the region preceded by - two '@' signs is used. The end of a Code Region is indicated by @@CODE. - End markers are optional in most cases.

- -

A lone @@ marker indicates the end of a rule, e.g.,

- -

  _NP <- _NOUN @@

- -

Case Sensitivity of Pass File Markers

- -

Pass file markers are case-insensitive.  A - marker such as @PRE can be written as @pre.  

- - - - - - -
-

Note: Unlike pass file markers, NLP++ functions - and actions are case-sensitive.  

- -

 

- -

See Also

- -

Code Zone
-Context Zone
-Minipass Zone
-Grammar Zone

- -

 

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Actions.htm b/Help/helps/NLP_PP_Stuff/Actions.htm deleted file mode 100644 index 1941de3..0000000 --- a/Help/helps/NLP_PP_Stuff/Actions.htm +++ /dev/null @@ -1,99 +0,0 @@ - - - - - - - - - - -Actions - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Actions

- -

Actions are NLP++ functions that are specialized to know about particular - contexts. For example, the POST Action single - assumes that a rule has successfully matched.

- -

Historically, actions preceded NLP++, and many do not allow full NLP++ - expressions as arguments. The CODE, PRE, CHECK, and POST regions all have - associated actions.

- -

Many actions, such as the PRINT actions, are superseded by the C++ -like - << operator and are retained for backward compatibility.

- - - - - - -
-

Note: - Some actions are slated for review.

- -

 

- -

See Also

- -

CODE - Action

- -

PRE - Action

- -

CHECK - Action

- -

POST - Action

- -

Functions -

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Code_Syntax.htm b/Help/helps/NLP_PP_Stuff/Code_Syntax.htm deleted file mode 100644 index 41d4989..0000000 --- a/Help/helps/NLP_PP_Stuff/Code_Syntax.htm +++ /dev/null @@ -1,82 +0,0 @@ - - - - - - - - - - -Code Syntax - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Code Syntax

- -

NLP++ code syntax is similar to C - or C++ code. Further topics within the Code Syntax section elaborate the - tokens, data types, variables, operators, functions, actions, expressions, - statements, arrays, and other constructs of the language.

- -

An interesting feature of NLP++ is that variables are named by strings - and not by identifiers. You can primarily refer to variables with the - G(), X(), - N(), and S() - special functions, which refer to a global, context, node, and suggested - node variable respectively.  Some - functions take variable names as arguments, e.g., pnvar.

- -

NLP++ data types include integer, float, string, - parse tree, and knowledge base types.

- -

See Also

- -

Tokens

- -

Variables

- -

Data - Types

- -

Special - Variables

- -

Functions - and Actions

- -

Operators - and Expressions

- -

Statements - and Blocks

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Code_Zone.htm b/Help/helps/NLP_PP_Stuff/Code_Zone.htm deleted file mode 100644 index e3b1e15..0000000 --- a/Help/helps/NLP_PP_Stuff/Code_Zone.htm +++ /dev/null @@ -1,65 +0,0 @@ - - - - - - - - - - -Code Zone - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Code Zone

- -

The Code Zone is reserved for regions of NLP++ code that are independent - of rule matching.

- -

Code Region

- -

Currently, the only region implemented within the Code Zone is the CODE - Region, which is delimited by @CODE and (optionally) by @@CODE, and operates - before any rule matching is executed - for the current pass file.

- -

A pass file may consist entirely of a Code Region.

- -

See Also

- -

CODE - Region

- -

CODE - Action

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Context_Zone.htm b/Help/helps/NLP_PP_Stuff/Context_Zone.htm deleted file mode 100644 index 6250c2d..0000000 --- a/Help/helps/NLP_PP_Stuff/Context_Zone.htm +++ /dev/null @@ -1,91 +0,0 @@ - - - - - - - - - - -Context Zone - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Context Zone

- -

The Context Zone of a pass file specifies methods for selecting - nodes of the parse tree. Each selected - node, or context node, - serves as a locus for rule matching. The @SELECT and @@SELECT markers - optionally delimit the Context Zone.

- -

Order in Pass Files

- -

The Context Zone can come either before or after the Code Zone.  The - Context Zone is also optional.  If - it is absent, it is equivalent to

- -

  @NODES _ROOT

- -

That is, rule matching will occur in the phrase of nodes immediately - under the root node of the parse tree.

- -

Rule Matching and Selectors

- -

The mode of rule matching depends on the selector (e.g., @NODES above) - and the pass algorithm type (e.g., Pattern or Recursive) for the current - pass. For example, with the selectors NODE or PATH, the phrase immediately - under the selected node is subjected to rule matching. With the MULTI - selector, every phrase in the subtree of the selected node is subjected - to rule matching.

- -

Specifying Selectors

- -

Within the Context Zone, a single SELECT Region may be specified. Currently, - a pass file may include at most one selector - (i.e., @NODES, @PATH, and @MULTI).

- -

See Also

- -

SELECT - Region

- -

NODES - Selector

- -

PATH - Selector

- -

MULTI - Selector

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Functions.htm b/Help/helps/NLP_PP_Stuff/Functions.htm deleted file mode 100644 index c854dc2..0000000 --- a/Help/helps/NLP_PP_Stuff/Functions.htm +++ /dev/null @@ -1,124 +0,0 @@ - - - - - - - - - - -Functions - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Functions

- -

NLP++ provides a variety of predefined functions for managing the parse - tree and knowledge base. Convenient functions also support output formatting, - string handling, and spelling. The spell functions can be used to help - with unknown word lookup by a text analyzer.

- -

As NLP++ evolved from an "action language" for rules, the - Actions section provides functions that expect to apply before, during, - and after the matching of a rule.  They - are "rule-aware," in that they can address the matched nodes, - the suggested concept, and the "dominating" or context nodes - within which the rule match occurred.  The - actions are deprecated in the sense that some of them cannot accept full - NLP++ expressions as function arguments, but require literal arguments. -  Some actions, - such as group, have been upgraded to full-fledged NLP++ functions and - are listed in both the Actions and Functions sections.

- - - - - - -
-

Note: - Boolean type is currently implemented by integer 1 - (or non-zero) for true and 0 for false. - Floating point type is newly (but incompletely) available.  Char - type is not implemented.

- -

See Also

- -

Database - Functions

- -

Special - Functions

- -

Knowledge - Base Functions

- -

Formatting - and I/O Functions

- -

Math Functions

- -

Parse - Tree Functions

- -

String - Functions

- -

Spelling - Functions

- -

Web - Functions

- -

 

- -

Actions

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Functions_and_Actions.htm b/Help/helps/NLP_PP_Stuff/Functions_and_Actions.htm deleted file mode 100644 index 35451ab..0000000 --- a/Help/helps/NLP_PP_Stuff/Functions_and_Actions.htm +++ /dev/null @@ -1,229 +0,0 @@ - - - - - - - - - - -Functions and Actions - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Functions and Actions

- -

Functions

- -

NLP++ supports function calls, which have a similar form to other programming - languages (such as C/C++):

- -

    functionName(arg1, - .... argN)

- -

where functionName is constrained to be alphabetic and where each argument - to the function may be an NLP++ expression.

- -

Only call-by-value arguments are supported at present.  That - means that when a variable is used as an argument, its value is not changed - by a change to the local variable that it becomes associated with in the - function.

- -

Functions have no template or declaration -- only a definition.  Even - so, you may use a function in one pass and define it in a subsequent pass - (forward reference).  The - parameters to functions take arbitrary data types that may vary dynamically. -  Return - statements similarly return arbitrary data types.  The - same function could return a numeric value in one call and a string in - another.  We - discourage the abuse of this extreme flexibility, however!  The - new L variable - reference has been introduced to support formal function parameters as - well as local variables within a function body.  Recursive - functions are also possible.  The - example below illustrates the definition and use of user-defined functions - in NLP++.

- -

@DECL

- -

xuf(L("var"))

- -

{

- -

"output.txt" - << L("var") << "\n";

- -

L("local") = 3;   # - create a local variable.

- -

"output.txt" - << L("local") << "\n";

- -

return - 1+2;

- -

}

- -

 

- -

facto(L("num"))  # - recursive factorial function.

- -

{

- -

if - (L("num") <= 0)

- -

return - 1;

- -

return - L("num") * facto(L("num") - 1);

- -

}

- -

@@DECL

- -

 

- -

@CODE

- -

G("x")[2] = 3;

- -

"output.txt" - << G("x") << "\n";

- -

xuf(G("x"));

- -

"output.txt" - << "4! = " << - facto(4) << "\n";

- -

@@CODE

- -

 

- -

The code above prints - out

- -

0 0 3

- -

0 0 3

- -

3

- -

4! = 24

- -

Actions

- -

Actions, e.g., single, are specialized - functions that don't take generalized NLP++ expressions as arguments. -  Though - some actions have been upgraded and now take NLP++ arguments (e.g. group), some actions have not been upgraded - and the argument must be a literal value.

- -

A unique feature of actions in the PRE Region is that they have a mandatory - range.  The - syntax is as follows:

- -

<num,num> - pre_action_name(args....)

- -

The action applies to each rule element in the <num, num> range - individually.

- -

See Actions as a starting point for per-page - descriptions for actions.

- -

Case Sensitivity

- -

NLP++ function and action names are case-sensitive.  

- -

Defining New Functions

- -

In VisualText 1.2 and later, user-defined NLP++ functions are fully - supported, including parameter lists, return statements, recursive calls, - and compilation of these.  Functions - are specified in the @DECL region, which must appear first within a pass - file, if present.

- -

New NLP++ functions can also be defined with C++ code, within an application's user - project. User-defined function calls must be preceded by the 'user' - scope qualifier, e.g., user::myfunction. - (Interpreted NLP++ will find the correct function and issue a warning, - but user functions without the scope qualifier will be compiled incorrectly.)

- -

See Also

- -

Actions

- -

Functions

- -

 

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Grammar_Region.htm b/Help/helps/NLP_PP_Stuff/Grammar_Region.htm deleted file mode 100644 index 91511c0..0000000 --- a/Help/helps/NLP_PP_Stuff/Grammar_Region.htm +++ /dev/null @@ -1,210 +0,0 @@ - - - - - - - - - - -Grammar Region - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Grammar - Region

- -

A Grammar Region is where you write NLP++ rules and actions.

- -

Region Marker and Action Regions

- -

The Grammar Region consists of one or more @RULES markers with associated - rules and optionally with associated action regions.  These - action regions are the PRE Region, the CHECK Region, and the POST Region. -  (See below.)

- -

Grammar Zone in RECURSE Region

- -

A RECURSE Region and the main Grammar Zone are the locations in which - a Grammar Region may appear.

- -

Grammar Region "Subregions"

- -

The following "subregions" may appear within a Grammar Region, - anchored by the RULES Region:

- - - - - - - - - - - - - - - - - - - - - - - -
-

REGION

 -

DESCRIPTION

-

@PRE

 -

Actions that represent additional conditions - on the matching of individual rule elements. NOTE: Does not allow general - NLP++ code.

-

@CHECK

 -

When a rule has matched, NLP++ code in this - region checks for self-consistency and/or builds semantic data. A rule - match may still be rejected from this region.

-

@POST

 -

NLP++ code that operates once a rule match - has been accepted.

-

@RULES

 -

Contains the actual NLP++ rules.

- -

The end markers @@PRE, @@CHECK, @@POST, and @@RULES are optional.

- -

See Also

- -

PRE - Region

- -

PRE - Action

- -

CHECK - Region

- -

CHECK - Action

- -

POST - Region

- -

POST - Action

- -

Grammar - Zone

- -

 

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Grammar_Zone.htm b/Help/helps/NLP_PP_Stuff/Grammar_Zone.htm deleted file mode 100644 index 44cb487..0000000 --- a/Help/helps/NLP_PP_Stuff/Grammar_Zone.htm +++ /dev/null @@ -1,121 +0,0 @@ - - - - - - - - - - -Grammar Zone - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Grammar Zone

- -

The Grammar Zone is where you place the main rules of a pass file.

- -

The Grammar Zone occurs as the last zone (or possibly the only zone) - of a pass file.  Unlike - other zones, the Grammar Zone does not have a distinguishing marker.  

- -

Grammar Region

- -

The Grammar Zone contains zero or more Grammar Regions.  The - Grammar Region is where rules and actions may be written.

- -

Invoking Minipasses

- -

If a pass file contains RECURSE Regions, then at least one rule in the - Grammar Zone must invoke a RECURSE Region to "start the ball rolling", - or no minipasses will ever be executed.  See - the phrase element modifier, recurse.)

- -

Common Error:  No - Grammar Zone with RECURSE Region

- -

Since a Grammar Zone has no beginning marker, a problem can occur if - a pass file looks like:

- -

@RECURSE abc

- -

@RULES

- -

...

- -

#Intended start of Grammar - Zone.

- -

@RULES

- -

In the above example, the last RULES Region is understood as part of - the previous RECURSE Region, rather than as the start of the Grammar Zone. -  To obtain - the right reading, the RECURSE Region terminator should be used.

- -

@RECURSE abc

- -

@RULES

- -

...

- -

@@RECURSE

- -

 

- -

#Intended start of Grammar - Zone.

- -

@RULES

- -

See Also

- -

About - Pass Files

- -

Grammar - Region

- -

PRE - Region

- -

CHECK - Region

- -

POST - Region

- -

RULES - Region

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Minipass_Zone.htm b/Help/helps/NLP_PP_Stuff/Minipass_Zone.htm deleted file mode 100644 index 76008c0..0000000 --- a/Help/helps/NLP_PP_Stuff/Minipass_Zone.htm +++ /dev/null @@ -1,66 +0,0 @@ - - - - - - - - - - -Minipass Zone - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Minipass Zone

- -

The Minipass Zone is used to write zero or more nested "minipasses" - within a pass file. Each minipass is called a RECURSE Region and has an - associated name. The Grammar Region within a RECURSE Region is not executed - directly, but rather it is invoked by a recurse - element action when a rule elsewhere in the pass file has matched.

- -

The idea is that a rule element, such as a wildcard, "roughs out" - a segment or phrase of nodes by matching or collecting these nodes. Then - that rule element invokes a minipass by name in order to characterize - or refine the matched phrase.

- -

Typically, minipasses are invoked from rules in the main Grammar Zone - of the pass file. But matches in one RECURSE Region may lead to invoking - a second RECURSE Region.

- -

See Also

- -

About - Pass Files

- -

RECURSE - Region

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Operators_and_Expressions.htm b/Help/helps/NLP_PP_Stuff/Operators_and_Expressions.htm deleted file mode 100644 index 1d6b4c6..0000000 --- a/Help/helps/NLP_PP_Stuff/Operators_and_Expressions.htm +++ /dev/null @@ -1,552 +0,0 @@ - - - - - - - - - - -Operators and Expressions - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Operators and Expressions

- -

Operators

- -

Most NLP++ operators are modeled on the C programming language. The - main addition is the confidence - operator, %%. This enables convenient accumulation of evidence - in terms of percent confidence. Following is an operator precedence table, - with highest precedence at the top.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

OPERATOR

 -

DESCRIPTION

 -

ASSOCIATIVITY

-

++ - --

 -

Post increment, decrement

 -

Left to right

-

++ --

 -

Pre increment, decrement

 -

Right to left

-

!

 -

Logical NOT (unary)

 -

Right to left

-

+ -

 -

Unary plus, minus

 -

Right to left

-

% * / %%

 -

Remainder, multiplication, division, confidence

 -

Left to right

-

+ -

 -

Addition, subtraction

 -

Left to right

-

 < > <= - == != >=

 -

Relational operators

 -

Left to right

-

&& ||

 -

Logical AND, OR

 -

Left to right

-

=

 -

Assignment

 -

Right to left (multiple assignment works)

-

*= /= += -= %%=

 -

Shorthand assignment (unimplemented)

 -

Right to left

-

<<

 -

Output operator

 -

Left to right

- -

Output Operator

- -

The << output operator is modeled - on the C++ operator. One form is:

- -

filename - << expression << expression - .... ;

- -

Example:

- -

"output.txt" << "Hello, NLP++!" << "\n";

- -

NOTES: - The file named by filename should have been created by a call to openfile(filename...) or a prior fileout(filename) code action. If that - hasn't been done, then the output operator will create the global variable - G(filename) with the output stream - as its value. The default output mode for the file is append, which can't - be modified by the user at present. Also, the default output directory - is the log directory for the current input file -- also not under user - control at present.

- -

Confidence Operator

- -

Say that you have 80% confidence - that you've found a name. You could write this into a global variable - as

- -

G("name confidence") = 80;

- -

Now - you want to account for a new piece of evidence. Say the standalone confidence - of that evidence is 70%. You could combine this as

- -

G("name confidence") = G("name confidence") %% 70;

- -

which - will yield a new percent confidence between 80% and 100%.

- -

The %% operator guarantees that "adding" - or "subtracting" confidence will always stay within the range - of 0 to +100, providing an intuitive way to gather evidence and compare - confidences using a single scale of measurement.

- -

String Catenation Operator

- -

The plus '+' operator, when given string args, serves as a built-in - string catenation operator.

- -

Expressions

- -

NLP++ expressions are modeled on the C/C++ programming languages and - can be nested arbitrarily and grouped using parentheses ( - ).

- -

Assignment expressions can be nested, just as in C/C++. For example

- -

G("a") = G("b") = 0;

- -

sets the global variables named "a" and "b" to 0.

- -

See Also

- -

Code - Syntax

- -

Tokens

- -

Variables

- -

Data - Types

- -

Special - Variables

- -

Functions - and Actions

- -

Statements - and Blocks

- -

 

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Pass_Files.htm b/Help/helps/NLP_PP_Stuff/Pass_Files.htm deleted file mode 100644 index 0c4b960..0000000 --- a/Help/helps/NLP_PP_Stuff/Pass_Files.htm +++ /dev/null @@ -1,3470 +0,0 @@ - - - - - - - - - - -Pass Files - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Pass Files

- -

A pass file (also called a rule file), consists of NLP++ code and rules. - A pass file represents a single pass of a text analyzer over a parse tree. - There are two different algorithms used to traverse the parse tree: PAT - (for pattern) and REC (for recursive).

- -

The PAT algorithm uses rules in a pattern-based way, traversing each - selected phrase in the parse tree only once, while the REC algorithm treats - the rules as a recursive grammar, traversing each selected phrase repeatedly - till no rules match within that phrase.

- -

Layout

- -

The pass file, which is used by a single analyzer PAT pass, consists - of sets of nonrepeating and repeating regions, which must be ordered in - a pass file as follows:

- - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

REGION

 -

REPEATABLE

 -

DESCRIPTION

-

Code

 -

NO

 -

Interpreted actions that precede the rule - matching process.

-

Context

 -

NO

 -

Specifies nodes of the parse tree. Rules will - match only against the phrases immediately under those nodes.

-

Minipass

 -

YES

 -

Recursive minipasses. Each recurse region - is named and can contain repeating rules regions. Rule elements in the - main RULES region (below) can invoke these named rule sets.

-

Grammar

 -

YES

 -

The main actions-rules regions for the current - file (i.e., the current pass).

- -

A pass file must contain at least a @CODE region or a main @RULES region.

- -

Code Region

- -

The CODE region consists of interpreted actions that operate before - the rule matching for the current file, but only once for the current - file. Each action looks like

- -

name(arg1,...argN)

- -

which is identical to the syntax for PRE, CHECK, and POST actions. (PRE - actions have an additional bracketed range at the start of each such action.) - See the CODE ACTIONS section below for details about available actions. - For a particular application, the developer can create new code actions - within the application's USER project.

- -

Select Region

- -

The Select Region allows the user to specify which nodes of the parse - tree to execute. The region is optional, and the default is to match the - rules of the file against the top-level phrase of the parse tree, that - is, the phrase directly under the _ROOT - node. The default could be made explicit, as in the following sample Select - Region:

- - - - - - - - - - - - - - - -
-

@SELECT

 -

# Start of Select Region

-

@NODES _ROOT

 -

# Specifies nodes whose underlying phrase - will be matched against.

-

@@SELECT

 -

# Optional end-marker for Select Region

- -

(Note that the region markers @SELECT and @@SELECT are also optional. - The sample Select Region above could be simply @NODES _ROOT.)

- -

If there is more than one - @NODES - marker, or if more than one name appears after a @NODES marker, those - names are all collected into a single list. The - pattern matcher makes a single pass - through the parse tree, looking for a node from this list. If such a node - is found, the phrase under it is matched against the rules of the file. - More than one @NODES is unimplemented.

- -

Instead of @NODES, one can specify the @MULTI - marker, which also takes a list of node names. This applies the - rules in the file to every - phrase in the subtree of the selected nodes, allowing multiple - matches in the subtree. Needless to say, the rule writer must be careful - in using rules with the singlet and tree keys in conjunction with the - MULTI specifier.

- - - - - - - - - - - - - - - - - - - -
-

SELECT TYPE

 -

DESCRIPTION

-

@NODES

 -

Directs the pattern matching algorithm (e.g., - PAT or REC) to nodes in the parse tree. The phrase immediately under such - a node will be traversed to look for rule matches. WARNING: The pattern matcher will only - find such nodes inside a path with the unsealed - attribute set to true in all nodes of the path.

-

@MULTI

 -

Directs the pattern matching algorithm (e.g., - PAT or REC) to nodes in the parse tree. The entire subtree under such - a node will be traversed to look for rule matches.

-

@PATH

 -

Directs the pattern matching algorithm (e.g., - PAT or REC) to a specific path of nodes in the parse tree. The phrase - immediately under the last node in the path list will be traversed to - look for rule matches. The list must always start with the node _ROOT. - The second name in the path must be a direct child of _ROOT, the third, - a child of that, and so on.

- -

If one SELECT node is inside the subtree of another SELECT node, then - the one inside will not be seen - by the pattern matcher. @MULTI in combination with [s] or singlet mode - can lead to unexpected behavior, such as matching the same pattern multiple - times. Normally, you will want to avoid using [s] in rule elements when - specifying a @MULTI traversal of a parse subtree.

- -

Recurse Region

- -

The Recurse Region allows the user to create several passes within a - single pass file. Each recurse region acts like an independent pass. But - a recurse pass can only be invoked from an element of a rule that has - matched. The phrase that the recurse pass operates on is the set of nodes - that matched the invoking rule element. Here's an example taken from the - RFA (Rules-File Analyzer), which parses the pass files themselves.

- -

# Start of recurse region. A name is required.

- -

@RECURSE actionpass

- -

# Start of one rules region within the recurse region.

- -

@RULES

- -

# The pass converts literals (e.g., "single") to actions in context.

- -

_ACTION <- _LIT @@

- -

# End of recurse region. (@@RULES marker is optional).

- -

@@RECURSE

- -

# ....

- -

# A main rules region.

- -

@RULES

- -

_POSTS [base] <- _soPOST

- -

_xWILD [matches=(_LIT _ACTION) recurse=(actionpass)]

- -

_eoPOST [opt] @@ # Simplified rule for @POST actions region.

- -

In the above example, when the RFA sees a plain word like "single" - in a file, it doesn't know whether to reduce it to an _ACTION or not. - But within the markers @POST and @@POST, the literal should be treated - as an action. This allows the parentheses in pass file actions to be optional, - when the action takes no arguments. E.g., "single()" and "single" - are equivalent.

- -

In general, RECURSE mini-passes enable - a mechanism for "roughing out" a text and then applying the - correct assignments when the context has been determined unambiguously. - They have many other uses, such as traversing nodes that matched a rule - element in order to perform cleanups, data gathering, deletions, and so - on.

- -

NOTE: A Recurse Region - cannot contain a Code Region or a Select Region. A Select Region would - be meaningless, since the recurse region operates on a phrase specified - by the rule element that invokes it. A code region could be implemented - if it proves useful.

- -

Rules Region

- -

The main Rules Regions (that follow the Code, Select, and Recurse Regions - in a pass file), as well as the Rules Regions nested inside a Recurse - Region, must be laid out as follows:

- - - - - - - - - - - - - - - - - - - - - - - -
-

SUBREGION

 -

DESCRIPTION

-

PRE

 -

Additional conditions on the matching of individual - rule elements.

-

CHECK

 -

When a rule has matched, these actions check - self-consistency and/or build semantic data. These actions can be used - to reject a rule match.

-

POST

 -

Actions that operate once a rule match has - been accepted.

-

RULES

 -

Subregion containing the actual rules that - are subject to immediately preceding Pre,Check, and Post subregions.
-If a Rules subregion has no preceding action subregions, then it resets - its actions to the default (single reduce).

- -

Each region has a set of markers - for it: @PRE and @@PRE, @CHECK and @@CHECK, @POST and @@POST, @RULES and - @@RULES. The end-markers are optional. Sections below detail the particular - actions available in each category.

- -

Rule Syntax

- -

Within each Rules Subregion, any - number of rules can occur. The syntax for a rule is

- -

 _SUGG - <- _PHRASE @@

- -

The _SUGG, or suggested rule - element is followed by the rule arrow, - then by a phrase of elements, - and finally the @@ end-mark. Each element, including the suggested element, - consists of a literal or nonliteral. Literals match tokens, - or "real text", while nonliterals match nodes with synthetic - names. A nonliteral must begin with an underscore, a literal must not - begin so.

- -

Optionally, each element can be followed by a list of key=values pairs - enclosed in square brackets. Though called pairs, some keys, such as trigger, - take no values. Some special rule elements, as well as the key=value pairs - that appear in phrase and suggested elements, are listed in tables below.

- -

Pass File Tokens

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

ITEM

 -

DESCRIPTION

-

#

 -

Comment. Text till the end of the line is - ignored.

-

AAA or NNN

 -

Literal tokens. Completely alphabetic or completely - numeric.

-

_AAA

 -

Nonliteral tokens must begin with underscore. - All the remaining letters MUST be alphabetic. (May expand to underscore,num,dash - someday.)

-

"XXX"

 -

Strings get double quoted. Everything within - string taken literally (except backslashed characters).

-

\X

 -

Backslash any nonalphanumeric char you want - taken literally.

-

\\

 -

Literal backslash character.

-

\

 -

At end of line, escapes end of line to NOTHING. - [Unimplemented]

-

\X

 -

where X=abnrftv special escapes in C/C++.

-

@AAA

 -

Reserved start-of-region markers in the pass - file (e.g., @RULES).

-

@@AAA

 -

Reserved end-of-region markers in the pass - file (e.g., @@RULES).

-

<-

 -

Rule arrow.

-

@@

 -

Rule terminator.

- -

Punctuation

- -

Each punctuation character is treated as a separate entity or "token" - in the rule file, even if there is no separating space. Constructs like - _xWILD [matches=(_noun)] and _xWILD [ matches = ( _noun ) ] are parsed - identically.

- -

Larger Structures

- - - - - - - - - - - - - - - -
-

[AAA=XXX - AAA=(XXX XXX) AAA]

 -

PAIRS. Square brackets enclose the key=value - pairs for a rule element (phrase element or suggested element). A sample - rule with pairs:
-_np <- _adj _xWILD [match=( a b) ren="noun" s] _noun [trigger] - @@

-

AAA(XXX,XXX....)

 -

ACTION. check and post action language.

-

<NUM,NUM> - XXX(XXX,XXX...)

 -

PRE ACTION. <NUM,NUM> is a range of - rule elements that must execute the action during rule matching.

- -

 
-ALL MATCHING IS CASE-INSENSITIVE. The rule-writer can add restrictive actions, - like uppercase(), as needed. Currently, wildcard cannot appear as the - start, end, or trigger of a rule. Optional elements cannot be triggers.

- -

Rule Elements

- -

Special rule elements are given in the following table.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

RULE - ELEMENT

 -

DESCRIPTION

-

_xWILD

 -

Unrestricted wildcard. Key-value pairs may - add restrictions on number of nodes matched and on what is matched. (Note: - with a match or fail - list, _xWILD becomes an "OR" matching function.)

-

_xANY

 -

Matches any single node. (See warning [1] - below.)

-

_xNIL

 -

Designates a suggested element when the rule - performs a special action, such as removing the matched nodes from the - parse tree. (_xNIL has no special action, but serves as documentation - for the rule writer.)

-

_xALPHA

 -

Matches an alphabetic token, including accented - and other extended ANSI chars.

-

_xCTRL

 -

Matches control and nonalphabetic extended - ANSI characters. (see _xALPHA)

-

_xNUM

 -

Matches a numeric token.

-

_xPUNCT

 -

Matches a punctuation token.

-

_xWHITE

 -

Matches a whitespace token, including - newline.

-

_xBLANK

 -

Matches a whitespace token, excluding - newline. Equivalent to _xWILD - [match=(\ \t)]

-

_xCAP

 -

Matches an alphabetic with uppercase first - letter.

-

_xEOF

 -

Matches the end of file.

-

_xSTART

 -

Matches if at the start of a phrase (or "segment").

-

_xEND

 -

Matches if at the end of a phrase (or "segment").

- -

Pairs for a Phrase Element

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

KEY

 -

VALUE

 -

DESCRIPTION

-

trigger,trig,
-t

 -

NONE

 -

Match the current element first. E.g.,
-_np <- _det _quan - _adj _noun [t] @@

-

min

 -

NUM

 -

Match a minimum of NUM nodes. 0 means the - current element is optional. E.g., _boys - <- the [min=0 max=1] boys @@

-

max

 -

NUM

 -

Match a maximum of NUM nodes. 0 means the - current element can match an indefinite number of nodes. E.g., _htmltag - <- \< _xWILD [min=1 max=100] \> @@

-

optional,option,opt,
-o

 -

NONE

 -

Optional element. Match a minimum of 0 and - a maximum of 1 node. (Short for min=0 max=1). E.g., _vgroup - <- _modal [opt] _have [opt] _be [opt] _verb @@

-

one

 -

NONE

 -

Match exactly one node. (Short for min=1 max=1).

-

star

 -

NONE

 -

Indefinite repetition. Match a minimum of - 0 up to any number of nodes. (Short for min=0 max=0).

-

plus

 -

NONE

 -

Indefinite repetition. Match a minimum of - 1 up to any number of nodes. (Short for min=1 max=0).

-

group, gp

 -

NAME

 -

Gather nodes that matched elt under a single - node with NAME as its name. Analogous to group post-action.

-

rename,ren

 -

NAME

 -

Rename every node that matched the current - element to NAME. E.g., _locfield <- - location \: _xWILD [ren=_location] \n @@

-

singlet,
-s

 -

NONE

 -

Search a node's descendants for a match. Stop - looking down when a node has more than one child or has the BASE attribute - set (see base below). E.g., _abbr <- - _unk \. [s] @@

-

tree

 -

NONE

 -

Search node's entire subtree for a match. - (Overuse of this may degrade analyzer performance.)

-

matches,
-match

 -

LIST

 -

For _xWILD element only. Restricted wildcard - succeeds only if one of the list names matches a node. E.g., _james - <- _xWILD [match=(jim - jimmy james) singlet min=1 max=1] - @@

-

fails,
-fail

 -

LIST

 -

For ANY - element. Match fails if node matches anything on the list. E.g., _par <- _xWILD [fail=(_endofpar - _par) min=1 max=0] @@

-

excepts, - except

 -

LIST

 -

For ANY element. Must be accompanied by a - single match or fail - list. Matching an item on the except list negates the effect of a match - on the match or fail list.

-

lookahead, - look

 -

NONE

 -

Designates the first lookahead element of - a rule. The first node matching the lookahead element or to the right - of it will be the locus where the pattern matcher continues - matching. WARNING: A reduce action such as singler - or noop should be used to ensure - that the lookahead node and nodes to its right are not included in the - current rule reduction.

-

layers, layer,
-attrs, attr

 -

LIST

 -

Layer additional attributes for the element - in the parse tree, as "mini-reductions". Use the names in the - list to name nodes. Note: each node that matched current rule element - will be layered.

-

recurse, - nest,
-passes, pass

 -

LIST

 -

Invoke a recursive rules pass on nodes that - matched the current rule element. E.g., _tag - <- \< _xWILD [nest=(tagrules)] \> @@. (See - the Recurse Region discussion.)

- -

WARNING:
-[1] Don't write rules with _xANY [max=0]. They won't work, because _xANY - is not implemented as a wildcard. The nearly equivalent _xANY _xWILD works.
-
-[2] More generally, matching with MIN and MAX is greedy, - and no backup is implemented. - So the rule
-
-  _a <- _b [min=1 max=3] _c [s] @@

- -

fails on the phrase _b _b(subnode _c), - because the repeating _b element greedily gobbles up the second _b node. - The pattern matcher will not back up to find that the rule can be made - to work. Unrestricted _xWILD is the only - rule element that is "backup-aware".

- -

Pairs for a Suggested - Element

- - - - - - - - - - - - - - - - - - -
-

base

 -

NONE

 -

The suggested node is the bottom-most node - to search when looking down the parse tree for a match (see singlet above). - (Takes no value).

-

unsealed

 -

NONE

 -

The suggested - node will be searched for select nodes (i.e., nodes specified by @NODES).

-

layers
-layer
-attrs
-attr

 -

LIST

 -

After normal reduce, perform additional reduces, - naming the nodes as in the list. This enables layering of attributes in - the parse tree.

- -

NLP++ Functions

- -

The code, pre, check, and post actions - listed in subsequent sections are still valid. However, a suite of general - functions that can occur in any region is being implemented, and some - of the old actions may go away or be generalized to work in any region - that allows NLP++ code. Reduction and parse tree manipulation functions - should still go in the post region, because that code executes when a - rule match is deemed to be successful. The check region may be folded - into the post region, since NLP++ functions can be implemented to pronounce - that a rule match has succeeded or failed. The pre region still does not - allow general NLP++ expressions. NLP++ functions are described - in the NLP++ and Knowledge Base help pages.

- -

WARNING: Many of the old-style actions are not recognized by the compiled - analyzer engine. Use NLP++ expressions wherever possible, rather than - the old hard-wired actions. Compiling with MS Visual C++ points out which - functions are invalid. The documentation below also marks old-style actions - slated to be removed from NLP++.

- -

Output Actions

- -

Code actions and rule - actions (pre, check and post) - are meant to work together in many cases. The code region in a file operates - before any rule-matching is invoked, so that is where the user should - initialize variables and perform other functions, such as printouts that - are independent of any rule.

- -

There are three output modes: (1) - output to standard output, (2) output to a file named in the actions themselves, - and (3) output to the "main output file" (default name output.txt) that is set by the environment that calls - the analyzer. Then we throw into the mix actions that print in the code - region as well as actions that print when a rule matches.

- -

Rather than have three types of output - (or "print") actions, we maintain two: actions that take a file - argument and equivalents that do not. By default, output actions that - take no file argument print to standard output. However, when delimited - by the code actions startout() and stopout(), these actions print to the main output - file, rather than standard output. If print actions in the rules are also - to write to the main output file, then the stopout() action must be placed in the code region - of a subsequent pass.

- -

For output actions with a filename - argument, the file must have been previously declared with a fileout() - action.

- -

Output is complex, but to allow the - analyzer developer as much flexibility as possible, we must support many - modes.

- -

Code Actions

- -

Code actions are not tied to any - rule, but execute operations for the entire pass. They are placed within - the @CODE and @@CODE markers at the head of the file, called the code region. There is only one such region per file, - and its actions operate before the pattern matching of the rules in the - file. This region represents an embryonic programming language that enables - the analyzer developer to perform useful tasks without having to write - and compile regular programming language code.

- -

Code actions operate before - the rules in the file operate on the parse tree. At present, there is - no "post code" region to execute actions after the rules for a pass have matched. The user - must place such code actions in a subsequent analyzer pass.

- -

NLP++ is now valid in the code region.

- -

Variable names must be strings, at - present.

- -

A. VARIABLE ACTIONS

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

ACTION(ARG1,ARG2...)

 -

DESCRIPTION

-

var(varname,str)

 -

Create global - variable with name varname and - initial value str. If str2 is - all numeric, then the code action inc() - can increment the value of the variable (this implements a counting variable). - (Old-style. Use NLP++)

-

varstrs(varname)

 -

Create an empty multi-string-valued global variable with name varname. - The post action addstrs() adds - values to this type of variable.

-

sortvals(varname)

 -

Sort the strings in multi-string-valued global variable varname.

-

gtolower(varname)

 -

Convert the strings in multi-string valued - global variable to lower case.

-

guniq(varname)

 -

Remove redundancies in a sorted, multi-string - valued global variable.

-

lookup(var,file,flag)

 -

Specialized word lookup. Global - variable var - has multiple words as values. file - is a file of strings, one per line. flag - tells which bit-flag of the word's symbol table entry to modify. E.g., - lookup("Words", "dict.words", - "word") looks up all the values in the Words - variable in the dict.words file - and modifies the word bit-flag - (which says whether the word is a proper English word).

- -

B. PRINT ACTIONS

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

print(str)

 -

Print str to the standard output (usually - the console or DOS window). (See startout - below).

-

printvar(var)

 -

Print the values of the global - variable named var to standard - output. (See startout - below).

-

fprintvar(file, - var)

 -

Print the values of the global - variable named var to file.

-

prlit(file, - str)

 -

Print the literal string str - to file.

-

gdump(filename)

 -

Dump the all the global variables and their - values to the given filename.

-

fileout(file)

 -

Open file - for appending. File becomes a - variable useable in print actions with a file argument, e.g. prlit().

-

startout(0)

 -

Divert standard output to the main output - file, set by the caller of the analyzer (default is output.txt).

-

stopout(0)

 -

Stop diverting standard output to the main - output file. Subsequent file-less output is to standard output.

- -

 

- -

Pre Actions

- -

Pre actions extend the pattern matcher's - mechanism for matching each rule element. Unlike other actions, pre actions - have a mandatory range. Their form is

- -

<num,num> - pre_action_name(args....)

- -

The pre action applies to each rule - element in the <num, num> range individually.

- -

Pre actions operate only on tokens, - or leaf nodes of the parse tree, at present. So, if the current node covers - more than a single leaf node, the pre action automatically - fails. Another thing to be careful of is that - pre actions disregard the base attribute of a node. That is, even if the - current node or a subnode has the base attribute set, the pre action will look - under it for the leaf node.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

uppercase()

 -

Succeed if leaf token is all-uppercase.

-

lowercase()

 -

Succeed if leaf token is all-lowercase.

-

cap()

 -

Succeed if leaf token has capitalized first - letter.

-

length(num)

 -

Succeed if leaf token has length exactly equal - to num.

-

lengthr(num1,num2)

 -

Succeed if leaf token length is in the range - of lengths (num1,num2), inclusive.

-

numrange(num1,num2)

 -

Succeed if leaf token is numeric in the given - range.

-

unknown()

 -

Succeed if leaf token is an unknown word. - Meaningful only if a prior pass has performed lookup() - code action.

-

debug()

 -

Succeed unconditionally. Used to place a C++ - breakpoint at a particular rule.

- -

Check Actions

- -

Check actions apply after the pattern - matcher has succeeded in matching a rule. These represent the final opportunity - to reject the rule match based on inconsistencies spanning more than one - rule element. For example, before accepting a noun-phrase rule, we could - check for agreement between the determiner and noun (e.g., accept "this - company" but reject "these company").

- -

See the section on NLP++ for a more - generalized way to express conditions in the Check region.

- - - - - - - -
-

or(num1,num2)

 -

Succeed if any of the rule elements from num1 to num2 - has matched an actual node. Meaningful only if this is a range of optional rule elements. (This action - is just a sample. As Text Analysis International develops its generic analyzer, more check actions will - be implemented.)

- -

Post Actions

- -

Post actions execute after the pattern - matcher has successfully matched a rule and the match is accepted.

- -

A. PARSE TREE - ACTIONS

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

single()

 -

Single-tier reduce. Reduce the entire set - of nodes that matched a rule phrase.

-

singler(num1,num2)

 -

Single-tier reduce of a range of rule elements. - eg, if finding a period to be end-of-sentence in context, we only want - to reduce the period to end-of-sentence, not the whole context.

-

singlex(num1,num2)

 -

Single-tier reduce of a range of rule elements. - All nodes not in the range are EXCISED. eg, if matching a keyword html - tag, we only want to reduce the keywords and to zap the rest of the tag.

-

singlezap()

 -

Single-tier reduce. Excise all the nodes in - the matched phrase.

-

merge()

 -

Single-tier reduce that DISSOLVES each top-level - node in the matched phrase.

-

merger(num1,num2)

 -

Single-tier reduce that DISSOLVES each top-level - node in the matched range.

-

listadd(olist,oitem - [, keep])

 -

Add a new node to a list node's children. - If the item occurs after the list (olist - < oitem), - it is added as the last child. If the item occurs before the list, it - is added as the first child. The optional keep arg can be "true" - or "false". If "true", it keeps the nodes between - the list and the item as children of list. If "false", it excises - all the intervening nodes.

-

excise(num1,num2)

 -

Excise the nodes matching the range of elements - from the parse tree. For multiple excise actions, operations must be ordered - such that the highest number is listed first in the operation.

-

splice(num1,num2)

 -

DISSOLVE the top level nodes of given range. - [NOT TESTED whether a reduce such as single can follow this action.]

-

xrename(name - [, num])

 -

Rename the numth - context node to name. - If num is - absent or 0, rename last context node.

-

setbase(num, - bool)

 -

Set the BASE attribute of the numth - node to "true" or "false".

-

setunsealed(num, - bool)

 -

Set the UNSEALED attribute of the numth - node to "true" or "false".

-

group(num1,num2,label_str)

 -

Reduce the range of rule elements (num1,num2) - and name the group node label. - Unlike other reduce actions, this one can be repeated. WARNING: - The rule element numbers change after each group action, so subsequent - post actions must account for the new numbering of the rule elements. - Reimplemented as an NLP++ function so that args can be NLP++ expressions.

-

noop(0)

 -

Perform no post action. This is a convenience - to disable the default single() - reduce action.

- -

 

- -

B. PRINT - ACTIONS

- -

Code actions and Post actions have distinct sets of printing actions, - though we've made them as similar as possible.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

print(str)

 -

Print the literal string str - to standard output.

-

printr(num1,num2)

 -

Print the text for the rule element range - from num1 to num2 - to standard output.

-

prchild(file, - num, name)

 -

Look for name - immediately under the node matching rule element num. - Print its text to file, if found.

-

prtree(file, - num, name)

 -

Look for name - anywhere under the node matching rule element num. - Print its text to file, if found.

-

prxtree(filename, prestr,
-  ord, name, poststr)

 -

To the filename, - print the first node called name - found in the ordth element's tree, - preceded by prestr and followed - by poststr. If named node is not - found, nothing is printed out. E.g., prxtree("out.txt", "date: - ", 3, "_date", "\n") prints out a line like "date: - 3/9/99 <cr>" if a _date node is found within the subtree of - elt 3.

-

prlit(file, - str)

 -

Print the literal string str - to file.

-

fprintnvar(file, - var, ord)

 -

To the file, - print the value of the variable var - in the ordth element's node.

-

fprintxvar(file, - var, ord)

 -

To the file, - print the value of the variable var - in the ordth context node.

-

fprintgvar(file, - var)

 -

To the file, - print the value of the global variable var.

-

gdump(file)

 -

Dump all the global variables and their values - to file.

-

xdump(file, - ord)

 -

Dump all the variables in the ordth - context node and their values to file.

-

ndump(file, - ord)

 -

Dump all the variables in the ordth - phrase element's node and their values to file.

-

sdump(file)

 -

Dump all the variables in the suggested node - and their values to file.

-

prrange(file, - num1,num2)

 -

Print the text under a range of rule elements - (num1,num2) to - file.

-

pranchor(file,num1,num2)

 -

Print a web URL to file. - Treat the range (num1,num2) as - a URL. Use the global variable named Base - to resolve and print complete relative URLs. (a prior pass must find the - <base> HTML tag and set Base - appropriately. NOT yet using the file's URL, so not handling cases where - <base> itself is a relative URL.)

- -

 

- -

C. VARIABLE ACTIONS

- - - - - - - -
-

addstrs(var,num)

 -

Add a string to the multi-string global - variable var. Fetch the string - from the nodes matching rule element num. - var should have been declared - using the code action varstrs(). - (No NLP++ equivalent yet.)

- -

A batch of post actions for the RFA - (Rules-File Analyzer) are not described, as they are not relevant to the - VisualText user. Their names - all begin with the prefix RFA.

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Phrase_element_modifiers.htm b/Help/helps/NLP_PP_Stuff/Phrase_element_modifiers.htm deleted file mode 100644 index 8ce7b0e..0000000 --- a/Help/helps/NLP_PP_Stuff/Phrase_element_modifiers.htm +++ /dev/null @@ -1,772 +0,0 @@ - - - - - - - - - - -Phrase Element Modifiers - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Phrase Element Modifiers

- -

Element modifiers that can be used in a rule's phrase are listed in - the table below. The element modifier must be enclosed in square brackets [ ] and placed after - the element. Some element modifiers have synonyms.  The - modifier trigger, for example can be written as [trigger], - [trig], or [t]. -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

KEY

 -

VALUE TYPE

 -

DESCRIPTION

-

deaccent, deacc, da

 -

NONE

 -

Match alphabetic, ignoring presence or absence - of accent marks on characters. NEW - IN VERSION 2.

-

trigger,trig,
-t

 -

NONE

 -

Match the current element first. E.g.,
-_np <- _det _quan - _adj _noun [t] @@

-

min

 -

NUM

 -

Specify the minimum number of nodes to match. -  0 means - the current element is optional. E.g., _boys - <- the [min=0 max=1] boys @@

-

max

 -

NUM

 -

Specify the maximum number of nodes to match. -  0 means - the current element can match an indefinite number of nodes. E.g., _htmltag <- \< _xWILD [min=1 max=100] - \> @@

-

opt,option,optional,
-o

 -

NONE

 -

Match a minimum of 0 and a maximum of 1 node. - (Short for min=0 max=1). E.g., _vgroup - <- _modal [opt] _have [opt] _be [opt] _verb @@

-

one

 -

NONE

 -

Match exactly one node. (Short for min=1 max=1).

-

star

 -

NONE

 -

Indefinite repetition. Match a minimum of - 0 up to any number of nodes. (Short for min=0 max=0).

-

plus

 -

NONE

 -

Indefinite repetition. Match a minimum of - 1 up to any number of nodes. (Short for min=1 max=0).

-

group, - gp

 -

NAME

 -

Gather nodes that matched the current element - under a single node with NAME as its name. Analogous to POST Action group.

-

rename,ren

 -

NAME

 -

Rename every node that matched the current - element to NAME. E.g., _locfield <- - location \: _xWILD [ren=_location] \n @@

-

singlet,
-s

 -

NONE

 -

Search a node's descendants for a match. Stop - looking down when a node has more than one child or has the BASE attribute - set. E.g., _abbr <- _unk \. [s] - @@  See - BASE.

-

tree

 -

NONE

 -

Search node's entire subtree for a match. - (Overuse of this may degrade analyzer performance.)

-

matches,
-match

 -

LIST

 -

For _xWILD element only. Match if element - is in a match list. Used as a restricted wildcard and succeeds only if - one of the list names matches a node. E.g., _james - <- _xWILD [match=(jim - jimmy james) singlet min=1 max=1] - @@

-

fails,
-fail

 -

LIST

 -

For ANY - element. Fail to match if node matches anything on the list. E.g., _par <- _xWILD [fail=(_endofpar - _par) min=1 max=0] @@

-

excepts, - except

 -

LIST

 -

For _xWILD only. Matching an item on the except - list negates the effect of a match on the match - or fail list.  Must - be accompanied by a single match - or fail list.

-

lookahead, - look

 -

NONE

 -

Designate the first lookahead element of a - rule. The first node matching the lookahead element or to the right of - it will be the locus where the rule matcher continues - matching. WARNING: A reduce action such as singler - or noop should be used to ensure - that the lookahead node and nodes to its right are not included in the - current rule reduction.

-

layers, - layer,
-attrs, attr

 -

LIST

 -

Layer additional attributes for the element - in the parse tree, as "mini-reductions". Use the names in the - list to name nodes. Note: each node that matched current rule element - will be layered.  (These - modifiers can also be used as suggested element modifiers.)

-

recurse, - nest,
-passes, pass

 -

LIST

 -

Invoke a recursive rules pass on nodes that - matched the current rule element. E.g., _tag - <- \< _xWILD [nest=(tagrules)] \> @@. (See RECURSE - Region.)

- - - - - - -
-

WARNING: -

-
    - -

  1. Don't - write rules with _xANY [max=0]. They won't work, because _xANY is not - implemented as a wildcard. The nearly equivalent _xANY _xWILD works.

    2. - -

  2. More generally, matching - with MIN and MAX is greedy, and - no backup is implemented. So - the rule
    -
    -   _a <- _b [min=1 max=3] _c [s] @@

    3. -
-

fails on the phrase _b _b(subnode _c), because the repeating - _b element greedily gobbles up the second _b node. The rule matcher will - not back up to find that the rule can be made to work. Unrestricted _xWILD - is the only rule element that - is "backup-aware".

- -

 See - Also

- -

Special - Rule Elements

- -

Suggested - Element Modifiers

- -

Rule - Syntax 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Rule_Element_Modifiers.htm b/Help/helps/NLP_PP_Stuff/Rule_Element_Modifiers.htm deleted file mode 100644 index 83bd613..0000000 --- a/Help/helps/NLP_PP_Stuff/Rule_Element_Modifiers.htm +++ /dev/null @@ -1,109 +0,0 @@ - - - - - - - - - - -Rule Element Modifiers - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Rule Element Modifiers

- -

Each rule element can be followed by optional - element modifiers.  Element - modifiers can be used for both - the suggested element (_SUGG) - and the phrase of elements (_PHRASE) - on the right-hand side of a rule.

- -

Element modifiers place constraints on how elements in a rule are matched. - For example, if you want to specify that the rule matcher should match - a specific element in the phrase of elements on a rule's right-hand side - first, you would use the trigger - element modifier with the rule element.

- -

The modifier must be enclosed in square - brackets [ - ] after the element. The element modifier is made - up of a key (the modifier name itself) and sometimes - a value. A value can be a parenthesized list, a name - or number. Some element modifiers do not have values. Some examples follow:

- -

_xWILD [s one match=(_adj _noun \n)]

- -

hello [rename=_hello]

- -

For a list of element modifiers used with the suggest rule element, - see Suggested Element Modifiers. - For a list of element modifiers for the phrase of elements, see Phrase - Element Modifiers.

- - - - - - -
-

WARNING: When - writing lists for an element modifier, we recommend not enclosing them - in double quotes, since they are not treated as strings. E.g., write [match=(_adj _adv)] rather than [match=("_adj" - "_adv")].

- -

See Also

- -

Rule Syntax

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Rule_syntax.htm b/Help/helps/NLP_PP_Stuff/Rule_syntax.htm deleted file mode 100644 index 9695ba8..0000000 --- a/Help/helps/NLP_PP_Stuff/Rule_syntax.htm +++ /dev/null @@ -1,219 +0,0 @@ - - - - - - - - - - -Rule Syntax - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Rule Syntax

- -

All rules are written in the RULES - Region of a pass file. Within each RULES Region, any number of rules can - occur.

- -

The syntax for a rule is

- -

 _SUGG - <- _PHRASE @@

- -

Rule Elements

- -

_SUGG, is referred to as the - suggested - rule element. It is the node that is created if a rule matches. - _SUGG is often referred to as the left-hand side of the rule.

- -

_PHRASE is the phrase - of elements that the rule matcher is trying to match. If the rule - matcher finds a match for the listed phrase, the node will suggest - or reduce to what is specified - in _SUGG. _PHRASE is referred to as the right-hand side of the rule.

- - - - - - -
-

Note: - Rule elements that make up the phrase of elements do not have to be written - on the same line. That is, the rule can extend over several lines with - each rule element in _PHRASE occupying its own line.  The - two rules below are equivalent:

-

_XYZ <- _ABC _DEF - @@

-

 

-

_XYZ <-

-

_ABC

-

_DEF

-

@@

-

When each rule element in the phrase is written - on its own line, the rule is said to be written in standard - rule format.  For - more information on rule formats, see Standard - Rule Format.

- -

<- is the rule arrow. It - divides the rule between the suggested element and the phrase of rule - elements.

- -

@@ indicates the end of a rule. - Rules must always end in the @@ marker.

- -

Rule Element Types

- -

Each rule element, including the suggested element, consists of a literal or a nonliteral. - What gets matched for each is different:

- - - -

A literal is a terminal node - (or leaf node) in the parse tree. It represents words, numbers and punctuation.

- -

A nonliteral is a nonterminal node - (also called an internal node). It represents a node which dominates other - nodes.

- -

The initial underscore character "_" is an NLP++ convention - marking nonterminal or internal nodes. A terminal node cannot begin with - an underscore character.

- -

Special Rule Elements

- -

NLP++ has a set of predefined rule elements called special - rule elements. These rule elements match various types of tokens - such as alpha characters, punctuation, wildcards, etc.. They have properties - that make writing rules easier. For a list of these predefined rule elements, - see Special Rule Elements.

- -

Rule Element Modifiers

- -

Each rule element can be followed by optional - element modifiers.  This - includes the suggested element - and phrase of elements on a rules - right-hand side.  

- -

Element modifiers place constraints on how elements in a rule are matched. - For example, if you want to specify that the rule matcher should match - a specific element in the phrase of elements on a rule's right-hand side - first, you would use the trigger - element modifier with the rule element. Element modifiers can be used - for both the suggested element - and the phrase of elements on the right-hand side of a rule.

- -

The modifier must be enclosed in square - brackets [ - ] after the element. The element modifier is made - up of a key (the modifier name itself) and sometimes - a value. A value can be a parenthesized list, a name - or number. Some element modifiers do not have values. Some examples follow:

- -

_xWILD [s one match=(_adj _noun \n)]

- -

hello [rename=_hello]

- -

For a list of element modifiers used - with the suggest rule element, see Suggested - Element Modifiers. For a list of element modifiers for the phrase - of elements, see Phrase Element - Modifiers.

- - - - - - -
-

WARNING: When - writing lists for an element modifier, we recommend not enclosing them - in double quotes, since they are not treated as strings. E.g., write [match=(_adj _adv)] rather than [match=("_adj" - "_adv")].

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Special_Variables.htm b/Help/helps/NLP_PP_Stuff/Special_Variables.htm deleted file mode 100644 index 4172706..0000000 --- a/Help/helps/NLP_PP_Stuff/Special_Variables.htm +++ /dev/null @@ -1,1215 +0,0 @@ - - - - - - - - - - -Special Variables - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Special Variables

- -

To get useful information from the parse tree and other sources when - analyzing a text, some special variable names have been provided. All - start with a dollar sign ($). - No such variable names are usable with suggested (S) - variables, at present.

- -

Notes on Special - Variables

- -
    - -

  1. Special variable - names must be quoted. E.g., N("$text", 2).

    2. - -

  2. You should - not assign to or create your own "dollar" variables. NLP++ reserves - all variable names that start with a dollar sign.

    3. -
- -

Types of Special Variables

- -

These special variables have been organized into the following four - groups:

- - - -

A consolidated table of NLP++ special variables is provided below. Each - item in the table is also detailed in individual topics in the Special - Variables section.

- -

Table of Special Variables

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

VARIABLE NAME

 -

VARIABLE TYPES

 -

DESCRIPTION

-

$treetext

 -

N, X

 -

Fetch the 'cleaned up' (removing leading and - trailing, convert separators to single space) text gleaned from a node's - subtree. Uses the parse tree as input, not text buffer. Analogous to $text. - Difference between $treetext and $text is $text returns text from the - input text buffer while $treetext returns text from the parse tree. So, - if you have removed some nodes from a tree with an excise(), $treetext - returns the text minus the excised elements, while $text returns everything - in the input text buffer, including the text under excised nodes.

-

$text

 -

N, X

 -

Fetch the text covered by the node. Cleanup - whitespace (removing leading and trailing, convert separators to single - space). Uses the original text buffer, rather than the subtree under the - node, in order to gather text. See also $treetext and $treeraw.

-

$treeraw

 -

N, X

 -

Fetch the text covered by the node from the - parse tree. Text is not 'cleaned up'. Analogous to $treetext. See also - $treetext and $raw.

-

$raw

 -

N, X

 -

Fetch the 'cleaned up' text covered by the - node from the text buffer. See also $treetext, $treeraw and $text.

-

$xmltext

 -

N, X

 -

Same as $raw, but converts characters - that are special to HTML and XML. E.g., "<" is converted - to "&lt;".

-

$length

 -

N, - X

 -

Get the length of node's - text.

-

$ostart

 -

N, X

 -

Start offset of the referenced node in the - input text.

-

$oend

 -

N, X

 -

End offset of the referenced node in the input - text.

-

$start

 -

N, X

 -

Check if referenced node has left sibling - in parse tree. Evaluates to 1 if - the referenced node has no left sibling, else 0.

-

$end

 -

N, X

 -

Check if referenced node has right sibling in parse tree. - Evaluates to 1 if referenced node - has no right sibling, else 0.

-

$apppath

 -

G

 -

Get base directory path for current application. - E.g., "D:\dev\apps\Resume".

-

$anaspath

 -

G

 -

Get parent directory for the analyzers. - NEW 2.1.0

-

$kbpath

 -

G

 -

Get base directory path for the kb / user directory. - E.g., "D:\dev\apps\Resume\kb\user". - NEW 1.27.0

-

$input

 -

G

 -

Get full input file. E.g., "D:\apps\Resume\input\Dev1\rez.txt"

-

$inputparent

 -

G

 -

$inputparent returns "Dev1" assuming input file is "D\apps\Resume\input\Dev1\text.txt". - NEW 1.26.0

-

$inputpath

 -

G

 -

Get input file path. E.g., "D:\apps\Resume\input\Dev1"

-

$inputname

 -

G

 -

Get input file name. E.g., "rez.txt"

-

$inputhead

 -

G

 -

Get input file head. E.g., "rez"

-

$inputtail

 -

G

 -

Get input file tail. E.g., "txt"

-

$passnum

 -

G

 -

Get current pass number.  NEW - 2.0.2.4

-

$rulenum

 -

G

 -

Get current rule number within current pass. - NEW 2.0.2.4

-

$isdirrun

 -

G

 -

Check if the engine is processing a directory of files instead of an individual file. Returns 1 if it is analyzing a directory, else 0. - NEW 1.26.0

-

$isfirstfile

 -

G

 -

Check if is first file processed in a directory. Returns 1 if it is the first file, else 0. - NEW 1.24.0

-

$islastfile

 -

G

 -

Check if is last file processed in a directory. Returns 1 if it is the last file, else 0. - NEW 1.24.0

-

$allcaps
-$uppercase

 -

N

 -

Check if token underlying node is all uppercase. -  Returns - 1 if token is all uppercase. If multiple words (even if all are all-caps), - returns 0. WARN: Not identical to the uppercase() PRE Action.

-

$lowercase

 -

N

 -

Check if token underlying node is all lowercase. - Returns 1 if token is all lowercase If multiple words (even if all are - all-caps), returns 0. WARN: Not identical to the lowercase() PRE Action.

-

$cap

 -

N

 -

Check if token underlying node is a capitalized - word. Returns 1 if token is a capitalized word. WARN: Not identical to - the cap() - PRE Action.

-

$mixcap

 -

N

 -

Check if token underlying node is a mixed-capitalized - word. Returns 1 if token is a mixed-capitalized word. Examples: "MIchigan", - "abcD".

-

$exists

 -

N

 -

Check if node exists in parse tree. Returns - 1 if node exists, else 0.

-

$unknown

 -

N

 -

Check if token underlying node is an unknown - word. Returns 1 if token underlying node is unknown. WARN: Not identical - to the unknown() - PRE Action. Requires a lookup() - CODE Action prior to any use of this special variable.

- - - - - - -
-

Notes: - The $uppercase, $lowercase, $cap, and - $unknown differ from the analogous PRE Actions in that they all traverse - down to the "leafiest" node they can find, in order to perform - their check. The corresponding PRE Actions only chain down if the [s] - flag is present in the associated phrase element(s). Also, the $ variables - are not stopped by a [base] attribute in a node.

- -

See Also

- -

Variables

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Special_rule_elements.htm b/Help/helps/NLP_PP_Stuff/Special_rule_elements.htm deleted file mode 100644 index 8e7b499..0000000 --- a/Help/helps/NLP_PP_Stuff/Special_rule_elements.htm +++ /dev/null @@ -1,439 +0,0 @@ - - - - - - - - - - -Special Rule Elements - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Special Rule Elements

- -

NLP++ has a set of predefined rule elements that make writing rules - easier.  These - special rule elements match various - types of tokens such as alpha characters, punctuation, wildcards, etc. -  For examples - of these special rule elements, refer to the individual pages included - in the Special Rule Elements section.

- -

Special rule elements are given in the following table.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

RULE - ELEMENT

 -

DESCRIPTION

-

_xWILD

 -

Match anything.  _xWILD - is an unrestricted wildcard. Key-value pairs may add restrictions on number - of nodes matched and on what is matched. (Note: with a match - or fail list, _xWILD becomes - an "OR" matching function.)

-

_xANY

 -

Match any single node. (WARNING:  A - rule written with _xANY [max=0] will not work.  _xANY - is not implemented as a wildcard.  Instead, - the near equivalent _xANY _xWILD should be used.)

-

_xNIL

 -

Match nothing.  _xNIL - designates a suggested element when the rule performs a special action, - such as removing the matched nodes from the parse tree. (_xNIL has no - special action, but serves as documentation for the rule writer.)

-

_xALPHA

 -

Match an alphabetic token, including accented - and other extended ANSI chars.

-

_xCTRL

 -

Match control and nonalphabetic extended ANSI - characters. (See _xALPHA.)

-

_xNUM

 -

Match a numeric token.

-

_xPUNCT

 -

Match a punctuation token.

-

_xWHITE

 -

Match a whitespace token, including - newline.

-

_xBLANK

 -

Match a whitespace token, excluding - newline. Equivalent to _xWILD - [match=(\ \t)]

-

_xCAP

 -

Match an alphabetic with uppercase first letter.

-

_xCAPLET

 -

Match an alphabetic consisting of a single - capitalized letter. NEW

-

_xLET

 -

Match an alphabetic consisting of a single - letter. NEW

-

_xEOF

 -

Match the end of file.

-

_xSTART

 -

Match if at the start of a phrase (or "segment").

-

_xEND

 -

Match if at the end of a phrase (or "segment").

- -

See Also

- -

Phrase - Element Modifiers
-Suggested Element Modifiers

- -

Rule - Syntax

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Statements_and_blocks.htm b/Help/helps/NLP_PP_Stuff/Statements_and_blocks.htm deleted file mode 100644 index bf1749c..0000000 --- a/Help/helps/NLP_PP_Stuff/Statements_and_blocks.htm +++ /dev/null @@ -1,116 +0,0 @@ - - - - - - - - - - -Statements and Blocks - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Statements and Blocks

- -

Expression Statement

- -

NLP++ expressions can be treated as statements by terminating them with - a semicolon.

- -

Statements and Blocks

- -

This too is modeled after the C programming language. Curly braces { and } - delimit blocks of statements. Statements should end with a semicolon (;). A statement can be an arbitrary - expression, a block, or fully nested if-else - statements.

- -

Assignment Statement

- -

Assignment is treated as an expression rather than as a statement. As - in C/C++, assignments can be chained.

- -

If-Else Statement

- -

NLP++ supports if-else statement syntax identical to C/C++.

- -

While Statement

- -

The only loop construct currently supported is the while - loop. Here's an example that prints out 0 thru 9:

- -

G("counter") = 0;
-G("loop_limit") = 10;
-while( G("counter") < G("loop_limit")) {
-"output.txt" << "Loop counter = " << G("counter") << "\n";
-G("counter")++;
-}

- -

See Also

- -

Code - Syntax

- -

Tokens

- -

Variables

- -

Data - Types

- -

Special - Variables

- -

Functions - and Actions

- -

Operators - and Expressions

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Suggested_element_modifiers.htm b/Help/helps/NLP_PP_Stuff/Suggested_element_modifiers.htm deleted file mode 100644 index d3da221..0000000 --- a/Help/helps/NLP_PP_Stuff/Suggested_element_modifiers.htm +++ /dev/null @@ -1,220 +0,0 @@ - - - - - - - - - - -Suggested Element Modifiers - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Suggested Element Modifiers

- -

Element modifiers for the suggested element are listed in the - table below. The element modifier must be enclosed in square brackets [ ] and placed after the element. The element - modifier layers has several synonyms, - layer, attrs and attr.

- - - - - - - - - - - - - - - - - - - - - - - -
-

KEY

 -

VALUE TYPE

 -

DESCRIPTION

-

base

 -

NONE

 -

The suggested node is the bottom-most node - to search when looking down a singlet - chain for a match. See also the singlet - modifier.

-

unsealed

 -

NONE

 -

The suggested - node will be searched for select nodes (i.e., nodes specified after @NODES - marker).

-

layers
-layer
-attrs
-attr

 -

LIST

 -

After normal reduce, perform additional reduces, - naming the nodes as in the list. This enables layering of attributes in - the parse tree.

- -

See Also

- -

Phrase - Element Modifiers

- -

Rule - Syntax

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Tokens.htm b/Help/helps/NLP_PP_Stuff/Tokens.htm deleted file mode 100644 index 7ec83a8..0000000 --- a/Help/helps/NLP_PP_Stuff/Tokens.htm +++ /dev/null @@ -1,224 +0,0 @@ - - - - - - - - - - -Tokens - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tokens

- -

Following are some token types that occur in NLP++ code.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

ITEM

 -

DESCRIPTION

-

#

 -

Comment. Text till the end of the line is - ignored.

-

AAA or NNN

 -

Literal tokens. Completely alphabetic or completely - numeric.

-

_AAA

 -

Nonliteral tokens must begin with underscore. - All the remaining letters MUST be alphabetic. Currently, only alphabetics - are allowed after the initial underscore.

-

"XXX"

 -

Strings get double quoted. Everything within - string taken literally (except backslashed characters).

-

\X

 -

Backslash any nonalphanumeric char you want - taken literally.
-If X = one of a b n r f t v, these are special escapes analogous to C/C++.

-

\\

 -

Literal backslash character.

- -

 

- -

See Also

- -

Code - Syntax

- -

Variables

- -

Data - Types

- -

Special - Variables

- -

Functions - and Actions

- -

Operators - and Expressions

- -

Statements - and Blocks

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/Variable_types.htm b/Help/helps/NLP_PP_Stuff/Variable_types.htm deleted file mode 100644 index a86c16e..0000000 --- a/Help/helps/NLP_PP_Stuff/Variable_types.htm +++ /dev/null @@ -1,545 +0,0 @@ - - - - - - - - - - -Variable Types - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Data Types

- -

NLP++ data types are shown in the following table.

- -

Numeric overflow and underflow are not flagged.  For - example, the maximum int is 2147483647 - in Windows and must be checked by the programmer as needed.  For - large numbers, the float type - should be used.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

NAME

 -

ABBREVIATION

 -

CHARACTERISTICS

-

string

 -

STR

 -

Sequence of characters: e.g., G("name") - = "w3c";

-

integer

 -

INT

 -

32-bit integers: e.g., G("num") - = 32776;

-

float

 -

FLOAT

 -

32-bit floating point numbers: e.g., G("flt") = 123.456;

-

boolean

 -

BOOL

 -

1 or 0. (Boolean implemented as a numeric - nonzero for true and the numeric 0 for false.)

-

ostream

 -

OSTREAM

 -

Output stream. G(filename) << expr; - prints by referring to the ostream value of G(filename). - See the openfile - NLP++ function.

-

concept

 -

CON

 -

Concepts are the main object in the knowledge - base. They have string names, and optionally have one or more attributes. - e.g. G("myConcept") = makeconcept(G("root"),"My - Name");

-

attribute

 -

ATTR

 -

Attributes are associated with concepts and - KB nodes. They have string names, and optionally have one or more values. - e.g. G("myAttribute") = addattr(G("myConcept"),"is");

-

value

 -

VAL

 -

Values are the values of KB attributes. They - can be strings, numerics, or concepts: e.g. addconval(G("myConcept"), - "is", "John Smith");

-

phrase

 -

PHR

 -

A phrase is a sequence of nodes owned by a - single concept. E.g., create the first node of a phrase: G("USPhrase") - = makephrase(G("US"),"United"); Add - the last node of the phrase: addcnode(G("US"), - "States");

-

node

 -

NODE

 -

Node in the KB. Differs from a concept in - that it is not in the KB hierarchy, but rather in a concept's phrase. - A node serves as a "proxy" or "instance" of a concept.

-

array

 -

ARRAY

 -

The values of an array, or multi-valued variable, - can be any simple values.  Note - that nested ARRAY types are not currently supported.

-

pnode

 -

PNODE

 -

Parse tree node. Unrelated to a KB node.

- -

Parse Tree Nodes

- -

NLP++ can also reference the nodes of a parse tree directly. For example, - N(2) refers to the (first) node - that matched the 2nd element of a rule's right-hand-side phrase. X(3) refers to the 3rd node in the current - context path as specified by @PATH. These constructs are used in functions - that deal with the parse tree.  For - more information on parse tree functions, see Parse - Tree Functions.

- - - - - - -
-

Note: - Knowledge base nodes are unrelated to parse tree nodes. The former are - called NODE and the latter PNODE.

- -

See Special - Node References for more on referencing parse tree nodes.

- -

Arrays

- -

Arrays can consist of combinations of simple types. Nested array types - are not implemented. An example follows.

- -

# Create an array and step through it.
-G("counter") = 0;
-G("array_size") = 10;
-# Arrays can optionally be "declared" by padding them with 0.
-G("int_array")[G("array_size") - 1] = 0;
-
-while ( G("counter") < G("array_size") ) {
-G("int_array")[G("counter")] = G("counter");
-"output.txt" << G("int_array")[G("counter")] << "\n";
-G("counter") ++;
-}
-# print it out all at once
-"output.txt" << G("int_array") << "\n";.
-# copy it all at once:
-G("another_array") = G("int_array");

- -

 

- -

See Also

- -

Code - Syntax

- -

Tokens

- -

Variables

- -

Special - Variables

- -

Functions - and Actions

- -

Operators - and Expressions

- -

Statements - and Blocks

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/NLP_PP_Stuff/pncopyvars_action.htm b/Help/helps/NLP_PP_Stuff/pncopyvars_action.htm deleted file mode 100644 index 69fadb8..0000000 --- a/Help/helps/NLP_PP_Stuff/pncopyvars_action.htm +++ /dev/null @@ -1,46 +0,0 @@ - - - - - - - - - - -pncopyvars_action - - - - - - - - - - - - - - - - - - - - - - - - - - - -

pncopyvars

- -

See pncopyvars - in the Functions section.

- - - - diff --git a/Help/helps/NLP_PP_Stuff/today_format.htm b/Help/helps/NLP_PP_Stuff/today_format.htm deleted file mode 100644 index 2d924f2..0000000 --- a/Help/helps/NLP_PP_Stuff/today_format.htm +++ /dev/null @@ -1,46 +0,0 @@ - - - - - - - - - - -today_format - - - - - - - - - - - - - - - - - - - - - - - - - - - -

today

- -

See today in the - Special Functions section.

- - - - diff --git a/Help/helps/NODES.htm b/Help/helps/NODES.htm deleted file mode 100644 index 96fb59f..0000000 --- a/Help/helps/NODES.htm +++ /dev/null @@ -1,126 +0,0 @@ - - - - - - - - - - -@NODES Selector - - - - - - - - - - - - - - - - - - - - - - - - - - - -

NODES Selector

- -

The NODES Selector allows you to specify a list - of node names to be searched for independently in the parse tree. The - phrase immediately under such a context node will be traversed to look - for rule matches.

- -

Using the NODES Selector

- -

Below are some cautionary notes on using the NODES Selector.

- -

Common Error 1: Selecting Nodes inside Sealed - Nodes

- -

The selection algorithm will only find such a node when all of its ancestors - have the unsealed attribute set.

- -

Example: - Say in one pass you have a rule

- -

_paragraph <- _sentence [s plus] @@

- -

and in the next pass you try to match within the _sentence nodes as - follows:

- -

@NODES _sentence

- -

_np <- _det [s star] _quan [s star] _adj [s star] _noun [s plus] - @@

- -

When you run this analyzer on a text, the rule creating an _np (i.e., - noun phrase) will never apply, assuming all sentences are within paragraphs. - The reason is because _paragraph nodes are sealed, - that is, they don't have the unsealed - attribute set. In order for the second pass to succeed, the rule in the - first pass must be written as

- -

_paragraph [unsealed] <- _sentence [s plus] @@

- -

Having NODES look only within unsealed contexts enables you to build - more efficient analyzers. The idea is that you will flag "roughed - out" segments as unsealed, so that you may characterize them further - later on. But this mechanism may sometimes have you scratching your head - as to why a rule didn't match.

- -

Common Error 2: Nested Context Nodes

- -

If one context node is inside the subtree of another context node, then - the buried one will not be seen - by the selection algorithm.

- -

Common Error 3: Using the X() Context Variable - with a Position Number

- -

With NODES as the selector in a pass file, it is usually a mistake to - write expressions such as X(2) or X("varname", 3). That is because - these variables refer to the second and third node, respectively, when - traversing down a parse tree path, - starting at the root node. That is, they are appropriate for use with - the PATH Selector.

- -

Because the names in a NODES Selector are independent of each other, - you will, in general, not know the distance of the current context node - from the root of the parse tree. So, in this case, the correct variables - are X() and X("varname"), which refer to the current context - node selected by NODES.

- -

See Also

- -

Context - Zone

- -

SELECT - Region

- -

PATH - Selector

- -

MULTI - Selector

- -

 

- -

 

- - - - diff --git a/Help/helps/New_in_Current_Version.htm b/Help/helps/New_in_Current_Version.htm deleted file mode 100644 index 497dced..0000000 --- a/Help/helps/New_in_Current_Version.htm +++ /dev/null @@ -1,1010 +0,0 @@ - - - - - - - - - - -New in Current Version - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

MAJOR RELEASE (2.0 NON-UNICODE + UNICODE BETA)

- -

[1] VisualText - executables. Version 2.0.1.X and later include both Unicode and - non-Unicode (or ANSI) executables.

- -

bin\VisualText.exe          Non-Unicode - ANSI version
-bin\VisualTextU.exe        Unicode - BETA version (only works with Unicode input files)

- -

With 2.1.0.0 and later, the following is also supplied:

- -

bin\VisualTextQ.exe -        Non-Unicode - ANSI version that uses QDBM database manager under the hood.  Note - that this is a beta version for growing arbitrarily large knowledge bases, - but is not yet optimized for speed.

- -

The ANSI version shares the same source code with the Unicode version, - but replicates the same basic functionality as earlier versions of VisualText.

- -

[2] Upgrading. -  As always, - care must be taken to upgrade or update existing analyzers and their knowledge - bases.  We - have added the topic Updating - Analyzers and KBs, which you should read and follow BEFORE - uninstalling an older VisualText and installing the latest.   Also, - analyzer project structure has changed slightly.  See - Structure - of Analyzer Projects.

- -

[3] Unicode - hints.  We've - added the topic Working with Unicode - to help here.

- -

Linux and Gnu C++. -  Analyzers - built with VisualText can be configured to run on Linux with Gnu C++. -  TAI periodically - upgrades the port to the VisualText runtime libraries.  The - libraries are not currently bundled with the VisualText release, and you - may inquire about them by contacting us at info@textanalysis.com .

- -

Miscellaneous

- -

(2.7.2.0) Rework of crawling functions eg urltofile, - to handle HTTPS as well as HTTP.
-(2.7.0.9) New function pnmove to enable moving - nodes within the parse tree.
-(2.7.0.0) Overhauls, optimizations, and bug fixes focusing on the recursive - NLP++ pass algorithm.
-(2.6.0.3) New function dejunk to clean out non-ASCII - and extended ASCII characters (sometimes replacing with a question mark - character).
-(2.6.0.0) Important fix to NLP++ pattern matcher re feature-based match - with wildcards.  dicttokz - pass, a variant of dicttok that removes whitespace from parse tree at - the outset, puts in attributes NOSP (no space before), NL (new line before). -  Adds capitalization - attributes to parse tree as well.  (This - will be an ongoing development in 2.6.x.x).
-(2.5.0.0) New dicttok - black-box tokenization pass that uses KB from TAIParse to put part-of-speech - tags on parse tree automatically. See DictTok_Sample - analyzer.
-(2.4.x.x) Maintenance releases, various fixes in feature-based matching, - compiled analyzers, etc.
-(2.3.2.0) In text windows ctrl-A now is Select All. ctrl-G is Find Again.
-(2.3.1.9) Maintenance release. Fixes to error reporting and recovery, compiled - vs interpreted analyzers.. group now returns the - new "reduce" node that it creates, an important convenience. -  group - also reports when it is including the lookahead - node -- an error.  listadd - no longer treats the list node as the "reduce" node.
-(2.3.1.8) xmlstr now converts characters to &#123; - where 123 is a character number, rather than to strings such as &frac14;.
-(2.3.1.6) Added special rule elements _xLET and _xCAPLET to conveniently match single-character - words.
-(2.3.1.0) VISUAL BASIC .NET API for - calling VisualText analyzers.  See - the sample program called VB_to_VT, which also - lists the API functions.
-(2.3.0.0) Compiling analyzers now generates separate code files for each - pass, enabling large analyzers to be compiled into C++.  For - example, the latest versions of TAIParse should now compile.  Note - that a Microsoft Visual Studio project may now need to have files added - to it, within an analyzer's RUN folder.
-(2.2.2.13) New PRE actions called regexp for matching - using regular expressions, with regexpi case - insensitive version.
-(2.2.2.11) New function for sorting a knowledge base concept's hierarchy - and children: sorthier and sortchilds.
-(2.2.2.8) Compiling analyzers once again generates PRETTY, or human-readable, - code.  We - tried very long lines to circumvent the 65K file line limit of some compilers, - but too-long lines are an issue as well.  In - subsequent versions, we'll be splitting files as needed.  For - now, huge analyzers may need their generated C++ code (in the RUN folder) - to be tweaked manually in order to compile.
-(2.2.2.5) EXITPASS, SUCCEED, and FAIL.  These - functions have been restored in that they are allowed in practically any - context without triggering error messages.  The - compiled analyzer version uses C++ exception handling mechanisms in order - to instantly handle these in deeply embedded contexts (function within - a function, etc.)  Parse - tree display ("yellow box") now places node variables in two - columns for convenient viewing of long variable lists. Parse tree nodes - now point back to rules that renamed them via xrename - and pnrename.  Compile-during-analyzer-load - popup of MS DOS window has been removed.  Interpreted - user functions improperly returned a non-zero value in the absence of - a return statement.
-(2.2.2.4) Incorrect VisualText TAB SIZE in Vista has been fixed. SAVE KB - and COMPILE KB default popup answer is now NO.  (Unless - you are explicitly trying to build a KB, then you don't want to save the - KB.)  Parse - tree for lone underscore character is now a terminal punctuation char. -  Listadd - in recursive passes when building a list right-to-left now fixed (compiled - version hang fixed). excise forbids deleting - a lookahead node.
-(2.2.2.2) The functions exitpass, succeed, - fail, can appear only at the top level of code - regions in compiled analyzers (FIXED. - See 2.2.2.5 above).  Fixes - to NLP++ syntax relating to negation.  A - fix to float conversion.  Further - fixes to optimized recursive pass algorithm.
-(2.2.2.1) Further fixes to align compiled and interpreted analyzers (fixes - to zeroed items in relational operators and others).  Fix - of an annoying VisualText crash when brought into focus from the background. -  A substantial - fix to the optimized recursive NLP++ pass algorithm in both the interpreted - and compiled analyzer frameworks.
-(2.2.1.1) New function pninsert for programmatically - inserting nonliteral nodes into the parse tree.
-(2.2.1.0) NEW. A parse tree view mode that reflects items removed from - the parse tree, such as HTML tags.  See - Tree Text Mode in the - View Menu.
-Fix to some NLP++ compiled functions to enable zero (0) as a null string. -  E.g., strsubst("ab,c",",",0); - will replace commas by empty string in a compiled analyzer.
-Note that this behavior already works as expected in interpreted analyzers, - e.g., those that run inside VisualText.
-(2.2.0.0) A substantially cleaned up VisualText release (see 2.1.0.9 notes - below).  Tree - Text view does not leave everything expanded on reentering VisualText, - but remembers what folders have been closed up in the view. The RUN NEXT - button behaves better.
-(2.1.0.9) A CLEANUP RELEASE.  ALIGNED - INTERPRETED AND COMPILED ANALYZERS, FIXED MISC BUGS AND LEAKS.  Fixed - a substantial bug that hobbled compiled analyzers.  Memory - leaks plugged in eltnode and lasteltnode.  The - PROG project or sample driver for analyzers now sets the BATCHSTART flag - at the start of a run, to enable a compiled analyzer to test the NLP++ - batchstart() function/flag. (So that an analyzer can perform initializations - when it runs the first file of a batch of files.)
-DEFAULT configuration of VisualText is now full screen, kb safe mode off, - toolbars in a standard configuration.  TIP - OF THE DAY offers real suggestions.
-Fixed bugs in the PROG driver of each analyzer, for the analysis of multiple - files and folders.
-(2.1.0.8) Fix to functions like resolveurl, strchr, strrchr. Fix to parse - tree nodes with large string-valued variables.  Fixes - to Tab Window and Find Window show/hide bugs.  Change - to VisualText naming in the Windows registry.  Fixed - a memory leak in rmattrs.
-(2.1.0.6) Fixes to file and directory handling.  (See - the PROG project of a newly created analyzer project.)
-(2.1.0.3) Misc fixes, including parse tree display and return of arrays - in NLP++ user-defined functions.
-(2.1.0.0) In order to support huge knowledge bases, a VisualTextQ version - is provided that utilizes the QDBM database management freeware under - the hood.  For - backward compatibility, one can treat the knowledge base built in this - way as the "compiled" knowledge base.  Loading - the interpreted knowledge base will read kb command files into a QDBM - file system.  "Compiling" - the kb will save the latest version of the QDBM files -- not compiling - the changes will discard the latest modifications to the kb.
-
-Other notes: Removed the Windows MS-DOS Command Prompt popups from most - system calls.
-
-(2.0.2.9) The levenshtein edit-distance computation - is provided.
-(2.0.2.8) Optimizations to string manipulation using L (local) variables. -  These are - not interned, as with S,X,G,N variables.  Locally - assigned strings are completely
-freed when a scope (function, @CODE, @CHECK) is exited, while globally - assigned strings are interned and persist throughout a text analysis. -  Analyzers - can
-be optimized by using local L variables wherever feasible.
-(2.0.2.7) fltval knowledge base value fetch.
-(2.0.2.5) Linux upgrade.
-(2.0.2.4) New variables G("$passnum") - and G("$rulenum") e.g. to track - rule context and performance.
-(2.0.2.3) New NLP++ function eltnode, analogous - to lasteltnode.
-(2.0.2.2) Recovery from very long tokens.
-(2.0.2.1) Refix of an infrequent crash in using $treetext.
-(2.0.2.0) Optimization to recursive passes.
-(2.0.1.1) Fixes to the Find in Files - popup menu.  Bugs - were introduced in VisualText 2.0: (a) missing patterns that start at - the beginning of a line, (b) incorrect display of capitalization in found - patterns.

- -

NEW IN CURRENT VERSION (2.0 NON-UNICODE BETA)

- -

VisualText 2.0 is the first to be compiled with Microsoft Visual Studio - .NET.  Within - a few releases, Unicode support will be featured as well.

- -

Professional version: While code generation is still automated, compiling - analyzers and KBs will need to be done manually for version 2.  Similarly, - user projects and other code will need to be upgraded to .NET and compiled, - in order to work properly.  See - the topic VisualText - Basics > Compiled and Standalone Analyzers for further guidelines. -  Similarly - for updating existing analyzers to version 2.

- -

A new feature is the deaccent element modifier, which allows alphabetics - to be matched without regard to accents on characters.  NOTE: - Within a rule element, the deaccent keyword must precede any match, fail, - and except lists, in order to work correctly.

- -

Miscellaneous

- -

(beta) The analyzer sequence can include folders, for organization and - readability.  Passes - can be dragged into the folders.

- -

Fix to NOOP rule matches in recursive pass.  Matching - continues after last node matched by such a rule.

- -

Better infinite loop tracking in recursive passes, to catch A -> - B and B -> A grammar loops ("2-state tracking").

- -

FEATURE-BASED PATTERN MATCHING.  The - functions below enable matching based on whether a node has a variable - with a particular value.

- -

var(var_name)    # - If var has nonzero value.
-varz(var_name)  # - If zero or no var.
-vareq(var_name,str)  # - If var equals str.
-vareq(var_name,num)  # - If var equals num.
-varne(var_name,str)  # - If var not equal str.
-varne(var_name,num)  # - If var not equal num.

- -

NEW IN CURRENT VERSION (1.8)

- -

NLP++ now features initial mathematical - functions.

- -

NEW IN CURRENT VERSION (1.7)

- -

The 1.7 release includes various enhancements to VisualText.  Attention - has been given to compiling and running analyzers within the Microsoft - .NET environment.  Also, - enhancements have been made to scalability: for example, the analysis - of each text now uses and frees up local tables, so that analyzers should - not grow as new texts are successively analyzed.  (Of - course, an analyzer may save things to the KB, in which case overall memory - size will increase.)

- -

The knowledge base files in the kb\user directory are now managed in - a new way, to better support scale up in the interpreted KB.  A - main.kb top-level file invokes the other command files in that directory - in order to load the interpreted knowledge base.  Also, - the attr.kb file has been replaced by multiple files named attr1.kb, attr2.kb, - and so on, depending on the size of the interpreted knowledge base.  Old-style - knowledge base directories will automatically be upgraded whenever a knowledge - base is saved.

- -

See Sample Analyzers > POStagger for - an example analyzer that integrates a 3rd party part-of-speech tagger.

- -

Important: See - the Analyzer Upgrade Note for 1.6, below, regarding upgrading existing - analyzers.

- -

Miscellaneous GUI

- -

GUI behaviors have been changed.  Questions - about compiling and saving the KB will be encountered as appropriate. -  Also, when - the analyzer sequence or properties are changed, a popup will ask to save - those changes.

- -

The GUI no longer uses the gui.kb file to record the analyzer sequence - and the input file hierarchy.  Rather, - the input file hierarchy is obtained from the file system, and the analyzer - sequence definition is fetched from the analyzer's spec\analyzer.seq human-readable - file.  The - GUI still uses the KB to store this information on a temporary basis when - the analyzer is loaded.

- -

Some log files accessible in the Text Tab are now tracked by the Page - Mode mechanism.  That - is, their display will automatically update in tandem with the input file - selected in the Text Tab.

- -

Timing data for each pass of an analyzer is available in the dbg.log - file.

- -

The Finder Window now presents matches using the order of files in the - current analyzer sequence.

- -

Browser and input text type-in windows can now accept up to 1024 characters.

- -

Some stale or broken pointers to VisualText Help topics have been fixed.

- -

By default, new analyzers have the Verbose and Logs Toggle turned off.

- -

Miscellaneous NLP++

- -

Scalability enhancement: Separate information tables are built and removed - for the analysis of each input text.  This - enables analyzers to run large numbers of texts without slowing down and - without consuming extra memory (depending on whether the analyzer updates - the KB).

- -

Incompatible changes to API header files, User project, and standalone - sample code.  Existing - analyzers must be upgraded and recompiled.  Standalone - code upgraded to recursively process directories of input files.  user::loadanalyzer - fixed to create globally managed knowledge base.  NLP, - CG, and VTRUN objects created and deleted only within the runtime libraries - via API calls in VisualText\include\Api\lite\vtrun.h.

- -

Storage tables in NLP++ and the Conceptual Grammar KBMS have been enlarged - and their performance has been enhanced.  The - differences can be substantial for larger input files.

- -

Some automatic infinite loop monitoring is performed in passes with - the recursive - algorithm type set.

- -

A bug in the fail - plus except - keyword logic has been fixed, within rule element modifiers.

- -

Missing KB functions for floating point data type have been installed, - as well as some missing compiled variants of NLP++ functions.

- -

Some table sizes have been increased, e.g., for buffers returned by - an analyzer called with user::analyzefile - and user::analyzebuf.

- -

Enhancements to analyzer error statements.  More - warnings are issued if an NLP++ function is called with too many arguments. -  Correct - pass and line numbers have been added to many error messages.

- -

The dictgetword function finds a dictionary - word-concept, else makes a new one.  A - renaming of addword.

- -

The take - command now handles relative paths.

- -

The sqlstr function converts strings to a format - for SQL entry into a database.  (Merely - converts single quote to two single quotes.)

- -

Bug fixes in compiled versions, string catenation and string functions, - both in scaled up string sizes and 0 as a string.  Removed - some null string warnings.

- -

A compiled vs interpreted issue is that C/C++ conditional evaluation - works in compiled but not interpreted.  So - in expressions like A && B, assume that both the A and the B sides - will always be evaluated.

- -

(1.7.4) New analyzers now come with a prebuilt Microsoft Visual C++ - user project, including the files prog.dsw - and prog.dsp.  For - professional version users, these enable convenient setup of a standalone - sample analyzer.  See - the Help topic VisualText - Basics > Compiled and Standalone Analyzers.

- -

(1.7.4.5)

- -

Fixed _xSTART - rule element modifier within a match list.

- -

Fixed rare race condition in opting to not save knowledge base when - exiting VisualText.

- -

Fixed handling of inactive passes and changing active/inactive pass - property.

- -

Fixed split function to handle large string, - added compiled variant of function.

- -

Better handling of "composite" literals formed by rules such - as:  redo - <- re \- do @@

- -

Fixed returns of some string processing functions.  Fixed - crash on catenating null strings.

- -

New function: deaccent replaces accented - characters in strings with unaccented characters.

- -

Improved error checking in the @DECL region.

- -

Fixed parse tree display for analyzing binary files.

- -

GUI convenience: When logs turned off, clicking passes in Ana Tab will - not redraw the parse tree.  Nicer - to work with large input files now.

- -

Fixed display of parse tree view for binary input files.

- -

(1.7.4.6)

- -

Fixed parse tree display of variables in yellow node popup when Node - Variables toggle is off.

- -

(1.7.5)

- -

Added a new sample analyzer called POStagger (in VisualText\apps), which - invokes an external part-of-speech tagger.  This - analyzer shows how one can extend VisualText with 3rd party software. -  See the - Sample Analyzers > POStagger description - for more details.

- -

Implemented the setlookahead action, - allowing dynamic setting of the lookahead node after a rule match.

- -

(1.7.6)

- -

Housekeeping: Further alignment of compiled and interpreted. Fixed compiled-version - of splice action and _xSTART in a match list.  Fixed - Save KB crashes due to long concept names.  Fixed - infinite loop on some optional rules with lookahead.

- -

Added error messages for fully optional rules, range errors in NLP++ - actions.

- -

(1.7.7)

- -

Fixed crashes with catenating empty strings.

- -

Some fixes to the third party POS tagging handshake, handles paths with - space characters.

- -

The permuten function is useful for permuting - lists of values.

- -

NEW IN VERSION (1.6)

- -

Analyzer Upgrade Note: -  Analyzers - built with prior versions of VisualText will not load properly with the - 1.6 version.  To - upgrade existing analyzers, you'll want to compare and merge your existing - analyzer project files with those within the VisualText\Templates\Bare - analyzer template.  (If - an analyzer was built using another template, best to compare with that - one.)  After - merge, you'll want to recompile the analyzer, knowledge base, user projects, - and stand-alone driver program, if any.  Important - places to look for mergeable files are:

- - - -

Warning: - Prior to any merge, we strongly recommend (1) making a safe copy of all - your existing analyzers; (2) doing Save KB for analyzers that have a compiled - knowledge base, in order to save the interpreted form of the KB; (3) making - a comparison with each existing analyzer and the Bare template (or other - template that you started with), using a tool such as WinDiff that comes - with Microsoft Visual C++.

- -

 

- -

Simple - merge: If you have only modified the analyzer sequence and pass - files, then the merge is simple.  Copy - everything in VisualText\Templates\Bare EXCEPT - the spec folder, kb - folder and .ana file, if any, - to your existing analyzer project.  Then - copy the top-level files in the Bare kb - folder, EXCLUDING - the user sub-folder, to your existing - analyzer project.

- -

External and Multiple Analyzers

- -

The NLP++ runtime system now enables analyzers to invoke other analyzers - and allows multiple analyzers and knowledge bases to cleanly operate within - the same executable program.

- -

In the User project of an analyzer, the sample function User\Ucode.cpp\useranalyzefile - may be called from an analyzer, e.g.,

- -

G("buffer for result from analyzer") - = user::analyzefile("analyzername","c:\\forexample\\inputfile.txt", - "data to pass to analyzer via $datum variable", G("ostream"));

- -

to invoke another analyzer (assumed to be - in the area defined by the APPS environment variable). The called analyzer - may grab the data from the 3rd argument via the special variable G("$datum"). -  Writing - to the cbuf() stream enables the called analyzer to return data to the - calling analyzer.

- -

Another function, user::useranalyzebuf(), - illustrates calling a second analyzer, as follows

- -

G("result") = user::analyzebuf("analyzername", - "input buffer...", "info for $datum special variable", - G("ostream"))

- -

so that a buffer can be processed by a called - analyzer, which returns a buffer result to the calling analyzer.

- -

The VisualText programmer's API has been enhanced - to enable multiple analyzers to be loaded into memory once, while invoking - each other multiple times.  Creating - an NLP() object now takes an optional name string, so that analyzers may - register themselves in memory and call each other by name.  The - function findana will help load once and run - many times within VisualText itself.  (This - capability is still under development.)

- -

In order to enable load one and run many times - for an analyzer, we also supply the function user::loadanalyzer(analyzername_str). -  Thus one - can set up to call a second analyzer as follows:

- -

G("analyzer handle") = user::loadanalyzer("myanalyzer");

- -

G("result") = user:analyzefile(G("analyzer - handle"), G("input file"), G("stuff to pass down to - $datum variable"), G("ostream"));

- -

The second statement could be placed in a - loop in order to analyze a set of input files, thereby saving on analyzer - and knowledge base load time.

- -

Note that the first argument to analyzefile - and analyzebuf can be the analyzer name or the "handle", which - currently are equivalent.  In - anticipation of VisualText allowing multiple instances of the same analyzer, - we recommend using the "handle" returned by loadanalyzer.

- -

New: analyzefile and analyzebuf now also take - an ostream argument (which can be 0), enabling the calling analyzer to - specify an output stream for the called analyzer.

- -

Caveats: - Recursive calls to the same analyzer are not supported in the current - version.  They - may clobber some working areas such as the knowledge base and the output - folder of an analyzer project.  Infinite - calling loops must be guarded against (e.g., by passing down a counter). -

- -

Compiling Knowledge Base

- -

If automated compilation of an analyzer's knowledge base is failing, - we recommend defining the user environment variable (in Start Button > - Control Panels > System > Advanced > Environment Variables):

- -

VCBIN = C:\Program Files\Microsoft Visual Studio\VC98\bin

- -

This assures that the correct linker (named LINK.EXE) will be invoked - by the makefile that compiles an analyzer's knowledge base.  (If - your computer has several development programs on it, such as Visual Basic - or Symantec Visual Cafe, then the wrong program named LINK.EXE may be - found inadvertently by the system when linking C++ programs.  The - above environment variable is used by VisualText to eliminate this conflict.)

- -

Miscellaneous

- -

HOURGLASS CURSOR BUG - On some - machines, especially new or fast machines, analyzing a file in VisualText - may have led to a hung VisualText session.  This - bug, due to a race condition between two threads, has been fixed.

- -

Runtime Library Upgrades - Knowledge base and runtime libraries have - been upgraded to remove global variables and install object-oriented programming - constructs.

- -

Compilation of larger analyzers has been enhanced by changes to the - C++ code generated from pass files.  Some - internal table sizes also have been enlarged.

- -

Global string and hash tables used by the analyzer have been enlarged.

- -

For display purposes, the new menu item Edit > Change Fonts... allows - a user to bring up a standard font selection popup window.  In - this way, text files may be displayed using fonts loaded in the computer -  (e.g., - Japanese fonts).

- -

Some important fixes to problems with multi-valued NLP++ variables, - e.g., going into and out of user-defined functions.

- -

Deployment - The /silent command line option for standalone analyzers - now eliminates extraneous outputs to the console.

- -

Some fixes have been made to the error messages printed in the VisualText - log window.  This - overhaul is still in progress, and so some error messages may still not - point to the file and line of code that led to a syntax or runtime error.

- -

The exitpass function has been upgraded to - work in the @POST region.  It - also works within if-else statements, lists of statements, and while-statements. -  It should - not be used within user-defined NLP++ functions, however.

- -

Compiled variants of NLP++ code have been fixed, including exitpass - and assignment statements (where an OSTREAM is being assigned to a variable). -  rightjustifynum - was fixed to handle null values identically in compiled and interpreted - variants.

- -

L local variables - may now be used in the @CODE, @POST, and @CHECK regions.  They - go out of scope when execution of code in a region is completed.

- -

A bug has been fixed in matching rules that contained a triggered element - along with wildcard elements to its left.  (Appeared - only in compiled analyzers.)

- -

When a browser window with a local HTML file was left open on exiting - VisualText, the file could not be found when reloading the analyzer in - VisualText.  This - bug has been fixed.

- -

The new function nodeowner has been implemented, - usurping the functionality of of nodeconcept. -  The former - fetches the concept that owns the current phrase, while the latter fetches - the concept that the given node is a proxy for.

- -

The new function sortconsbyattr sorts - an array of KB concepts by comparing one of their attributes.

- -

The new function inputrange conveniently - fetches substrings of the input text. For efficient printing of such ranges, - inputrangetofile is provided.

- -

The new function strchrcount counts the - occurrences of a character in a string.

- -

The new functions dictfirst and dictnext - enable convenient traversal of the words in the KB dictionary hierarchy.

- -

The new Perl-like function split enables splitting - a string into a multi-valued variable (i.e., an NLP++ array) using a given - character as a separator.

- -

New functions in support of web page processing: urltofile - fetches a web page as a file and resolveurl - converts a relative URL to an absolute URL.  In - conjunction with analyzers calling analyzers, these functions support - web navigation (crawling, spidering, searching) within VisualText.  urltofile - now has reasonable timeouts set for fetching a page from the web (but - user cannot modify the timeout setting).

- -

The new function xmlstr converts characters - in a string appropriately for outputting XML or HTML files (e.g., converts - "hello&bye" to "hello&amp;bye").

- -

Fix to the splice NLP++ action.  Developers - should avoid addressing rule elements after actions such as splice - and merge, because numbering - is element numbering is unpredictable after such actions.

- -

Variants of the NLP++ strpiece function have - been fixed.

- -

Fixes to NLP++ expressions that compare non-numeric types with 0.  E.g. - "G("str") == 0".  String - catenation with one empty argument has been fixed (e.g., "hello" - + 0).

- -

Assignments involving OSTREAM data types have been fixed in the compiled - version.

- -

Many error messages have been fixed, but due to overhauls, the pass - and line number are not always available to an error message.

- -

Compiled rules with a trigger element and wildcards to its left have - been fixed.  Compiled - versions of Boolean testing for knowledge base data types, e.g., in an - if-statement, have been fixed.

- -

Local files in the Browser window now display properly when loading - an analyzer.

- -

Some enhancements have been made to "retokenization".  For - example, one can write rules like

- -

oversight <- over sight @@

- -

in which the suggested concept is a literal, and have the new "oversight" - node in the parse tree matched by _xALPHA.  Getting - the text from N("$treeraw") will also work (though not from - N("$text")).

- -

In some 1.6 versions, cout() does not work.  This - has been fixed.

- -

NEW IN VERSION (1.5)

- -

Version 1.5 is primarily concerned with housekeeping.

- -

Changes to the VisualText user interface behaviors make it more convenient - to work with analyzers.  When - you click the Run button, the cursor immediately changes to an hourglass. -  In viewing - parse trees regardless of settings, the yellow popup box for a node always - displays its variables and values.  When - Logs are turned off, the parse tree always displays the final parse tree. -  When commenting - text in the Editor, only highlighted lines get commented (fixing the "extra - line commented" feature).

- -

The runtime system for compiled analyzers has received major treatment. -  For example, - the compiled and interpreted versions of the latest TAIParse can have - their intermediate log files and output files compared on a file-by-file - basis using a program such as WinDiff.  Because - interpreted/compiled analyzers are now well-aligned, it is easy to find - any differences between the versions, as a debugging aid.  Missing - compiled variants of NLP++ functions have been added and debugged.

- -

NEW IN VERSION (1.4)

- -

Database Connectivity

- -

NLP++ now features database connectivity functions, more or less mirroring - the standard set of ODBC functions.  All - database types other than BLOBS are handled.  Values - returned from a database are mapped to string, numeric, and float NLP++ - types, as appropriate.  See - Database Functions.

- -

A sample analyzer illustrating connection to a database and other new - features is in Samples\database1. -  See also - Database1 Analyzer.

- -

Attribute Editor Enhancements

- -

The Attribute - Editor has been substantially overhauled.  Management - of the knowledge base value types (string, concept, and numeric) is more - coherent.  Methods - have been introduced for specifying and viewing a concept as the value - of an attribute.  Up - and down navigation arrows have been added, enabling convenient perusal - of the knowledge base while staying within the Attribute Editor.

- -

Miscellaneous

- -

An analyzer may now cause the creation of a popup window within VisualText, - with the new functions exittopopup() - and getpopupdata().

- -

The built-in database function dbbindcol() - is the first to employ call-by-reference within NLP++.  User-defined - functions may not use call-by-reference at this time.

- -

A silent flag in standalone - analyzer API calls minimizes the amount of debugging and logging output - produced during the operation of a text analyzer.

- -

A function for gathering command line arguments has been placed within - the sample standalone code, e.g., in Bare\main.cpp.

- -

NEW IN VERSION 1.3.1

- -

Miscellaneous

- -

The VisualText GUI now highlights contiguous text regions with alternating - dark and light colors.  That - is, dark blue and light blue alternate to display new nodes built in a - pass, while dark and light green alternate to show nodes matched in a - pass.

- -

The conversion functions num() and - str() now handle floating point.

- -

A call-by-reference capability has been implemented for NLP++ builtin - functions.  Not - yet available to user-defined NLP++ functions.

- -

ind wattr, a command for more - conveniently adding knowledge to the kb via command files.  See - About KB Command - Files.

- -

Fixed an overflow buf in spelling functions.

- -

NEW IN VERSION 1.3

- -

Output to User-Supplied Stream and Buffer

- -

In standalone/embedded analyzers, users may now provide an output stream - and an output buffer to the analyzer.  Output - to these is enabled by the new NLP++ functions cout() - and cbuf().  coutreset() - enables rebinding the output stream to a file, while interactive() - checks to see if the analyzer is running in an interactive environment, - e.g., within VisualText.  Further - documentation is available in the topic Compiled - and Standalone Analyzers.

- -

NEW IN VERSION 1.2

- -

Enhanced User-Defined NLP++ Functions

- -

In version 1.1, we introduced the @DECL region and user-defined function - definitions within NLP++.  With - Version 1.2, support for NLP++ functions is much improved, including parameter - lists, arguments, return statements, recursive function calls, and compilation - of all the above.  To - implement parameters, we introduce the local variable reference L, similar to the S, X, - N, and G - variable references that are key to NLP++.  L - is used not only for parameters, but can define variables local to the - function body.

- -

NEW IN VERSION 1.1.2

- -

Event-Based Functions in User Project

- -

The functions ucodeIni() and - ucodeFin() are available in the - User project of an analyzer.  The - analyzer developer may place initializations in ucodeIni that operate - immediately after the analyzer is loaded.  Analogously, - cleanups may be placed in ucodeFin, which operates immediately before - the analyzer is closed.

- -

Improved Checking on Overly Large Input Texts

- -

Enhancements are in progress that check against an overly large input - text file.  More - generally, when the analyzer has expended all available space resources, - then the system will degrade more gracefully rather than crashing.  It - is recommended that an input file of about 1MB maximum be run through - a typical text analyzer.

- -

As always, huge texts will need to be split up into buffers or multiple - files before presentation to a text analyzer, in general.  The - VisualText knowledge base can serve to manage information across multiple - segments of a huge document.

- -

 

- -

NEW IN VERSION 1.1

- -

User-Defined NLP++ Functions

- -

 VisualText - 1.1 introduces user-defined NLP++ functions.  Up - to now, users could only add functions via the C++ user - project associated with each analyzer project, or by requesting - that a function be added by TAI to the NLP++ builtin functions.

- -

The current version is a pre-release of the user-defined function capability. -  No return - statements and no parameter lists can be written in functions yet.  Nor - are there templates that define the inputs and outputs to a function, - nor is overloading of function names allowed.  Even - so, the new capability substantially enhances the power of NLP++ for complex - tasks.  Rather - than repeating the same code in many places, functions enable the code - to be written in a single centralized location.

- -

Functions are placed in the new @DECL region, which must be the first - region in a pass file, when present.  Functions - may be called from passes that precede the function definition (i.e., - forward reference).   A - sample is

- -

@DECL

- -

# User-defined function definitions.

- -

myfunction()
-{
-"output.txt" << "Entering myfunction." << - "\n";
-G("tmp") = 3;
-}

- -

@CODE
-myfunction();   # - A call to user-defined function.

- -

An interesting feature of NLP++ functions is that their operation can - optionally depend on the region they are called from.  For - example, in functions called from a @POST region, the user may use the - S, N, - and X variable - references as though the function's code were placed directly within the - region.

- -

Double-Click Parse Tree Node

- -

A new backtrace and debugging capability is the following.  In - a Parse Tree window, double-click on a node, and the pass file that created - that node will open in the Workspace.  Further, - the cursor will be placed at the start of the precise rule that created - the node.

- - - - diff --git a/Help/helps/PATH.htm b/Help/helps/PATH.htm deleted file mode 100644 index 45e160d..0000000 --- a/Help/helps/PATH.htm +++ /dev/null @@ -1,67 +0,0 @@ - - - - - - - - - - -@PATH Selector - - - - - - - - - - - - - - - - - - - - - - - - - - - -

PATH Selector

- -

The PATH Selector lets you specify a path - of node names that must start with _ROOT, the root of the parse tree. - The path specifies the root, its child, grandchild, and so on. The selection - algorithm traverses the parse tree starting at the root and searches for - a context node that matches the entire path. The phrase immediately under - such a context node will be traversed to look for rule matches. (PATH - does not check the unsealed flag.)

- -

See Also

- -

Context - Zone

- -

SELECT - Region

- -

MULTI - Selector

- -

NODES - Selector

- -

 

- - - - diff --git a/Help/helps/POST.htm b/Help/helps/POST.htm deleted file mode 100644 index c0f5fd8..0000000 --- a/Help/helps/POST.htm +++ /dev/null @@ -1,64 +0,0 @@ - - - - - - - - - - -@POST Region - - - - - - - - - - - - - - - - - - - - - - - - - - - -

POST Region

- -

NLP++ code in the POST Region executes - after - the rule matcher has successfully matched a rule and the match is accepted. - Element actions are also executed prior to the POST Region.

- -

See Also

- -

Grammar - Region

- -

POST - Action

- -

PRE Region

- -

CHECK - Region

- -

RULES - Region

- - - - diff --git a/Help/helps/POStagger.htm b/Help/helps/POStagger.htm deleted file mode 100644 index 5c2a48c..0000000 --- a/Help/helps/POStagger.htm +++ /dev/null @@ -1,92 +0,0 @@ - - - - - - - - - - -POStagger - - - - - - - - - - - - - - - - - - - - - - - - - - - -

POStagger

- -

NOTE: - We are no longer bundling and distributing this third party tagger, as - its licensing is incompatible with the commercial uses of VisualText.

- -

Our most recent part-of-speech tagger that keys off our TAIParse general - analyzer for English is described at TAIParse.

- -

The POStagger analyzer project can be found in VisualText\apps.  As - an example, it invokes the LexTagger part-of-speech tagger from UMIST, - which is available in VisualText\apps\LexTagger-1.0 and also by contacting - the author, Argyris Vasilakopoulos, for updates and documentation at

- -

a.vasilakopoulos@postgrad.umist.ac.uk

- -

The POStagger analyzer consists of only two passes.  The - first is the default tokenize pass that initiates all VisualText analyzers. -  The second - pass looks like

- -

@CODE
-user::postagger();
-@@CODE

- -

which invokes the LexTagger software with the current input file, then - parses the result and layers it into the analyzer's current parse tree. -  The handshake - code can be found in POStagger\User, within the Ucode.cpp and Umisc.cpp - files.

- -

In order to use the analyzer, you'll need to edit the run.bat - and the Properties.txt files within - the LexTagger-1.0 folder, in order to fix the pathnames for your machine. -  You'll - also need to edit the paths in the User project to point to your copy - of LexTagger.

- -

Note that the current handshake code in the User project copies the - input file to the LexTagger-1.0\test folder.  A - more elegant handshake would write out the Properties.txt file for each - input, so that large input files won't need to be copied.

- -

Other part-of-speech taggers could easily be swapped in for LexTagger - with some editing of the example handshake in the User project.

- -

 

- -

 

- - - - diff --git a/Help/helps/PRE.htm b/Help/helps/PRE.htm deleted file mode 100644 index 6c04f29..0000000 --- a/Help/helps/PRE.htm +++ /dev/null @@ -1,91 +0,0 @@ - - - - - - - - - - -@PRE Region - - - - - - - - - - - - - - - - - - - - - - - - - - - -

PRE Region

- -

The PRE Region is used to extend the pattern matcher's mechanism for - matching each rule element.   Currently, - the PRE Region does not support general NLP++ expressions and statements.

- -

PRE Action Syntax

- -

The actions used in the PRE Region have a mandatory range. Their form - is

- -

<num,num> - pre_action_name(args....)

- -

The PRE Action applies to each rule - element in the <num, num> range individually.

- -

PRE Actions operate only on tokens, - or leaf nodes of the parse tree, at present. So, if the current node covers - more than a single leaf node, the PRE Action automatically - fails. Another thing to be careful of is that - PRE Actions disregard the base attribute of a node. That is, even if the - current node or a subnode has the base attribute set, the PRE Action will look - under it for the leaf node.

- -

See Also

- -

Grammar - Region

- -

PRE - Action

- -

CHECK - Region

- -

POST - Region

- -

RULES - Region

- -

 

- -

 

- - - - diff --git a/Help/helps/PROG_Analyzer.htm b/Help/helps/PROG_Analyzer.htm deleted file mode 100644 index 0e0faeb..0000000 --- a/Help/helps/PROG_Analyzer.htm +++ /dev/null @@ -1,52 +0,0 @@ - - - - - - - - - - -PROG_Analyzer - - - - - - - - - - - - - - - - - - - - - - - - - - - -

PROG_C++_Analyzer

- -

Every analyzer project comes with an MS Visual Studio project called - PROG in the top-level folder.  This - shows how to call an analyzer from VisualText.

- -

See Compiled - and Standalone Analyzers for more details.

- -

 

- - - - diff --git a/Help/helps/Parse_Tree.htm b/Help/helps/Parse_Tree.htm deleted file mode 100644 index 7d711a2..0000000 --- a/Help/helps/Parse_Tree.htm +++ /dev/null @@ -1,63 +0,0 @@ - - - - - - - - - - -Parse Tree - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Parse Tree

- -

The Parse Tree display is a tool for viewing the results of pattern - matching by an analyzer.

- -

NEW!  Double-clicking - on a node now brings up the pass file and rule that created that node. -  This is - a substantial debugging aid, and also empowers a faster mode of development - in which the Logs Toggle is turned off (so that analysis of a text is - faster).

- -

Clicking on the "+" in front of a node displays its subtree. -  Conversely, - clicking on the "-" in front of a node hides its subtree.

- -

See Also

- -

Parse Tree Menu, - Parse Tree - Popup

- - - - diff --git a/Help/helps/Possibly_JUNK/Basic_steps_for_creating_a_text_analyzer.htm b/Help/helps/Possibly_JUNK/Basic_steps_for_creating_a_text_analyzer.htm deleted file mode 100644 index ba4f3d9..0000000 --- a/Help/helps/Possibly_JUNK/Basic_steps_for_creating_a_text_analyzer.htm +++ /dev/null @@ -1,272 +0,0 @@ - - - - - - - - - - -Basic steps for creating a text analyzer - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Basic steps for creating a text analyzer

- -

This topic outlines the basic steps for creating and running a text - analyzer using VisualText. Since VisualText can be used to create many - different types of text analyzers, we will give a rough outline of the - overall process of creating, running and defining a text analyzer.

- -

Creating a new analyzer project:

- -
    - -

  1. From the File - menu, select New Analyzer.

    2. - -

  2. In the New Analyzer - dialog, provide the following information:

    3. -
- - - -
    - -

  1. Click OK.

    2. -
- -

Creating text files to be analyzed:

- -
    - -

  1. In the Text - Tab, right click and select Add - > New Text File.

    2. - -

  2. In the dialog box, give the text file a name.

    3. - -

  3. Click OK. The newly created text file opens up in - the Workspace. (The text file is stored in the input folder of the analyzer project. You can - also add files to be analyzed directly into the input folder.)

    4. -
- -

Adding and saving text to a text file:

- -
    - -

  1. Double click - on the a file in the Text Tab to open it in the Workspace.

    2. - -

  2. Add text to - the file.

    3. - -

  3. Save the file - by right-clicking on the file and selecting Save. You can also use the save icon on the Main - Toolbar.

    4. -
- -

Defining the analyzer sequence:

- -

The analyzer sequence, made up of - passes, - is viewed and managed in the Ana - Tab. A pass is a discrete step - in the analyzer sequence that has its own pass algorithm. Each analyzer - project comes with a built-in pass in the analyzer sequence. The default - pass tokenize - tokenizes each of the characters in an input file into token types, such - as alpha - for alphabetic, num for numeric, punct for punctuation, etc. As you build your - analyzer, you add passes to the analyzer sequence after the tokenize pass. - Each pass in the analyzer sequence is associated with a pass - file that contains NLP++ code - and rules which instruct the rule matcher what to do. Here is how you - add passes to the analyzer sequence:

- -
    - -

  1. Select the - pass in the analyzer sequence where you want to add a pass.

    2. - -

  2. Right-click - and select, Add - > New Pass. - (The pass is inserted AFTER the pass you selected.)

    3. - -

  3. In the Pass - Properties dialog, provide the following:

    4. -
- - - -
    - -

  1. Click OK.

    2. -
- -

The new pass is created in the analyzer - sequence. To edit the pass, double click on the pass to open the associated - pass file in the Workspace. For a description on editing pass files, see - the Developer's section.

- -

Running an analyzer:

- -
    - -

  1. Select a file - in the Text - Tab.

    2. - -

  2. From the Analyzer - Menu, select Run. - You can also use the run icon from the Workspace toolbar. A red checkmark on the file icon indicates that the file - has been analyzed.

    3. -
- -

When an analyzer is run, a parse - tree is created for the analyzed text. How you define your analyzer sequence - and pass files will determine what kind of information will be displayed - in the parse tree.

- -

Viewing the parse tree of an analyzed - text

- -
    - -

  1. Select the - analyzed file in the Text Tab.

    2. - -

  2. Click on the - Parse Tree button. You can also open the file in the Workspace, - right-click on the file and select View > Parse - Tree.

    3. -
- -

Saving the Analyzer:

- -
    - -

  1. Select Save Analyzer - from the File menu. Save Analyzer - saves the analyzer project and the associated Knowledge Base. (You should - save your project often.)

    2. - -

  2. Answer Yes - to the prompt asking if you want to save the analyzer.

    3. -
- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/Possibly_JUNK/NLP++.htm b/Help/helps/Possibly_JUNK/NLP++.htm deleted file mode 100644 index 18a086d..0000000 --- a/Help/helps/Possibly_JUNK/NLP++.htm +++ /dev/null @@ -1,3674 +0,0 @@ - - - - - - - - - - -NLP++ - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

NLP++

- -

[Note: This topic is slated for removal. Refer also - to the NLP++ section.]

- -

NLP++� - is a novel programming language developed by Text Analysis International, - Inc. It enables a developer to focus on the conceptual task of building - a text analyzer while minimizing the need to delve into compiled programming - languages. NLP++ combines a C++ -like syntax (called the code - part) with a rule syntax (described in the Pass - Files topic). The current topic is concerned primarily with the code - part.

- -

One use of NLP++ is dynamic - semantics, that is, to create and use information "on the - fly". Another is to embed heuristics and algorithms into the text - analyzer under construction.

- -

Regions

- -

The NLP++ language can currently appear in the @CHECK, @CODE and @POST - regions of a pass file. In a @CHECK region, the functions succeed - and fail can be used to accept - or reject a rule match, respectively.

- -

The @PRE region, which operates during the pattern matching of a rule, - does not yet allow NLP++ constructs.

- -

WARNINGS:
-[1] The developer must keep in mind that actions that modify the parse - tree, such as single(), will disable references to the nodes of a phrase - and otherwise affect the execution of NLP++ code. It is recommended that - tree-modifying actions follow references to node variables.
-[2] If the developer must reference a newly created node, e.g., by the - group - action, he must manually figure out the new element number of the created - node. Note that following an action such as group, element numbers to the right of the grouped - node will also change.

- -

Tokens

- -

See the Pass File section on Tokens.

- -

Variables

- -

NLP++ currently supports global variables and variables associated with - particular nodes of the parse tree. The syntax for these variables is - as follows:

- - - - - - - - - - - - - - - - - - - - - - - -
-

VARIABLE

 -

DESCRIPTION

-

G(varname)

 -

Global variable.

-

S(varname)

 -

Variable belonging to the suggested concept - of a rule.

-

X(varname - [,num])

 -

Variable belonging to the numth - context node starting at the root of the parse tree. Usually used to refer - to the numth - node of the @PATH select list. - With @NODES, the preferred form - is X(varname).

-

N(varname - [,num])

 -

Variable belonging to a node that matched - the numth element of a rule phrase.

- -

The shorthand notations N(varname) - and X(varname) refer to the last - element of the phrase and the last context node, respectively. The former - is convenient for rules that have only one element in their phrase. The - latter is convenient when using the @NODES - selector. N(varname, 0) and X(varname, 0) have the identical meaning.

- -

There is no notion of a local variable or pointer, at this time. S variables are often used as an idiom - for local variables (the difference is that they "hang around" - if a new node is suggested by the current rule).

- -

NOTE: Variable names must be quoted. E.g., N("unknown - word", 2)

- -

WARNING: X(varname,num) does NOT refer to the numth member of a @NODES list. That list is - not a path in the parse tree. We recommend using X(varname) with @NODES, to avoid "blindly" - traversing down the parse tree. Using X(varname,2), for example, implies - that you know what the nodes immediately below the root are. Such a use - should be clearly documented.

- -

Special Variable Names

- -

To get useful information from the parse tree and other sources when - analyzing a text, some special variable names have been provided. All - start with a dollar sign ($). - No such variable names are usable with suggested (S) - variables, at present.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

VARIABLE NAME

 -

VARIABLE TYPES

 -

DESCRIPTION

-

$treetext

 -

N, X

 -

Fetch the 'cleaned up' (removing leading and - trailing, convert separators to single space) text gleaned from a node's - subtree. Uses the parse tree as input, not text buffer. Analogous to $text. - Difference between $treetext and $text is $text returns text from the - input text buffer while $treetext returns text from the parse tree. So, - if you have removed some nodes from a tree with an excise(), $treetext - returns the text minus the excised elements, while $text returns everything - in the input text buffer, including the text under excised nodes.

-

$text

 -

N, X

 -

Fetch the text covered by the node. Cleanup - whitespace (removing leading and trailing, convert separators to single - space). Uses the original text buffer, rather than the subtree under the - node, in order to gather text. See also $treetext and $treeraw.

-

$treeraw

 -

N, X

 -

Fetch the text covered by the node from the - parse tree. Text is not 'cleaned up'. Analogous to $treetext. See also - $treetext and $raw.

-

$raw

 -

N, X

 -

Fetch the 'cleaned up' text covered by the - node from the text buffer. See also $treetext, $treeraw and $text.

-

$xmltext

 -

N, X

 -

Same as $raw, but converts characters - that are special to HTML and XML. E.g., "<" is converted - to "&lt;".

-

$length

 -

N, - X

 -

Get the length of node's - text.

-

$ostart

 -

N, X

 -

Start offset of the referenced node in the - input text.

-

$oend

 -

N, X

 -

End offset of the referenced node in the input - text.

-

$start

 -

N, X

 -

Evaluates to 1 - if the referenced node has no left sibling in the parse tree, else - 0.

-

$end

 -

N, X

 -

Evaluates to 1 if - the referenced node has no right sibling in the parse tree, else 0.

-

$apppath

 -

G

 -

Get base directory path for current application. - E.g., "D:\dev\apps\Resume".

-

$input

 -

G

 -

Get full input file. E.g., "D:\apps\Resume\input\Dev1\rez.txt"

-

$inputpath

 -

G

 -

Get input file path. E.g., "D:\apps\Resume\input\Dev1"

-

$inputname

 -

G

 -

Get input filename. E.g., "rez.txt"

-

$inputhead

 -

G

 -

Get input file head. E.g., "rez"

-

$inputtail

 -

G

 -

Get input file tail (i.e., extension). E.g., - "txt"

-

$allcaps
-$uppercase

 -

N

 -

Returns 1 if the token underlying the node - is all uppercase. If multiple words (even if all are all-caps), returns - 0. WARN: Not identical to the uppercase() pre action.

-

$lowercase

 -

N

 -

Returns 1 if the token underlying the node - is all lowercase If multiple words (even if all are all-caps), returns - 0. WARN: Not identical to the lowercase() pre action.

-

$cap

 -

N

 -

Returns 1 if the token underlying the node - is a capitalized word. WARN: Not identical to the cap() pre action.

-

$mixcap

 -

N

 -

Returns 1 if the token underlying the node - is a mixed-capitalized word. Examples: "MIchigan", "abcD".

-

$exists

 -

N

 -

Returns 1 if the node exists, 0 otherwise.

-

$unknown

 -

N

 -

Returns 1 if the token underlying the node - is an unknown word. WARN: Not identical to the unknown() pre action. Requires - a lookup() code action prior to - any use of this special variable.

- -

NOTE:
-[1] Special variable names must be quoted. E.g., N("$text", 2).
-[2] You should not assign to or create your own "dollar" variables. - NLP++ reserves all variable names that start with a dollar sign.
-[3] The $uppercase, $lowercase, $cap, and $unknown differ from the analogous - pre actions in that they all traverse down to the "leafiest" - node they can find, in order to perform their check. The corresponding - pre actions only chain down if the [s] flag is present in the associated - phrase element(s). Also, the $ variables are not stopped by a [base] attribute - in a node.

- -

Parse Tree Nodes

- -

NLP++ can also reference the nodes of a parse tree directly. For example, - N(2) refers to the (first) node - that matched the 2nd element of a rule's right-hand-side phrase. X(3) refers to the 3rd node in the current - context path as specified by @PATH. These constructs are used in functions - that deal with the parse tree (see the Parse Tree Functions section below). - (Note that knowledge base nodes are unrelated to parse tree nodes. The - former are called NODE and the latter PNODE, as in the Types section below.)

- -

Types

- -

The types of NLP++ variables available are shown in the following table.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

NAME

 -

ABBREVIATION

 -

CHARACTERISTICS

-

string

 -

STR

 -

Alpha-numerics: e.g. G("foobar") - = "w3c";

-

integer

 -

INT

 -

32-bit integers: e.g. G("foorbar") - = 32776;

-

boolean

 -

BOOL

 -

1 or 0 (Boolean implemented as a numeric 1 - for true and the numeric 0 for false.)

-

ostream

 -

OSTREAM

 -

C++ output stream. - fileout(filename) creates a variable - with ostream value and filename << expr; - prints by referring to the ostream value of G(filename). See also - the ofile NLP++ function.

-

concept

 -

CON

 -

Concepts are the main object in the kb and - appear in the kb hierarchy. They have string names, and optionally have - one or more attributes. e.g. G("myConcept") - = makeconcept(G("root"),"My Name");

-

attribute

 -

ATTR

 -

Attributes are associated with concepts and - kb nodes. They have string names, and optionally have one or more values. - e.g. G("myAttribute") = addattr(G("myConcept"),"is");

-

value

 -

VAL

 -

Values are the values of kb attributes. They - can be strings, integers or concepts: e.g. addconval(G("myConcept"), - "is", "John Smith");

-

phrase

 -

PHR

 -

A phrase is a sequence of nodes owned by a - single concept. E.g., create the first node of a phrase: G("USPhrase") - = makephrase(G("US"),"United"); Add - the last node of the phrase: addcnode(G("US"), - "States");

-

node

 -

NODE

 -

Node in the KB. Differs from a concept in - that it is not in the KB hierarchy, but rather in a concept's phrase. - A node can serve as a "proxy" for a concept.

-

array

 -

ARRAY

 -

The values of an array, or multi-valued variable, - can be any single value (i.e., any type other than a nested ARRAY).

-

pnode

 -

PNODE

 -

Node in the parse tree. Unrelated to a KB - node.

- -

Arrays

- -

Arrays can consist of combinations of simple types. Nested array types - are not implemented. An example follows.

- -

# Create an array and step through it.
-G("counter") = 0;
-G("array_size") = 10;
-# Arrays can optionally be "declared" by padding them with 0.
-G("int_array")[G("array_size") - 1] = 0;
-
-while ( G("counter") < G("array_size") ) {
-G("int_array")[G("counter")] = G("counter");
-"output.txt" << G("int_array")[G("counter")] << "\n";
-G("counter") ++;
-}
-# print it out all at once
-"output.txt" << G("int_array") << "\n";.
-# copy it all at once:
-G("another_array") = G("int_array");

- -

Operators

- -

Most NLP++ operators are modeled on the C programming language. The - main addition is the confidence - operator, %%. This enables convenient accumulation of evidence - in terms of percent confidence. Following is an operator precedence table, - with highest precedence at the top.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

OPERATOR

 -

DESCRIPTION

 -

ASSOCIATIVITY

-

++ - --

 -

Post increment, decrement

 -

Left to right

-

++ --

 -

Pre increment, decrement

 -

Right to left

-

!

 -

Logical NOT (unary)

 -

Right to left

-

+ -

 -

Unary plus, minus

 -

Right to left

-

% * / %%

 -

Remainder, multiplication, division, confidence

 -

Left to right

-

+ -

 -

Addition, subtraction

 -

Left to right

-

 < > <= - == != >=

 -

Relational operators

 -

Left to right

-

&& ||

 -

Logical AND, OR

 -

Left to right

-

=

 -

Assignment

 -

Right to left (multiple assignment works)

-

*= /= += -= %%=

 -

Shorthand assignment (unimplemented)

 -

Right to left

-

<<

 -

Output operator

 -

Left to right

- -

Output Operator

- -

The - << output operator is modeled on the C++ operator. Only the following - form is supported, currently:

- -

filename - << expression << expression - .... ;

- -

Example:

- -

"output.txt" << "Hello, NLP++!" << "\n";

- -

NOTES: - The file named by filename should have been created by a prior fileout(filename) code - action. If that hasn't been done, then the output operator will create - the global variable G(filename) with - the output stream as its value. The default output mode for the file is - append, which can't be modified by the user at - present. Also, the default output directory is the log directory for the - current input file -- also not under user control at present.

- -

Confidence Operator

- -

Say - that something in the parse tree gives you 80% confidence that you've - found a name. You could write this into a global variable as

- -

G("name confidence") = 80;

- -

Now - you want to account for a new piece of evidence. Say the standalone confidence - of that evidence is 70%. You could combine this as

- -

G("name confidence") = G("name confidence") %% 70;

- -

which - will yield a new percent confidence between 80% and 100%.

- -

The - %% operator guarantees that "adding" or "subtracting" - confidence will always stay within the range of -100 to +100, providing - an intuitive way to gather evidence and compare confidences using a single - scale of measurement. (Actually, %% bottoms out at 0%. A planned change - to the operator will enable it to bottom out at -100%, which can be thought - of as 100% confidence against something.)

- -

Expressions

- -

NLP++ expressions are modeled on the C/C++ programming languages and - can be nested arbitrarily and grouped using parentheses ( - ).

- -

One useful enhancement: addition with string arguments or string variables - equates to string catenation.

- -

Functions and Actions

- -

NLP++ supports function calls. Actions, e.g., single, - are specialized functions that don't take generalized NLP++ expressions - as arguments (see Pass Files in the Developers - section). Currently, new NLP++ functions must be defined with C++ code, within an application's user - project. User-defined functions must be preceded by the 'user' - scope qualifier, e.g., user::myfunction. - (Interpreted NLP++ will find the correct function and issue a warning, - but user functions without the scope qualifier will be compiled incorrectly.)

- -

Statements and Blocks

- -

This too is modeled after the C programming language. Curly braces { and } - delimit blocks of statements. Statements should end with a semicolon (;). A statement can be an arbitrary - expression, a block, or fully nested if-else - statements.

- -

While Loop

- -

The only loop construct current supported is the while - loop. Here's an example that prints out 0 thru 9:

- -

G("counter") = 0;
-G("loop_limit") = 10;
-while( G("counter") < G("loop_limit")) {
-"output.txt" << "Loop counter = " << G("counter") << "\n";
-G("counter")++;
-}

- -

Parse Tree Functions

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

arraylength(VAR)

 -

INT

 -

Return number of values in given variable - or expr. (Like a call-by-reference)

-

lasteltnode(elt_num)

 -

PNODE

 -

Get the last node that matched the elt_numth - element of a rule.

-

phrasetext()

 -

STR

 -

Get the text that matched the right hand side - phrase of a rule. Analogous to $text.

-

phraseraw()

 -

STR

 -

Get the raw text that matched the right hand - side phrase of a rule. Analogous to $raw

-

pndeletechilds(pnode)

 -

None

 -

Delete children of given pnode.

-

pnname(pnode)

 -

STR

 -

Return pnode's name.

-

pnnext(pnode)

 -

PNODE

 -

Return pnode's right sibling, if any.

-

pnprev(pnode)

 -

PNODE

 -

Return pnode's left sibling, if any.

-

pnrename(pnode, - str)

 -

STR

 -

Rename given pnode. Returns the interned name.

-

pnroot()

 -

PNODE

 -

Return root of parse tree

-

pnsingletdown(pnode)

 -

PNODE

 -

Return pnode's child, obeying rules about - going down a "singlet chain." I.e., get child only if no branching - and if not going past node with BASE flag set.

-

pnup(pnode)

 -

PNODE

 -

Return pnode's parent, if any. Only leftmost - pnode in a list has a parent.

-

pnvar(pnode, - var_str)

 -

(var values)

 -

Get pnode's variable value(s).

-

varinlist(var_str, - elt_num)

 -

(var values)

 -

Get var_str variable's value(s) in any pnode - that matched elt_numth element

- -

String Functions

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

strchar(str, - num)

 -

STR

 -

Index to numth char of string

-

strchr(str, - ch_str)

 -

STR

 -

Find first occurrence of char in string. Return - string headed by char.

-

strclean(str)

 -

STR

 -

Remove leading, trailing, and repeat whitespace - separators.

-

strcontains(str1,str2)

 -

BOOL

 -

Is str1 contained in str2.

-

strcontainsnocase(str1,str2)

 -

BOOL

 -

Is str1 contained in str2, ignoring letter - case.

-

strendswith(str, - suffix_str)

 -

strstartswith(str, - suffix_str)

 -

BOOL

 -

Does given str end with given suffix_str

-

strequal(str1,str2)

 -

BOOL

 -

Is str1 identical to str2.

-

strisalpha(str)

 -

BOOL

 -

Is str all alphabetic chars.

-

strisdigit(str)

 -

BOOL

 -

Is str all numeric chars.

-

strnotequal(str1,str2)

 -

BOOL

 -

Does str1 differ from str2.

-

strequalnocase(str1,str2)

 -

BOOL

 -

Is str1 identical to str2, ignoring letter - case.

-

strnotequalnocase(str1,str2)

 -

BOOL

 -

Does str1 differ from str2, ignoring - letter case.

-

strlessthan(str1,str2)

 -

BOOL

 -

Is str1 lexically before str2.

-

strgreaterthan(str1,str2)

 -

BOOL

 -

Is str1 lexically after str2.

-

strlength(str)

 -

INT

 -

Return length of string.

-

strpiece(str, - start_num, end_num)

 -

STR

 -

Return substring from start_num to end_num - indexes. Zero-based

-

strrchr(str, - ch_str)

 -

STR

 -

Find last occurrence of char in string. Return - string headed by char

-

strsubst(str, - old_str, new_str)

 -

STR

 -

In str, substitute instances of old_str with - new_str.

-

strtolower(str)

 -

STR

 -

Return lowercase form of string.

-

strtotitle(str)

 -

STR

 -

Capitalize string as in a title or header.

-

strtoupper(str)

 -

STR

 -

Return uppercase form of string

-

strtrim(str)

 -

STR

 -

Remove leading and trailing whitespace from - str.

-

unpackdirs(dir_str)

 -

STR ARRAY

 -

Unpack directory names from a full directory - path string.

-

strwrap(str, - num)

 -

STR

 -

Fold string successively at every numth character.

-

num(str)

 -

NUM

 -

Convert string to number, if possible. - (Also accepts num arg).

-

str(num)

 -

STR

 -

Convert numeric value to string. (Also - accepts string arg).

- -

Miscellaneous Functions

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

fail()

 -

None

 -

In @CHECK region, abort the rule that has - just matched. Noop elsewhere.

-

succeed()

 -

None

 -

In @CHECK and @CODE region, succeed without - performing further code actions in that region.

-

batchstart()

 -

BOOL

 -

True if current file is first in a batch run.

-

debug()

 -

BOOL

 -

No op. Convenient breakpoint for debugging - from C++ compiler.

-

hitconf(hits_num, - total_num, fudge_num)

 -

0 to 100

 -

Percent confidence for hits. Fudge factor - is typically 3 to 20.

-

LJ(num, fieldsize_num)

 -

STR

 -

Left-justify num in given field size.

-

ofile(path_str - [, mode_str [, mode_str [, mode_str]]] )

 -

OSTREAM

 -

Open an output file for the given path; analogous - to C++ open file function. Zero to three mode strings selected from: 1) - "app" = append; 2) "ate" = special - C++ append; 3) "binary" = open in binary mode (text is default). - WARNINGS: Default mode is OVERWRITE, not APPEND. Unlike fileout(), does - not create a global variable for the file stream. [Upgraded - doc to conform to standard C++ i/o libraries.]

-

percentstr(numerator_num, - denominator_num)

 -

STR (" - 0" to "100")

 -

Format a percentage (right-justified in field - of 3 chars).

-

rightjustifynum(num, - fieldsize_num)

 -

STR

 -

Format a percentage (right-justified in field - of 3 chars).

-

spellcandidates(word_str)

 -

STR

 -

Return list of space separated candidates - for spell-correcting given word.

-

spellcorrect(word_str)

 -

STR

 -

If known word, return as-is. If not, return - best spelling guess.

-

spellword(word_str)

 -

BOOL

 -

See if given word is a known English word.

-

strspellcandidate(word_str, - list_str)

 -

STR

 -

Choose best spell-correct candidate for a - word, given a list of space-separated candidates (as returned by spellcandidates, - e.g.)

-

strspellcompare(str1, - str2)

 -

INT

 -

Return spelling "distance" between - two given words. The smaller the number, the fewer the corrections to - get from one to the other.

-

system(str)

 -

BOOL

 -

Execute str as operating system command

- -

Note: It is necessary to escape the directory delimiter backslash ('\') - with another ('\'). For example, suppose you wanted to echo characters - to a file and then copy that file to some other directory. Here's some - example code:

- -

System Function Example

- -

@NODES _ROOT
-@POST
-# echo.txt is put in your default console location, which for
-# me happens to be D:\WINNT\Profiles\Administrator\Desktop
-system("echo this is a test > echo.txt");
-# echo1.txt ends up in D:\WINNT\Profiles\Administrator\Desktop
-system("copy echo.txt echo1.txt");
-# echo2.txt ends up on D:
-system("copy echo1.txt D:\\echo2.txt");
-# echo3.txt ends up on D:\apps
-system("copy D:\\echo2.txt D:\\apps\\echo3.txt");
-"output.txt" << "copy.\n" ;
-noop();
-@@POST
-@RULES
-_xNIL <- _xALPHA @@

- - - - diff --git a/Help/helps/RECURSE.htm b/Help/helps/RECURSE.htm deleted file mode 100644 index d31923b..0000000 --- a/Help/helps/RECURSE.htm +++ /dev/null @@ -1,201 +0,0 @@ - - - - - - - - - - -@RECURSE Region - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

RECURSE Region

- -

The RECURSE Region is a nested "minipass" within a pass file. - It can only be invoked from an element of a rule that has matched. The - phrase that the recurse minipass operates on is the set of nodes that - matched the invoking rule element. Each RECURSE Region contains within - it a Grammar Region.

- -

Example of RECURSE Region

- -

Here's an example taken from the RFA (Rules-File Analyzer), which parses - the pass files themselves.

- -

@RECURSE actionpass # Start of a RECURSE Region. A name, e.g. "actionpass", is required.

- -

@RULES # Start of one RULES Regions within the RECURSE minipass.

- -

# The minipass converts literals (e.g., "single") to actions in context.

- -

_ACTION <- _LIT @@

- -

@@RECURSE # End of the RECURSE minipass. Note that @@RULES is optional.

- -

# .... # Additional RECURSE minipasses may be specified here.

- -

# The GRAMMAR ZONE (i.e., the main Grammar Region of the pass file) starts here. It has no marker.
-# Note that in this case, the @@RECURSE terminator is required, so that the main Grammar Region is not
-# taken to be part of the previous RECURSE minipass.

- -

@RULES

- -

_POSTS [base] <- # Simplified rule for identifying a @POST Region.

- -

_soPOST # @POST
-_xWILD [matches=(_LIT _ACTION) recurse=(actionpass)]
-_eoPOST [opt] # @@POST is optional.

- -

@@

- -

In the above example, when the NLP++ parser (RFA or RFB) sees a plain - word like "single" in a file, - it doesn't know whether to reduce it to an _ACTION or not. But within - the markers @POST and @@POST, the literal should be treated as an action. - This allows the parentheses in pass file actions to be optional, when - the action takes no arguments. E.g., "single()" and "single" - are equivalent.

- -

In general, RECURSE mini-passes enable a mechanism for "roughing - out" a text and then applying the correct assignments when the context - has been determined unambiguously. They have many other uses, such as - traversing nodes that matched a rule element in order to perform cleanups, - data gathering, deletions, and so on.

- -

For a phrase element modifier that invokes - a recursive rules pass, see the recurse - modifier.  

- - - - - - -
-

Note: - A RECURSE Region cannot contain a CODE Region or a SELECT Region. A SELECT - Region would be meaningless, since the RECURSE Region operates on a phrase - specified by the rule element that invokes it.

- -

See Also

- -

Minipass - Zone

- -

About - Pass Files

- -

Grammar - Region

- - - - diff --git a/Help/helps/RULES.htm b/Help/helps/RULES.htm deleted file mode 100644 index e3b0465..0000000 --- a/Help/helps/RULES.htm +++ /dev/null @@ -1,111 +0,0 @@ - - - - - - - - - - -@RULES Region - - - - - - - - - - - - - - - - - - - - - - - - - - - -

RULES Region

- -

The RULES Region is where you write the actual NLP++ rules.  The - RULES Region begins with the @RULES marker and optionally terminates with - the @@RULES marker.

- -

Action Regions (i.e. PRE, CHECK, and POST) that precede a RULES Region - apply to every rule within the RULES Region.

- -

A RULES Region with no preceding Action Regions gets the default reduce - action, single() applied to every - rule match within it.

- -

Common Error: Not Resetting Actions

- -

In order to reset the actions that apply to a RULES Region, one must - reset actions by adding a @RULES marker before the rules to be reset.

- -

Example:

- -

@POST
-N("noun", 3) = "true";

- -

single();
-@RULES
-_np <- _det _quan _xALPHA @@
-_np <- _det _xALPHA @@

- -

In the above, we want to mark alphabetic tokens as possible nouns based - on their context. But the code in the POST Region sets a variable called - "noun" to "true" in the node matching the THIRD element - of each rule in the RULES Region. However, this will fail, since the second - rule has only two elements on the right hand side. Therefore, this example - must be rewritten as

- -

@POST
-N("noun", 3) = "true";

- -

single();
-@RULES
-_np <- _det _quan _xALPHA @@

- -

 

- -

@POST
-N("noun", 2) = "true";

- -

single();
-@RULES
-_np <- _det _xALPHA @@

- -

It's even easier to forget to add a lone @RULES so as to reset to the - default action for subsequent rules.

- -

See Also

- -

Grammar - Region

- -

Rule - Syntax

- -

PRE Region

- -

CHECK - Region

- -

POST - Region

- - - - diff --git a/Help/helps/Resume_Analyzer.htm b/Help/helps/Resume_Analyzer.htm deleted file mode 100644 index 55ed07c..0000000 --- a/Help/helps/Resume_Analyzer.htm +++ /dev/null @@ -1,66 +0,0 @@ - - - - - - - - - - -Resume_Analyzer - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Resume_Analyzer

- -

NOTE: - Our Resume analyzer project ended in 2001, - with an ancient version of VisualText and NLP++. The project emphasized - development of the RUG automated rule generation capability (a prototype - capability in VisualText).  Please - don't judge VisualText and NLP++ - by this prehistoric analyzer!  Instead, - look at TAIParse as an example - of the things you can do with VisualText and NLP++.

- -

An advanced resume analysis prototype is available via download from - our website.  It - extracts contact information, experience and education records from resumes - at 80% accuracy (90% precision and 75% recall/completeness).  It - is available at

- -

http://dev.textanalysis.com/Rez0_3_1.zip - (latest compiled version)

- -

http://www.textanalysis.com/Apps/Information_Extraction/information_extraction.html

- -

 

- - - - diff --git a/Help/helps/S.htm b/Help/helps/S.htm deleted file mode 100644 index 6f2d780..0000000 --- a/Help/helps/S.htm +++ /dev/null @@ -1,50 +0,0 @@ - - - - - - - - - - -S - - - - - - - - - - - - - - - - - - - - - - - - - - - -

S

- -

[1] Reference to a parse tree node or its variable.  See - Variables.

- -

[2] Reference to a phrase element modifier.  s is an - alternate form for the phrase element modifier singlet.

- - - - diff --git a/Help/helps/SELECT.htm b/Help/helps/SELECT.htm deleted file mode 100644 index ddd50d3..0000000 --- a/Help/helps/SELECT.htm +++ /dev/null @@ -1,257 +0,0 @@ - - - - - - - - - - -@SELECT Region - - - - - - - - - - - - - - - - - - - - - - - - - - - -

SELECT Region

- -

The SELECT Region specifies methods for selecting context - nodes in the parse tree. Each selected - node serves as a locus for rule matching.

- -

Selectors

- -

The SELECT Region may include at most one selector, - chosen from the following:

- - - - - - - - - - - - - - - - - - - -
-

SELECTOR

 -

DESCRIPTION

-

@NODES

 -

A list - of node names to be searched for independently in the parse tree. The - phrase immediately under such a context node will be traversed to look - for rule matches. NOTE: - The selection algorithm will only find such a node when all of its ancestors - have the unsealed attribute set.

-

@PATH

 -

A path - of node names that must start with _ROOT, the root of the parse tree. - The path specifies the root, its child, grandchild, and so on. The selection - algorithm traverses the parse tree starting at the root and searches for - a context node that matches the entire path. The phrase immediately under - such a context node will be traversed to look for rule matches. (The PATH - Selector does not check the unsealed - flag.)

-

@MULTI

 -

A list - of node names to be searched for independently in the parse tree. The - entire subtree under such a context node will be traversed to look for - rule matches. (The MULTI Selector currently does not check the unsealed - flag.)

- -

In addition to the selector markers @NODES, @PATH, and @MULTI there - are optional markers that indicate the termination of a selector specification. -  The corresponding - markers are  @@PATH, - @@NODES, and @@PATH.

- -

The SELECT Region itself is optional. If the region is not included - in the pass file, the default is to match the rules of the file against - the top-level phrase of the parse tree, that is, the phrase directly under - the _ROOT node. The default could - be made explicit, as in the following:

- - - - - - - - - - - - - - - -
-

@SELECT

 -

# Start of SELECT Region

-

@NODES _ROOT

 -

# Specifies nodes whose underlying phrase - will be matched against.

-

@@SELECT

 -

# Optional end-marker for SELECT Region

- -

(The region markers @SELECT and @@SELECT are also optional. The sample - SELECT Region above could be simply @NODES _ROOT.)

- -

See Also

- -

Context - Zone

- -

NODES - Selector

- -

PATH - Selector

- -

MULTI - Selector

- -

 

- - - - diff --git a/Help/helps/SPLASH_SCREEN/Welcome.htm b/Help/helps/SPLASH_SCREEN/Welcome.htm deleted file mode 100644 index b118a73..0000000 --- a/Help/helps/SPLASH_SCREEN/Welcome.htm +++ /dev/null @@ -1,89 +0,0 @@ - - - - - - - - - - -Welcome - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

 

- -

- -

*** New to VisualText? Start by taking the Quick Tour! ***

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/Sample_Analyzers/Corporate_Analyzer.htm b/Help/helps/Sample_Analyzers/Corporate_Analyzer.htm deleted file mode 100644 index c8ad60c..0000000 --- a/Help/helps/Sample_Analyzers/Corporate_Analyzer.htm +++ /dev/null @@ -1,60 +0,0 @@ - - - - - - - - - - -Corporate Analyzer - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Corporate Analyzer Tour

- -

The tour of the Corporate Analyzer (in the VisualText\apps folder) is - intended to take you through the step by step process used in building - the Corporate Analyzer. It will give you an idea of the approach and techniques - used to build analyzers. This tour is a supplement to the VisualText Quick - Tour.  Study - of the tutorials is also recommended for a more complete understanding - of this tour.

- -

We recommend you run the Corporate Analyzer application as you go through - the tour. This will give you an opportunity to see how the analyzer works - as you read through the discussion.

- -

Tour the Corporate Analyzer.

- -

 

- - - - diff --git a/Help/helps/Sample_Analyzers/Database1_Analyzer.htm b/Help/helps/Sample_Analyzers/Database1_Analyzer.htm deleted file mode 100644 index fc410b0..0000000 --- a/Help/helps/Sample_Analyzers/Database1_Analyzer.htm +++ /dev/null @@ -1,100 +0,0 @@ - - - - - - - - - - -Database1 Analyzer - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Database1 Analyzer

- -

VisualText version 1.4 introduces database connectivity functions within - NLP++.  The - database1 sample analyzer (in VisualText\apps) connects to a database, - sends commands to it, retrieves data from it, and prints the results to - a file.  Some - features:

- - - -

To set up the precise example within this analyzer, you will first need - to install a database.  We - recommend downloading MySQL and MyODBC and configuring them as described - at their respective web sites.  Copy - the database1 folder from samples to c:\apps.  Then - set up an ODBC connection to a database called test (making sure to match - the account information to that in c:\apps\database1\data\createdb).  Then - one can run a command such as

- -

c:\mysql\bin\mysql test - < c:\apps\database1\data\createdb

- -

in order to create the same "test" database tables and values - as used in the database1 analyzer.

- -

The two pass files within the database1 analyzer contain further notes - regarding this application.

- -

 

- - - - diff --git a/Help/helps/Sample_Analyzers/GenTxtDocs/GenTxt1.htm b/Help/helps/Sample_Analyzers/GenTxtDocs/GenTxt1.htm deleted file mode 100644 index 125dbd1..0000000 --- a/Help/helps/Sample_Analyzers/GenTxtDocs/GenTxt1.htm +++ /dev/null @@ -1,226 +0,0 @@ - - - - - - - - - - -The GenTxt Sample Application - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

The GenTxt General Analyzer

- -

GenTxt is not currently supported. It is superseded - by TAIParse, which is available at

- -

http://www.textanalysis.com/Apps/POS_Tagger/pos_tagger.html

- -

An older version, - which includes a semantic analysis framework, is also available at

- -

http://www.textanalysis.com/Apps/Natural_Language_Processing/natural_language_processing.html

- -

A knowledge - base that embeds WordNet is also available on request.

- -

 

- - - - diff --git a/Help/helps/Setting_Rule_Generation_Properties.htm b/Help/helps/Setting_Rule_Generation_Properties.htm deleted file mode 100644 index 2bb6c65..0000000 --- a/Help/helps/Setting_Rule_Generation_Properties.htm +++ /dev/null @@ -1,776 +0,0 @@ - - - - - - - - - - -Setting Rule Generation Properties - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Setting Rule Generation Properties

- -

You can control the generation of particular rule sets by setting attributes - for concepts in the Gram (or sample) hierarchy. These attributes control - aspects of rule generation and specify actions to associate with generated - rules.

- -

Gram Concept Properties

- -

Properties are set using the Gram Concept - Properties dialog (below). The dialog box is accessed from the - Gram Tab by selecting Properties - from the context menu. The various attributes in the Gram Concept Properties - are outlined below.

- -

- - - --- - - - - - - - - - - - - - - - - - - - - - - - -
-

Item

 -

Description

-

Name

 -

Name of the current concept.

-

Sample File

 -

Associated file of samples to import/export, - if any.

-

Attributes

 -

Controls for rule generation (see tables below.)

-

Done

 -

Closes the Gram Concept Properties dialog.

-

Help

 -

Launches Help documentation.

- -

 

- -

Attributes for Rule Sets

- -

The user can choose to preserve or discard sets of rules that are automatically - generated from samples. Some sets are useful primarily for debugging the - rule generation process. They might also be useful in very highly specialized - contexts within a text.

- -

The Constrain rule set is by far the most generally useful set of rules.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

ATTRIBUTE

 -

VALUES

 -

DESCRIPTION

-

RAW

 -

true/false/<NONE>

 -

Keep rules that collect - each token of a sample into a list of all its suggested concepts. E.g.,
-_abbrev <- _xWILD [s one match=(_city _state NY)] @@

-

LIT

 -

true/false/<NONE>

 -

Keep rules that collect - the most literal possible form of each token. NOTE: If parts of a sample - are already reduced by prior rules, then the rule will not look as literal - as expected.

-

CONSTRAIN

 -

true/false/<NONE>

 -

Keep rules that merge samples with same length - and general composition, constrained as far as possible. E.g., given samples - "1261 Starlit Dr." and "3 Elm St.", build a rule like
-@PRE
-<3,3> cap()
-<5,5> cap()
-@RULES
-_street <- _xNUM [s layer=(_number)] _xALPHA [s layer=(_name)] _poROAD - [s layer=(_road)]  \.[s] - @@

-

SPLIT

 -

true/false/<NONE>

 -

Keep unconstrained rules that merge samples - as in GEN, with consistent labeling (i.e., layering). E.g.
-_street <- _xNUM [s layer=(_number)] _xALPHA [s layer=(_name)] _xALPHA - [s layer=(_road)]  \.[s] - @@

-

GEN

 -

true/false/<NONE>

 -

Keep rules that merge samples with same length - and general composition. E.g.,
-_street <- _xNUM [s] _xALPHA [s] _xALPHA [s] \.[s] @@

- -

Each literal rule is labeled as completely - as possible, depending on other samples that the rule merged with.

- -

Miscellaneous Attributes

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

ATTRIBUTE

 -

VALUES

 -

DESCRIPTION

-

CLOSED

 -

true/false/<NONE>

 -

(Operates on LABEL concepts. If set for a - nonlabel concept, applies to every element of every rule in the subhierarchy). - Merge the literal values due to each sample, rather than building generalized - elements such as _xALPHA and _xNUM.

-

LAYER

 -

string

 -

Create a layer key-value pair in the suggested concept for rules in - the Gram subhierarchy. E.g, layer=_Caps will create rules like _city [layer=(_Caps)] - <- New [s] York [s] @@

-

QUICKSEM

 -

true/false/<NONE>

 -

Create semantic variables - in the suggested concept of rules, using label concepts to name the variables. - E.g., if the Gram hierarchy has a rule concept humanName and under it - the label concepts  firstName - and lastName, then the rules built for humanName will create variables - called firstName and lastName with the appropriate text as their value. - Generates NLP++ POST Actions such as
-S("firstName") = N("$text",1);
-S("lastName") = N("$text",2);

-

LOCKED

 -

true/false/<NONE>

 -

If true, keeps the rule - generator from traversing the locked subtree. If there are new samples, - the rule generator will override the lock on the concepts owning the samples.

-

DIRTY

 -

true/false/<NONE>

 -

[INTERNALS] True if the - subtree will have rules for it regenerated with a Generate Dirty menu - item. To set this manually in VisualText, use the Mark For Generation - menu item.

-

XWHITE

 -

true/false/<NONE>

 -

Expand any whitespace - present in a rule to an element like _xWHITE - [s star]. (That is, zero or more white spaces.)

- -

Inherited attributes - By setting attributes - at a concept, the entire subtree under that concept inherits those attribute - values. In this way, the user can quickly control the way rules will be - generated for an entire sub-hierarchy. This method also enables the user - to override the settings wherever desired.

- -

Default - attribute values - The CONSTRAIN attribute and the XWHITE attribute - are true except where overridden in the Gram Hierarchy. Similarly, all - other attributes are empty or FALSE, as appropriate.

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/TAIParse.htm b/Help/helps/TAIParse.htm deleted file mode 100644 index 04f736f..0000000 --- a/Help/helps/TAIParse.htm +++ /dev/null @@ -1,82 +0,0 @@ - - - - - - - - - - -TAIParse - - - - - - - - - - - - - - - - - - - - - - - - - - - -

TAIParse

- -

TAIParse is a general analyzer for English.  It - is the best starting point for many text analysis and information extraction - tasks.

- -

NOTE: The latest TAIParse release now includes its compiled run.dll - library, making it easy to call outside VisualText.  (Still - need VisualText installed for its runtime libraries).  The - latest TAIParse (0.8.1.2 at this time) will be placed at

- -

http://dev.textanalysis.com/taiparse.zip

- -

NOTE 2: It is compiled with Microsoft Visual C++ 2008.  If - you are running an older version, you'll need to compile it for yourself - or get help from TAI.

- -

 

- -

Two older versions can still be downloaded from our website.

- -

1. - Part-of-speech tagger.  This - is the latest version that focuses on accurate part-of-speech (POS) tagging. -  The URL - below is for a beta version at about 90% accuracy.  It - can be run (and modified) within VisualText, and also can be run as a - standalone executable program.  We - also have available a newer version of the tagger at 94% - accuracy -- contact TAI for licensing the absolute latest.

- -

http://www.textanalysis.com/Apps/POS_Tagger/pos_tagger.html

- -

2. Another version of TAIParse is available, which incoporates a sample - semantic analysis system.  The - capabilities of this system are being migrated to the one above.

- -

http://www.textanalysis.com/Apps/Natural_Language_Processing/natural_language_processing.html

- -

 

- - - - diff --git a/Help/helps/Table_of_Array_Functions.htm b/Help/helps/Table_of_Array_Functions.htm deleted file mode 100644 index 8f01d0a..0000000 --- a/Help/helps/Table_of_Array_Functions.htm +++ /dev/null @@ -1,186 +0,0 @@ - - - - - - - - - - -Table of Array Functions - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Table of Array Functions

- -

Arrays and lists are treated equivalently in NLP++, and we'll continue - to call them both "arrays" for conciseness.

- -

The current section is meant to round out the treatment of list-like - capabilities such as pushing a value to the front.  Up - to now, arrays were grown mainly by indexing to the last element.

- -

Arrays can also be assigned and manipulated with parse tree functions - and kb functions.

- - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

push(var)

 -

ARRAY

 -

Push a value to the front of a variable's - value(s).  Allows - mixed types (string, num, flt, concept...), allows redundant values, does - not check uniqueness. NOTE: This will alter the array indexing, pushed - item will be the 0th element of array index. UNIMPLEMENTED

-

 

 -

 

 -

 

- -

 

- -

 

- -

See Also

- -

 

- -

 

- - - - diff --git a/Help/helps/Table_of_Database_Functions.htm b/Help/helps/Table_of_Database_Functions.htm deleted file mode 100644 index 1974aed..0000000 --- a/Help/helps/Table_of_Database_Functions.htm +++ /dev/null @@ -1,440 +0,0 @@ - - - - - - - - - - -Table of Database Functions - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Table of Database Functions

- -

The NLP++ functions for interfacing with a database are listed here - in table form for easy reference.  For - examples of these functions, refer to the individual function pages included - in the Database Functions section.  Also - available is a sample analyzer illustrating connection to a database, - in Samples\database1.  An - introductory page is available at Database1 - Analyzer.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

dballocstmt()

 -

BOOL

 -

Create a handle for an active database statement. -  (Stored - internally, allows only one active handle at a time.)

-

dbbindcol(column_num, type_str, - size_str, &var, &flag_var)

 -

BOOL

 -

Bind the given database column number to the - NLP++ variable var. type_str - specifies the database data type being fetched.  size_str - specifies the size needed to store the database value being retrieved - -- for some types, it is ignored.  flag_var - is a second NLP++ variable used to flag whether the currently fetched - row contains a value in the column being fetched.  Calls - to dbbindcol are followed by calls - to dbfetch in order to fetch successive - database rows.  Note - that the ampersands are needed to indicate call by reference, i.e., that - values are being passed up to the two NLP++ variables.  E.g., - dbbindcol(1,"string",255, - &G("employee name"),&G("result flag"));

-

dbclose()

 -

BOOL

 -

Close the currently open database.  (Only - one database can be open at a time.)

-

dbexec(cmd_str)

 -

BOOL

 -

Send a statement to the currently open database. -  A modular - function that creates a statement handle, sends the command string, then - closes the statement handle.

-

dbexecstmt(cmd_str)

 -

BOOL

 -

Send a statement using the currently open - statement handle.  Assumes - a statement has been opened with dballocstmt.

-

dbfetch()

 -

BOOL

 -

Fetch a row from a database.  Columns - bound by previous calls to dbbindcol - will get their variables filled with values from the columns in this database - row.  Returns - 1 as long as rows successfully fetched, else 0.

-

dbfreestmt()

 -

BOOL

 -

Frees the statement handle allocated by dballocstmt.  Only - one statement handle can be active at a time.

-

dbopen(db_str [, user_str] [, password_str])

 -

BOOL

 -

Open the given database name with user name - and password.  Depending - on the database and account set up, the user name and password may be - optional.

-

 

 -

 

 -

 

- -

 

- -

 

- -

See Also

- -

Database1 Analyzer

- -

 

- - - - diff --git a/Help/helps/Table_of_Formatting_and_I_O_Functions.htm b/Help/helps/Table_of_Formatting_and_I_O_Functions.htm deleted file mode 100644 index 494ce8a..0000000 --- a/Help/helps/Table_of_Formatting_and_I_O_Functions.htm +++ /dev/null @@ -1,455 +0,0 @@ - - - - - - - - - - -Table of Formatting and I/O Functions - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Table of Formatting and I/O Functions

- -

The NLP++ functions used formatting and performing input and output - operations are listed here in table form for easy reference.  For - examples of these functions, refer to the individual function pages included - in the Formatting and I/O Functions section.  

- -

The return types include:  STR - (string), OSTREAM (output stream) - and BOOL (Boolean).  Boolean - type is currently implemented by integer 1 - (or non-zero) for true and 0 for false.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

cbuf()

 -

OSTREAM

 -

Output to buffer that user supplied for the - current parse (standalone or embedded only).

-

cout()

 -

OSTREAM

 -

Output to standard output or user-supplied - stream (the latter in standalone or embedded only).

-

coutreset(fileNameStr)

 -

OSTREAM

 -

Bind cout() stream to the given filename.

-

closefile(fileOstream)

 -

BOOL

 -

Close the output file stream.

-

inputrangetofile(start_n,end_n,out_ostr)

 -

BOOL

 -

Print - a range of chars from the input text to an output stream.  (Listed - with Parse Tree Functions)

-

LJ(num, fieldsize_num)

 -

STR

 -

Left-justify num in given field size.

-

openfile(fileNameStr)

 -

OSTREAM

 -

Open output file stream.

-

percentstr(numerator_num, - denominator_num)

 -

STR (" - 0" to "100")

 -

Format a percentage (right-justified in field - of 3 chars).

-

rightjustifynum(num, - fieldsize_num)

 -

STR

 -

Right-justify num in given field size.

-

today()

 -

STR

 -

Format the current date and time as a string.

- -

See Also

- -

Special - Functions

- -

Knowledge - Base Functions

- -

Parse - Tree Functions

- -

String - Functions

- -

Spelling - Functions

- - - - diff --git a/Help/helps/Table_of_Knowledge_Base_Functions.htm b/Help/helps/Table_of_Knowledge_Base_Functions.htm deleted file mode 100644 index ff2e972..0000000 --- a/Help/helps/Table_of_Knowledge_Base_Functions.htm +++ /dev/null @@ -1,3307 +0,0 @@ - - - - - - - - - - -Table of Knowledge Base Functions - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Table of Knowledge Base Functions

- -

The functions used to operate on objects in the Knowledge Base have - been listed here in table form for easy reference.  For - examples of these functions, please refer to the individual function pages - within the subsections of the current Help section, Knowledge Base Functions.

- -

The names of KB objects are abbreviated as follows: concept (CON), - attribute (ATTR), value (VAL), and phrase (PHR). -

- -

The return types include: CON, - ATTR, VAL, - PHR, INT - (integer),  FLT - (float), STR (string), BOOL - (Boolean).  Boolean - type is currently implemented by integer 1 - (or non-zero) for true and 0 for false.) -  NONE - indicates nothing is returned.

- -

In the table below, a name must - be a STR, parent is a CON, attr_str is a STR, hier - is a CON.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FUNCTION(ARG1,ARG2...)

 -

RETURN TYPE

 -

DESCRIPTION

-

Command - File Functions

 -

 

 -

 

-

kbdumptree(root_con, file_str)

 -

BOOL

 -

Create a dump file of knowledge base concept.

-

take(file_str)

 -

BOOL

 -

Execute commands in a knowledge base command - file (.KB file)

-

writekb()

 -

BOOL

 -

Dumps the entire KB currently in memory to the kb/user directory which includes a word.kb, attrs.kb, and main.kb file.

-

Functions - that Find and Fetch

 -

 

 -

 

-

findroot()

 -

CON

 -

Find the root concept of the knowledge - base (named concept).

-

findconcept(con, - str)

 -

CON

 -

Find and - return the concept with name str under the parent concept con.

-

findconcept(con, - num)

 -

CON

 -

Find concept - number num - under the parent concept con.

-

findattr(con, - str)

 -

ATTR

 -

Fetch attribute named str - belonging to concept con.

-

findattrs(con)

 -

ATTR

 -

Fetch concept's list of attributes.

-

attrname(attr)

 -

STR

 -

Fetch attribute's name.

-

attrtype(concept,attrName)

 -

NUM

 -

Fetch an attribute's type. 0 = string, 1 = number, 2 = concept, 3 = float.

-

attrvals(attr)

 -

VAL

 -

Fetch attribute's list of values.

-

findvals(con, - name)

 -

VAL

 -

Fetch list of values for a concept's attribute.

-

numval(con, - name)

 -

NUM

 -

Fetch numeric-value of attribute (must be - first).

-

fltval(con, name)

 -

FLT

 -

Fetch float value of attribute (must be first).

-

strval(con, - name)

 -

STR

 -

Fetch string-value of attribute (must be first).

-

conval(con, - name)

 -

CON

 -

Fetch concept-value of attribute (must be - first).

-

attrwithval(con, attr_str, val_str)

 -

BOOL

 -

Check if concept's attribute has given - value (multiple-value aware). Note that second two arguments must be STR.

-

inheritval(con, - name, topCon)

 -

STR

 -

Find string attribute's value up the hierarchy. - con is the current concept, topCon is the root to be searched up - to, and name is the name of the - attribute. If hier==0, goes to root of KB.

-

conceptname(con)

 -

STR

 -

Fetch name of given concept.

-

conceptpath(con)

 -

STR

 -

Fetch entire path of given concept as a string.

-

pathconcept(str)

 -

CON

 -

Fetch concept, given the path str.

-

wordpath(str)

 -

STR

 -

Get entire path of dictionary concept for - the given string.

-

findwordpath(str)

 -

STR

 -

Find entire path of dictionary - concept for the given string. (If not present, don't add the word.)

-

wordindex(str)

 -

CON

 -

Fetch index entry for dictionary concept. -  Gets index - concept that str would be added under.

-

findhierconcept(name, - hier)

 -

CON

 -

Find a concept in the subhierarchy of a given - concept. e.g.: G("returnedConcept") = findhierconcept("child",G("ancestor")); - finds the concept named "child" anywhere under the concept pointed - to by G("ancestor").

-

dictfindword(str)

 -

CON

 -

Find dictionary concept - in dictionary hierarchy of KB.

-

dictfirst()

 -

CON

 -

Fetch the first word-concept in the KB dictionary - hierarchy. (New in 1.6)

-

dictnext(con)

 -

CON

 -

Fetch the word following the given word in - the KB dictionary hierarchy. - (New in 1.6)

-

attrexists(hier, - attr_s, val_s)

 -

BOOL

 -

Find attribute and value pair in the subhierarchy - of a concept.

-

attrchange(hier, - attr_s, val_s, new_s)

 -

BOOL

 -

Replace all matching attribute-value pairs - in the given hierarchy to have the new string value, new_s.

-

down(con)

 -

CON

 -

Fetch the first child of given concept.

-

up(con)

 -

CON

 -

Fetch the parent of the given concept.

-

prev(con)

 -

CON

 -

Fetch the left or previous sibling of concept.

-

next(con)

 -

CON

 -

Fetch the right or next sibling of concept.

-

nextattr(attr)

 -

ATTR

 -

Fetch the next attribute in a list of attributes

-

nextval(val)

 -

VAL

 -

Fetch the next value in a list of values.

-

getsval(val)

 -

STR

 -

Fetch value as a string from a numeric or - string-valued KB value object.

-

getstrval(val)

 -

STR

 -

Fetch the string value from a KB value object.

-

getfltval(val)

 -

FLOAT

 -

Fetch number from float VAL.

-

getnumval(val)

 -

INT

 -

Fetch number from numeric VAL.

-

getconval(val)

 -

CON

 -

Fetch concept from value.

-

Functions - that Create

 -

 

 -

 

-

makeconcept(parent, - name [, num] )

 -

CON

 -

Make concept under given parent concept. Optionally - as the numth child of the parent. num==0 or absent places concept at end - of list of children. name is the name of new concept.

-

addattr(con, - attr_s)

 -

ATTR

 -

Add an attribute with no value to given concept.

-

addsval(con, - name, num)

 -

NONE

 -

Add numeric - value num as a string to concept con's attribute called name.

-

addstrval(con, - name, str)

 -

NONE

 -

Add str - as string value to concept con's attribute called name.

-

addnumval(con, - name, num)

 -

NONE

 -

Add num - as numeric value to concept con's attribute called name.

-

addconval(con, - name, con_val)

 -

NONE

 -

Add concept - value con_val to concept's name attribute.

-

getconcept(parent, - name)

 -

CON

 -

Get or - make named concept under parent.

-

addword(str)

 -

CON

 -

Add - dictionary concept to the dictionary within the KB, if not present. In - either case, fetch the dictionary concept for the word.

-

dictgetword(str)

 -

CON

 -

Same as addword; - more principled function name.

-

sortconsbyattr(arr_c, - attr_s, numeric_b, descending_b)

 -

ARR

 -

Given an array (multi-valued - variable) of KB concepts, sort the concepts by the given attribute. numeric_b - = 1 if num, 0 if string type attribute.  descending_b - = 1 if descending order, 0 if ascending.  E.g.,

-

G("sorted") - = sortconsbyattr(G("cons"),"count",1,0);

-

will sort an array of KB - concepts by their "count" attribute in ascending order, assigning - a new sorted array to G("sorted").  The - given array in G("cons") is left unaltered. New - in 1.6

-

Functions - that Move and Remove

 -

 

 -

 

-

rmconcept(con)

 -

BOOL

 -

Remove concept con from Knowledge Base. - Removes entire subhierarchy.

-

rmchild(con, str)

 -

BOOL

 -

Remove child named str from parent concept con.

-

rmchild(con, num)

 -

BOOL

 -

Remove number num - child of parent concept con.

-

rmvals(con, str)

 -

BOOL

 -

Remove attributes and - values of concept con's attribute - named str.

-

rmval(attr, val)

 -

BOOL

 -

Remove value val - from attribute attr.

-

rmattrval(con, str1, str2)

 -

BOOL

 -

Remove value named str2 from attribute named str1 under concept con. Also removes the attribute.

-

rmattrs(con)

 -

BOOL

 -

Remove concept's attributes (except for system - internal ones).

-

rmattr(con,name)

 -

BOOL

 -

Remove concept's named - attr and values.

-

rmchildren(con)

 -

BOOL

 -

Remove concept's children - and phrase.

-

rmword(str)

 -

BOOL

 -

Remove dictionary concept from KB.

-

prunephrases(hier)

 -

BOOL

 -

Remove all phrases from - given subhierarchy.

-

replaceval(con, - name, str)

 -

NONE

 -

Replace named attribute's value(s) with str.

-

replaceval(con, - name, num)

 -

NONE

 -

Replace - named attribute's value(s) with num.

-

replaceval(con, - name, con_val)

 -

NONE

 -

Replace - named attribute's value(s) with concept con_val

-

renameconcept(con, - name)

 -

NONE

 -

Rename - given concept to name.

-

renamechild(con, - num, name)

 -

NONE

 -

Rename - con's numth child concept to name.

-

renameattr(con, - name, new)

 -

NONE

 -

Rename - con's named attribute to new.

-

movecleft(con)

 -

NONE

 -

Move concept before previous sibling.  (Moves - concept to the left in its list.)

-

movecright(con)

 -

NONE

 -

Move concept after next sibling.  (Moves - concept to the right in its list.)

-

sortchilds(con)

 -

NONE

 -

Sort concept's immediate children alphabetically.

-

sorthier(con)

 -

NONE

 -

Sort concept's subhierarchy alphabetically.

-

Phrase - and Node Functions

 -

 

 -

 

-

findphrase(con)

 -

PHR

 -

Fetch concept's phrase.

-

sortphrase(con)

 -

NONE

 -

Sort concept's phrase nodes alphabetically.

-

phraselength(con)

 -

INT

 -

Get number of nodes in concept's phrase.

-

nodeconcept(node)

 -

CON

 -

Fetch the concept that node is a proxy for. - (Changed in 1.6)

-

nodeowner(node)

 -

CON

 -

Fetch the concept that owns the node's phrase. (NEW in 1.6)

-

findnode(phrase, - name)

 -

CON

 -

Find first named node in phrase.

-

findnode(phrase, - num)

 -

CON

 -

Find the numth node in phrase.

-

listnode(node)

 -

CON

 -

Fetch the first node in given node's list.

-

firstnode(phrase)

 -

CON

 -

Fetch the first node in phrase.

-

lastnode(phrase)

 -

CON

 -

Fetch the last node in phrase.

-

makephrase(con, - name)

 -

PHR

 -

Make a phrase in con by creating named - node.

-

addcnode(con, - name)

 -

CON

 -

Add named - node at end of con's phrase.

-

addnode(phrase, - name, num)

 -

CON

 -

Add named - node as numth in phrase.

-

rmnode(con)

 -

NONE

 -

Remove - node from concept's phrase.

-

rmphrase(phrase)

 -

NONE

 -

Remove phrase from its concept.

-

rmcphrase(con)

 -

NONE

 -

Remove a concept's phrase.

-

renamenode(phrase, - name, new)

 -

NONE

 -

Rename phrase's named node to new.

-

renamenode(phrase, - num, new)

 -

NONE

 -

Rename phrase's numth node to new.

- -

See Also

- -

Special - Functions

- -

Formatting - and I/O Functions

- -

Parse - Tree Functions

- -

String - Functions

- -

Spelling - Functions

- - - - diff --git a/Help/helps/Table_of_Math_Functions.htm b/Help/helps/Table_of_Math_Functions.htm deleted file mode 100644 index 286ac1c..0000000 --- a/Help/helps/Table_of_Math_Functions.htm +++ /dev/null @@ -1,162 +0,0 @@ - - - - - - - - - - -Table of Math Functions - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Table of Math Functions

- -

NLP++ now features mathematical functions analogous to those provided - in C/C++.

- - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

logten(num_float)

 -

FLOAT

 -

Get base 10 logarithm of given number.  Analogous - to log10 function in C/C++.

-

 

 -

 

 -

 

- -

See Also

- - - - diff --git a/Help/helps/Table_of_Parse_Tree_Functions.htm b/Help/helps/Table_of_Parse_Tree_Functions.htm deleted file mode 100644 index 90f23ac..0000000 --- a/Help/helps/Table_of_Parse_Tree_Functions.htm +++ /dev/null @@ -1,1006 +0,0 @@ - - - - - - - - - - -Table of Parse Tree Functions - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Table of Parse Tree Functions

- -

The functions used to operate on parse trees have been listed here in - table form for easy reference.  For - examples of these functions, refer to the individual function pages included - in the Parse Tree Functions section.

- -

The return types include:  INT - (integer), PNODE (parse tree node), - STR (string), and VAR - VALUES (variable values). NONE - indicates nothing is returned.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

arraylength(VAR)

 -

INT

 -

Count number of values in given variable or - expr. (Like a call-by-reference)

-

eltnode(elt_num)

 -

PNODE

 -

Fetch the first node that matched the nth - element of a rule.

-

group(start_n,end_n,name_str)

 -

PNODE

 -

Perform a reduction in the @POST region. -  Replaces - the old group action. (2.3.1.9 return the created node.)

-

inputrange(start_n,end_n)

 -

STR

 -

Return a substring of the input text as specified - by start and end offsets.

-

inputrange(start_n,end_n,out_ostr)

 -

BOOL

 -

Print a range of chars from the input text - to an output stream.

-

lasteltnode(elt_num)

 -

PNODE

 -

Fetch the last node that matched the nth element - of a rule.

-

phrasetext()

 -

STR

 -

Fetch the text that matched the right hand - side phrase of a rule. Analogous to $text.

-

phraseraw()

 -

STR

 -

Fetch the raw text that matched the right - hand side phrase of a rule. Analogous to $raw

-

pncopyvars(Pnode1,Pnode2)

-

pncopyvars(Pnode1,aConcept)

-

pncopyvars()

-

pncopyvars(aPnode)

-

pncopyvars(0)

 -

NONE

 -

Copy a node's variables to the suggested node - of a rule match, or to a concept in the KB hierarchy. Must be called from the POST Region and can be considered - a new-style NLP++ action.

-

pndeletechilds(pnode)

 -

NONE

 -

Delete children of given pnode.

-

pninsert(name, pnode, - after_n)

 -

PNODE

 -

Insert nonliteral node with name into parse - tree.  If - after_n == 1, insert after pnode, else if 0 then before pnode.

-

pnmakevar(pnode, var_str, - val)

 -

NONE

 -

Make a variable for a given pnode.

-

pnname(pnode)

 -

STR

 -

Fetch the name of a pnode.

-

pnnext(pnode)

 -

PNODE

 -

Fetch a pnode's right sibling, if any.

-

pnpushval(pnode, var_str, - val)

 -

 

 -

Push a value onto pnode's variable. New in 2.7.3.2

-

pnprev(pnode)

 -

PNODE

 -

Fetch a pnode's left sibling, if any.

-

pnrename(pnode, - str)

 -

STR

 -

Rename given pnode. Returns the interned name.

-

pnreplaceval

 -

NONE

 -

Replace the value of pnode's variable.

-

pnroot()

 -

PNODE

 -

Fetch the root of a parse tree.

-

pnrpushval(pnode, var_str, - val)

 -

 

 -

Push a value onto end of pnode's variable. - New in 2.7.3.2

-

pndown(pnode)

 -

PNODE

 -

Fetch the child of a parse tree node.

-

pnsingletdown(pnode)

 -

PNODE

 -

Fetch the child of a parse tree node, obeying - rules about going down a "singlet chain." I.e., get child only - if no branching and if not going past node with BASE flag set.

-

pnup(pnode)

 -

PNODE

 -

Fetch a pnode's parent, if any. Only leftmost - pnode in a list has a parent.

-

pnvar(pnode, - var_str)

 -

ARRAY

 -

Fetch a pnode's variable value(s).

-

pnvarnames(pnode)

 -

ARRAY

 -

Fetch a node's variable names as an array.

-

varinlist(var_str, - elt_num)

 -

ARRAY

 -

Find var_str variable's value(s) in any pnode - that matched elt_numth element

- -

See Also

- -

Special - Functions

- -

Knowledge - Base Functions

- -

Formatting - and I/O Functions

- -

String - Functions

- -

Spelling - Functions

- - - - diff --git a/Help/helps/Table_of_Special_Functions.htm b/Help/helps/Table_of_Special_Functions.htm deleted file mode 100644 index 6121f47..0000000 --- a/Help/helps/Table_of_Special_Functions.htm +++ /dev/null @@ -1,658 +0,0 @@ - - - - - - - - - - -Table of Special Functions - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Table of Special Functions

- -

NLP++ has a set of special functions.  These - have been listed here in table form for easy reference.  For - examples of these functions, refer to the individual function pages included - in the Special Functions section.  

- -

The return types include:  STR - (string), INT (integer), and BOOL (Boolean). Boolean type is currently - implemented by integer 1 (or non-zero) - for true and 0 - for false.  NONE indicates - nothing is returned.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

batchstart()

 -

BOOL(1,0)

 -

Check if start of a batch analysis.  True - if current file is first in a batch run.

-

debug()

 -

BOOL(1,0)

 -

No op. Convenient breakpoint for debugging - from C++ compiler.

-

exitpass()

 -

NONE

 -

Exit from the current pass file immediately, - without performing rule matching (if any).

-

exittopopup(message_str,
-type_str)

 -

BOOL(1,0)

 -

Exit from current pass, invoking a popup if - within VisualText.

-

fail()

 -

NONE

 -

In CHECK Region, abort the rule that has just - matched.  Noop - elsewhere.  See - succeed().

-

findana

 -

STR

 -

See if named project is already resident in - memory. (New in 1.6)

-

getpopupdata()

 -

STR

 -

Get the data typed in by a call to exittopopup().

-

hitconf(hits_num, total_num, - fudge_num)

 -

INT (0 to 100)

 -

Calculate the keyword density.  Fudge - factor is typically 3 to 20.

-

interactive()

 -

BOOL(1,0)

 -

True if analyzer is being run in an interactive environment, e.g., inside - VisualText.

-

mkdir(dir_str)

 -

BOOL(1,0)

 -

Create a directory.  Operating - system-dependent.

-

permuten(tot_num)

 -

INT ARR

 -

Permute integers 0 to (tot_num - 1) such that array element k will never - be equal to k.

-

succeed()

 -

NONE

 -

In CHECK Region and CODE Region, succeed without executing further code. - See fail().

-

system(str)

 -

BOOL(1,0)

 -

Execute str as operating system command.

-

today()

 -

STR

 -

Format the current date and time as a string.

-

 

 -

 

 -

 

- -

See Also

- -

Knowledge - Base Functions

- -

Formatting - and I/O Functions

- -

Parse - Tree Functions

- -

String - Functions

- -

Spelling - Functions

- - - - diff --git a/Help/helps/Table_of_Spelling_Functions.htm b/Help/helps/Table_of_Spelling_Functions.htm deleted file mode 100644 index b1b9fd2..0000000 --- a/Help/helps/Table_of_Spelling_Functions.htm +++ /dev/null @@ -1,263 +0,0 @@ - - - - - - - - - - -Table of Spelling Functions - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Table of Spelling Functions

- -

The NLP++ functions used to check spelling are listed here in table - form for easy reference.  For - examples of these functions, refer to the individual function pages included - in the Spelling Functions section.  

- -

Return types can be STR (string), - INT (integer) or BOOL - (Boolean).  Boolean - type is currently implemented by integer 1 - (or non-zero) for true and 0 for false.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

spellcandidates(word_str)

 -

STR

 -

Make a list of space separated candidates - for spell-correcting given word.

-

spellcorrect(word_str)

 -

STR

 -

Correct spelling of word. If word is not known, - returns best spelling guess.

-

spellword(word_str)

 -

BOOL

 -

Check if word is a known English word.

-

strspellcandidate(word_str, - list_str)

 -

STR

 -

Select best spell-correct candidate for a - word, given a list of space-separated candidates (as returned by spellcandidates, - e.g.)

-

strspellcompare(str1, - str2)

 -

INT

 -

Measure spelling "distance" between - two given words. The smaller the number, the fewer the corrections to - get from one to the other.

- -

See Also

- -

Special - Functions

- -

Knowledge - Base Functions

- -

Formatting - and I/O Functions

- -

Parse - Tree Functions

- -

String - Functions

- -

 

- - - - diff --git a/Help/helps/Table_of_String_Functions.htm b/Help/helps/Table_of_String_Functions.htm deleted file mode 100644 index c95a993..0000000 --- a/Help/helps/Table_of_String_Functions.htm +++ /dev/null @@ -1,1649 +0,0 @@ - - - - - - - - - - -Table of String Functions - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Table of String Functions

- -

The NLP++ functions used to operate on strings are listed here in table - form for easy reference.  For - examples of these functions, refer to the individual function pages included - in the String Functions section.  

- -

The return types include:  STR - (string), STR ARRAY (string array), - INT (integer), and BOOL - (Boolean).  Boolean - type is currently implemented by integer 1 - (or non-zero) for true and 0 for false. -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

deaccent(str)

 -

STR

 -

Convert string to replace accented characters.

-

dejunk(str)

 -

STR

 -

Convert string to replace all but simple ASCII - chars. New in 2.6.0.3.

-

flt(str)

 -

FLOAT

 -

Convert string or int to float, if possible.

-

levenshtein(str1,str2)

 -

INT

 -

Levenshtein edit-distance between two strings. New in 2.0.2.9

-

num(str)

 -

INT

 -

Convert string to number, if possible. (Also - accepts num arg).

-

split(str, char_str)

 -

STR ARRAY

 -

Split a string using the given char as separator.

-

stem(str)

 -

STR

 -

Compute stem of noun and verb. Return lowercase string.

-

strchar(str, - num)

 -

STR

 -

Index to numth char of string

-

strchr(str, - ch_str)

 -

STR

 -

Find first occurrence of char in string. Return - string headed by char.

-

strchrcount(str, ch_str)

 -

INT

 -

Count the occurrences of a character in a - string.

-

strclean(str)

 -

STR

 -

Remove leading, trailing, and repeat whitespace - separators.

-

strcontains(str1,str2)

 -

BOOL

 -

Check if str1is contained in str2.

-

strcontainsnocase(str1,str2)

 -

BOOL

 -

Check if str1 is contained in str2, ignoring - letter case.

-

strendswith(str, - suffix_str)

 -

BOOL

 -

Does given str ends with given suffix_str

-

strstartswith(str, - suffix_str)

 -

BOOL

 -

Does given str starts with given suffix_str

-

strequal(str1,str2)

 -

BOOL

 -

Check if str1 is identical to str2.

-

strisalpha(str)

 -

BOOL

 -

Check if str is all alphabetic chars.

-

strisdigit(str)

 -

BOOL

 -

Check if str is all numeric chars.

-

strislower

 -

BOOL

 -

Check if first char of string is lowercase

-

strisupper

 -

BOOL

 -

Check if first char of string is uppercase.

-

striscaps

 -

BOOL

 -

Check if the entire string is uppercase.

-

strnotequal(str1,str2)

 -

BOOL

 -

Check if str1 differs from str2.

-

strequalnocase(str1,str2)

 -

BOOL

 -

Check if str1 is identical to str2, ignoring - letter case.

-

strnotequalnocase(str1,str2)

 -

BOOL

 -

Check if str1 differs from str2, ignoring - letter case.

-

strlessthan(str1,str2)

 -

BOOL

 -

Check if str1 is lexically before str2.

-

strgreaterthan(str1,str2)

 -

BOOL

 -

Check if str1 is lexically after str2.

-

strlength(str)

 -

INT

 -

Return length of string.

-

strpiece(str, - start_num, end_num)

 -

STR

 -

Fetch substring of string from start_num to - end_num indexes. Zero-based

-

struniquechars(str)

 -

STR

 -

Find all the unique characters in a string.

-

strrchr(str, - ch_str)

 -

STR

 -

Find last occurrence of char in string. Return - string headed by char

-

strsubst(str, - old_str, new_str)

 -

STR

 -

Replace substring old_str of a string str - with new_str.

-

strtolower(str)

 -

STR

 -

Convert string to lowercase.

-

strtotitle(str)

 -

STR

 -

Convert string to title capitalization.

-

strtoupper(str)

 -

STR

 -

Convert string to uppercase.

-

strtrim(str)

 -

STR

 -

Remove leading and trailing whitespace from - str.

-

strwrap(str, - num)

 -

STR

 -

Break string at specified length.

-

str(num)

 -

STR

 -

Convert numeric value to string. (Also accepts string arg).

-

strescape(str, charsToEscapeStr, escapeStr)

 -

STR

 -

Escape characters in string.

-

strunescape(str, escapedCharsStr, - escapeStr)

 -

STR

 -

Unescape characters in string.

-

suffix(wordStr, suffixStr)

 -

BOOL

 -

True if suffixStr may be the suffix of wordStr.

-

unpackdirs(dir_str)

 -

STR ARRAY

 -

Unpack directory names from a full directory path string.

-

xmlstr(str)

 -

STR

 -

Convert chars to appropriate chars for an XML/HTML file.

-

 

 -

 

 -

 

- -

See Also

- -

Special - Functions

- -

Knowledge - Base Functions

- -

Formatting - and I/O Functions

- -

Parse - Tree Functions

- -

Spelling - Functions

- - - - diff --git a/Help/helps/Table_of_Web_Functions.htm b/Help/helps/Table_of_Web_Functions.htm deleted file mode 100644 index afc49f7..0000000 --- a/Help/helps/Table_of_Web_Functions.htm +++ /dev/null @@ -1,249 +0,0 @@ - - - - - - - - - - -Table_of_Web_Functions - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Table_of_Web_Functions

- -

NLP++ functions are provided to help process web pages.

- -

Note: The User - Project within each analyzer project includes the functions user::analyzefile - and user::analyzebuf.  In - concert with the NLP++ web functions, a web traversal capability can be - built and run within VisualText.  (E.g., - a web crawler or web spider.)  By - integrating web traversal with HTML web page analysis, one may construct - a focused or intelligent web crawler.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

FUNCTION - NAME

 -

RETURNS

 -

DESCRIPTION

-

resolveurl(base_str,embed_str)

 -

STR

 -

Create an absolute URL from embed_str, using - base_str as a base URL.

-

urlbase(url_str)

 -

STR

 -

(Superseded by resolveurl) Fetch a normalized, - partial URL ending in a directory path.  Aids - in resolving relative URLs within a web page. - (New in 1.6)

-

urltofile(url_str,file_str)

 -

NONE

 -

Fetch the web page specified by the url_str - into a file located at file file_str on the local machine. (New - in 1.6)

-

 

 -

 

 -

 

- -

See Also

- -

Functions

- -

 

- - - - diff --git a/Help/helps/Text_Analysis_International/About_Text_Analysis_International.htm b/Help/helps/Text_Analysis_International/About_Text_Analysis_International.htm deleted file mode 100644 index b24b98a..0000000 --- a/Help/helps/Text_Analysis_International/About_Text_Analysis_International.htm +++ /dev/null @@ -1,63 +0,0 @@ - - - - - - - - - - -About Text Analysis International - - - - - - - - - - - - - - - - - - - - - - - - - - - -

About Text Analysis International

- -

The mission of Text Analysis International (TAI) is to revolutionize - the way in which people process text.

- -

TAI was founded in 1998 to develop technologies to analyze text deeply, - accurately, and extensibly. VisualText embodies that technological approach, - enabling a Natural Language Engineering (NLE) methodology for developing - text analyzers. With NLE, the user controls the order of execution, the - contexts to be worked in, and the actions to be taken in every step of - the analysis of a text.

- -

TAI's goals include an ever-greater degree - of automation in the analysis of text, but always with a pragmatic, rather - than theoretical, foundation. TAI is developing standardized bodies of - knowledge, methods for text analysis, and diverse applications as it broadens - its core technology. Rather than championing any particular paradigm, - TAI seeks to exploit all paradigms and methods and to integrate methods - wherever useful. Rather than over-committing to keyword, statistical, - neural network, pattern, recursive grammar, or other approaches, TAI seeks - to encompass all these approaches and more within its product sets.

- - - - diff --git a/Help/helps/Text_Analysis_International/Contact_Information.htm b/Help/helps/Text_Analysis_International/Contact_Information.htm deleted file mode 100644 index d4ebdee..0000000 --- a/Help/helps/Text_Analysis_International/Contact_Information.htm +++ /dev/null @@ -1,282 +0,0 @@ - - - - - - - - - - -Contact Information - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Contact Information

- -

Contacting Text Analysis International, Inc.

- -

Address:

- -

Text Analysis International, Inc.
-1669-2 Hollenbeck Ave. #501
-Sunnyvale, CA 94087

- -

Email:

- - - --- - - - - - - - - - - - - - - - - - - - -
-

Department

 -

Email Address

-

Technical Support

 -

support@textanalysis.com

-

Documentation Feedback

 -

help@textanalysis.com

-

Sales and Marketing

 -

sales@textanalysis.com

-

Corporate

 -

info@textanalysis.com

- -

 

- -

Telephone:

- - - --- - - - - - - - - - - - - - - - -
-

  Department

 -

  Telephone - Number

-

  Corporate, - Sales and Marketing

 -

1-877-235-6259 (USA - toll free)

-

  Technical - Support

 -

 949-376-8507

-

  

 -

  

- -

Please check for updates to this information - at our website,

- -

http://www.textanalysis.com

- -

 

- - - - diff --git a/Help/helps/Text_Tab_Popup.htm b/Help/helps/Text_Tab_Popup.htm deleted file mode 100644 index 94bb2eb..0000000 --- a/Help/helps/Text_Tab_Popup.htm +++ /dev/null @@ -1,1226 +0,0 @@ - - - - - - - - - - -Text Tab Popup - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Text Tab Popup

- -

The Text Tab Popup menu is launched by right clicking in the Text Tab - Window.

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Add

 -

Submenu for adding new or existing text files - to a project, a new folder under a selected folder, or a new top level - folder. (See below.)

-

Attributes

 -

Launches the Attribute - Editor dialog and displays the attributes belonging to the selected - file or folder in the Text Tab. User has option to delete, add or change - names of attributes and values.

-

Move to Top - Level

 -

Moves selected folder to the top level (i.e. - leftmost edge) of the Text Tab.

-

Refresh

 -

Refreshes the Text Tab display.

-

Delete

 -

Deletes selected file or folder in the Text - Tab.

-

Delete Children

 -

Deletes the children of a selected folder - but keeps the folder itself.

-

Delete Logs

 -

Deletes selected log file. Each log file contains - a representation of one pass of the analyzer sequence. When archiving - an analyzer by creating a zip file, there's no need to save the intermediate - parse files since these files can always be regenerated. Delete Logs can - be used to remove these files before archiving.

-

Dictionary

 -

Submenu to perform dictionary related functions. -  (See below.)

-

Edit/View

 -

Displays selected file in the Workspace. Same - effect as double clicking on the file.

-

Rename

 -

Launches a dialog box to rename selected file - or folder.

-

Tools

 -

Submenu for accessing VisualText tools. (See below.)

-

Go - to Web Page

 -

Opens a browser for selected html file.

-

Update Webpage(s)

 -

Updates selected URL(s).

-

View - As Webpage

 -

Displays selected html file as a web page.

-

Text Archive

 -

Submenu for archiving Text Tab files and folders. - (See below.)

-

Run Analyzer

 -

Runs analyzer on selected file. Same function - as selecting the Run - button on the Workspace Toolbar and Run - from the Analyzer - Menu.

-

Help

 -

Launches VisualText Help documentation.

- -

Add Submenu

- -

- - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Existing - Text File

 -

Displays Open dialog to select path and name - of existing file.

-

Folder

 -

Displays prompt to enter new folder name.

-

New Text - File

 -

Displays prompt to enter new file name.

-

Top Level

 -

Submenu for adding a folder or file to the - top level (i.e. leftmost edge) of the Text Tab.

- -

Add Top Level Submenu

- -

- -

Add an existing file, a new file, or a folder to the top level of the - Text Tab. The items in this menu are analogous to the items in the Text Tab Popup > Add menu.

- -

Dictionary Submenu

- -

The Dictionary menu option allows you to perform dictionary related - functions.  

- -

- - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Add Words

 -

Adds the words in a selected file to the dict - hierarchy in the Knowledge Base.  Only - single words can be added to the dictionary.  (If - a word list contains a phrase, only the first word of the phrase is added.) -  File name - and words added to the dict hierarchy are displayed in the Log Window. -  

-

Lookup

 -

Selecting Lookup displays the online dictionaries - specified in the Dictionary Lookup Preferences tab.  Selecting - an option in the Lookup menu starts a search in the online dictionary - source.  See - Dictionary - Lookup for more information on using this tool.

- -

 

- -

Tools Submenu

- -

- -

 

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

ASCII - Table

 -

Displays the ASCII - Table. ASCII Table displays the decimal and hexadecimal numbers and - their corresponding ASCII characters.

-

Character Viewer

 -

Displays the Character - Viewer for the selected file. Character Viewer shows the line number, - cumulative count of total characters seen per line, characters on a line, - and ASCII characters for each line of text in selected file. Under each - line is the hexadecimal value of each character above it.

-

Dictionize

 -

Activates the Dictionize - Tool.  

-

Email Tool

 -

Launches the Email - Tool. Email Tool enables sending email directly from the VisualText - interface.

-

Hex Viewer

 -

Launches the Hex - Viewer.

-

Jumble

 -

Activates the Jumble - Tool.  This - tool jumbles the letters in words in a text file.  This - tool is useful when looking for structural clues in a text, since it removes - distracting content.  Paragraph - and sentence structure are preserved.  

-

Mask

 -

Activates the Mask - Tool.  This - tool takes alphabetic content of a text with strings of the single letter - 'a', but leaves the form, the non-alphabetic structure, including punctuation - and white spaces. This tool is useful when looking for structural clues - in a text, since it removes distracting content.

-

Text File Info

 -

Displays the Text - File Info dialog. This tool provides statistics about the selected - text file, including creation date, character, word and line counts.

- -

 

- -

Text Archive Submenu

- -

Text Archive is used to view - and create archives of the text files in the Text Tab either on remote - servers or on local machines.  Archiving - is used to create quick backups of your work and to facilitate communication - between developers working on the same analyzers.  To - set archiving preferences, select Preferences - under the File Menu and click on the Archiving - tab.

- -

- - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Archive Local

 -

Launches a dialog box to create a local archive - of the current files and folders in the Text Tab hierarchy. The name of - the archive defaults to the name of the current analyzer suffixed with - the current date and time.

-

Archive Remote

 -

Launches a dialog box to create a remote archive - of files the Text Tab. The name of the archive file defaults to the name - of the current analyzer suffixed with the current date and time.  Archive - is created on the remote server.  

-

View Local

 -

Displays local archives in the Text Archive - > View Local dialog. Presents options to delete, rename, upload (send - to server) or load into current analyzer Workspace.  Listings - in archive can be sorted by clicking on column headers.

-

View Remote

 -

Displays server archives in the Text Archive - > View Remote dialog. Presents options to delete, rename, download - (send to local archive) or load into current analyzer Workspace.  Listings - in the archive can be sorted by clicking on column headers.

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/Toolbar_buttons/attr_edit_button.gif b/Help/helps/Toolbar_buttons/attr_edit_button.gif deleted file mode 100644 index 78abca9..0000000 Binary files a/Help/helps/Toolbar_buttons/attr_edit_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/attr_edit_input.gif b/Help/helps/Toolbar_buttons/attr_edit_input.gif deleted file mode 100644 index 0dc1463..0000000 Binary files a/Help/helps/Toolbar_buttons/attr_edit_input.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/attr_edit_search.gif b/Help/helps/Toolbar_buttons/attr_edit_search.gif deleted file mode 100644 index 4560085..0000000 Binary files a/Help/helps/Toolbar_buttons/attr_edit_search.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/browser_back_button.gif b/Help/helps/Toolbar_buttons/browser_back_button.gif deleted file mode 100644 index d3c9f0e..0000000 Binary files a/Help/helps/Toolbar_buttons/browser_back_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/browser_forward.gif b/Help/helps/Toolbar_buttons/browser_forward.gif deleted file mode 100644 index 8c9ab46..0000000 Binary files a/Help/helps/Toolbar_buttons/browser_forward.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/browser_stop_button.gif b/Help/helps/Toolbar_buttons/browser_stop_button.gif deleted file mode 100644 index 0ec4d7b..0000000 Binary files a/Help/helps/Toolbar_buttons/browser_stop_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/copy_button.gif b/Help/helps/Toolbar_buttons/copy_button.gif deleted file mode 100644 index 6aa1004..0000000 Binary files a/Help/helps/Toolbar_buttons/copy_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/cut_button.gif b/Help/helps/Toolbar_buttons/cut_button.gif deleted file mode 100644 index 00cefed..0000000 Binary files a/Help/helps/Toolbar_buttons/cut_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/delete_in_tab_win_button.gif b/Help/helps/Toolbar_buttons/delete_in_tab_win_button.gif deleted file mode 100644 index f267f71..0000000 Binary files a/Help/helps/Toolbar_buttons/delete_in_tab_win_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/find_next_button.gif b/Help/helps/Toolbar_buttons/find_next_button.gif deleted file mode 100644 index da49df3..0000000 Binary files a/Help/helps/Toolbar_buttons/find_next_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/find_previous_button.gif b/Help/helps/Toolbar_buttons/find_previous_button.gif deleted file mode 100644 index 8b81b86..0000000 Binary files a/Help/helps/Toolbar_buttons/find_previous_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/generate_dirty_button.gif b/Help/helps/Toolbar_buttons/generate_dirty_button.gif deleted file mode 100644 index cea6724..0000000 Binary files a/Help/helps/Toolbar_buttons/generate_dirty_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/new.gif b/Help/helps/Toolbar_buttons/new.gif deleted file mode 100644 index 253bd0e..0000000 Binary files a/Help/helps/Toolbar_buttons/new.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/new_browser_button.gif b/Help/helps/Toolbar_buttons/new_browser_button.gif deleted file mode 100644 index 6471492..0000000 Binary files a/Help/helps/Toolbar_buttons/new_browser_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/open_button.gif b/Help/helps/Toolbar_buttons/open_button.gif deleted file mode 100644 index 3079a45..0000000 Binary files a/Help/helps/Toolbar_buttons/open_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/paste_button.gif b/Help/helps/Toolbar_buttons/paste_button.gif deleted file mode 100644 index 9e7351a..0000000 Binary files a/Help/helps/Toolbar_buttons/paste_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/print_button.gif b/Help/helps/Toolbar_buttons/print_button.gif deleted file mode 100644 index 0cbe647..0000000 Binary files a/Help/helps/Toolbar_buttons/print_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/run_next_button.gif b/Help/helps/Toolbar_buttons/run_next_button.gif deleted file mode 100644 index 71425d8..0000000 Binary files a/Help/helps/Toolbar_buttons/run_next_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/save_all_button.gif b/Help/helps/Toolbar_buttons/save_all_button.gif deleted file mode 100644 index f898f46..0000000 Binary files a/Help/helps/Toolbar_buttons/save_all_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/save_url_text_buttonl.gif b/Help/helps/Toolbar_buttons/save_url_text_buttonl.gif deleted file mode 100644 index ac9bcc2..0000000 Binary files a/Help/helps/Toolbar_buttons/save_url_text_buttonl.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/slog_button.gif b/Help/helps/Toolbar_buttons/slog_button.gif deleted file mode 100644 index a8d9bca..0000000 Binary files a/Help/helps/Toolbar_buttons/slog_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/stop_analyzer_button.gif b/Help/helps/Toolbar_buttons/stop_analyzer_button.gif deleted file mode 100644 index 728dbcf..0000000 Binary files a/Help/helps/Toolbar_buttons/stop_analyzer_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/toggle_batch_mode_button.gif b/Help/helps/Toolbar_buttons/toggle_batch_mode_button.gif deleted file mode 100644 index 74deb18..0000000 Binary files a/Help/helps/Toolbar_buttons/toggle_batch_mode_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/toggle_find_win_button.gif b/Help/helps/Toolbar_buttons/toggle_find_win_button.gif deleted file mode 100644 index 400c83c..0000000 Binary files a/Help/helps/Toolbar_buttons/toggle_find_win_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/toggle_log_win_button.gif b/Help/helps/Toolbar_buttons/toggle_log_win_button.gif deleted file mode 100644 index 5440f36..0000000 Binary files a/Help/helps/Toolbar_buttons/toggle_log_win_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/toggle_tab_win_button.gif b/Help/helps/Toolbar_buttons/toggle_tab_win_button.gif deleted file mode 100644 index c7ea673..0000000 Binary files a/Help/helps/Toolbar_buttons/toggle_tab_win_button.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/view_pass_file.gif b/Help/helps/Toolbar_buttons/view_pass_file.gif deleted file mode 100644 index 1b8d6c3..0000000 Binary files a/Help/helps/Toolbar_buttons/view_pass_file.gif and /dev/null differ diff --git a/Help/helps/Toolbar_buttons/view_web_page_button.gif b/Help/helps/Toolbar_buttons/view_web_page_button.gif deleted file mode 100644 index d7c8715..0000000 Binary files a/Help/helps/Toolbar_buttons/view_web_page_button.gif and /dev/null differ diff --git a/Help/helps/Tutorials/Tutorial_Introduction.htm b/Help/helps/Tutorials/Tutorial_Introduction.htm deleted file mode 100644 index c6f8da0..0000000 --- a/Help/helps/Tutorials/Tutorial_Introduction.htm +++ /dev/null @@ -1,224 +0,0 @@ - - - - - - - - - - -Tutorial Introduction - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tutorial Introduction

- -

Tutorial Structure

- -

Six tutorials guide you through the basics of VisualText. You should - follow them in order; each takes about an hour.

- -

The tutorials generally build on each other, so that the analyzer used - in Tutorial 2 is the one that was created in Tutorial 1.  Because - of this, if you change file names, variable names, etc., from what is - presented in a tutorial, later tutorials may not work as described.  However, - the release provides six tutorial analyzer projects, one for each tutorial, - so that you could start Tutorial 2 with the analyzer built by Tutorial - 1, and so on.

- -

Tutorial Analyzers

- -

NOTE: In some versions of - VisualText, the zipped tutorial analyzers are absent or obsolete, so it - is best to start from scratch with Tutorial 1.

- -

The tutorial analyzer projects are in folders named

- -

C:\Program Files\TextAI\VisualText\docs\tutorialN

- -

where N is the tutorial number.  So, for - example, the analyzer created by following Tutorial 1 resides at

- -

C:\Program Files\TextAI\VisualText\docs\tutorial1\myAnalyzer1.zip

- -

(Note: The assumed installation location is C:\Program Files. The exact - path to these files will of course depend on where you installed the software.) - If you do not have decompression software to access these files, you can - download WinZip from a website such as:

- -

 http://www.winzip.com

- -

If you need to look at working code or start a new tutorial in sync - with the documentation, you can copy the tutorial analyzer in question - out of the docs folder in the install directories and into the working - directory.  We've set up the working directory for the tutorial analyzers - to be:

- -

C:\apps\myAnalyzerFolder\

- -

Suppose you wanted to start Tutorial 3 with a clean copy of the analyzer - created in Tutorial 2. You would go to the install directory containing - the zipped tutorial, in this case:

- -

C:\Program Files\TextAI\VisualText\docs\tutorial2

- -

and unzip the myAnalyzer2.zip - file. (When unzipping your files, it is best to keep the unzipped files - in a different directory than the working directory.) When you unzip the - file, it uncompresses to the folder called myAnalyzer.

- -

Place the myAnalyzer folder - into the working directory:

- -

C:\apps\myAnalyzerFolder\

- -

Since the first tutorial starts from scratch, you cannot start out Tutorial - 1 in this way.

- -

Additional Notes

- -

Here are a few extra notes before starting the tutorials.

- - - -

C:\Program Files\TextAI\VisualText\Help.chm

- - - -

Tutorial Description

- -

Six tutorials guide you through the basics of creating an analyzer. - Along the way, you will also be introduced to various tools and aspects - of the VisualText work environment. Below is a description of each tutorial. -  You may - find it useful to scan through the tutorials to familiarize yourself with - the content before actually starting the tutorials.

- - - -

 

- -

 

- - - - diff --git a/Help/helps/Tutorials/tutorial1/tutorial.html b/Help/helps/Tutorials/tutorial1/tutorial.html deleted file mode 100644 index 13545be..0000000 --- a/Help/helps/Tutorials/tutorial1/tutorial.html +++ /dev/null @@ -1,1002 +0,0 @@ - - - - - - - - - - -tutorial - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tutorial 1 - Creating a New Analyzer

- -

In Tutorial 1, you create and run a simple text analyzer. You give the - analyzer a name and a location, select an initial template, and save the - analyzer. Then you exit VisualText and reload the analyzer you created. - We also introduce the Tab Window and the Log Window portions of the interface - and explain how to create an input file. Finally, you run the analyzer - and view the parse tree, which holds the patterns matched in a text.

- -

Startup

- -

Launch VisualText by double clicking on the VisualText - icon on your desktop (or from Start button > TextAI > VisualText).  - You should see the following screen:

- -

- -

If your toolbar area does not look like the above, right-click - the mouse in the toolbar area, then enable all the toolbars.  We - recommend rearranging them as shown.

- -

Take a moment to familiarize yourself with the interface, which divides - into three general areas: toolbars, windows, and a workspace.

- -

The various toolbars give you access to commonly used functions. - In Tutorial 1 you will mostly be working with the Main Toolbar. (See the - Interface - Overview section for the naming of the VisualText components).

- -

At the bottom of the Main Window, you'll see the Log Window and the - Find Window (VisualText 2 separates these into tabs). This is where you - will be creating and monitoring your analyzer. You will learn more about - the Main Window later in this tutorial. The Workspace is where you can - view and manipulate files and tools for building and running text analyzers. - For more details on the function of all of the windows, toolbars and menus, - see the Interface - section of this manual.

- -

Create Analyzer

- -

Now, let's create a new analyzer.

- -

Select File > New - Analyzer from the Main Menu. Note the title bar "VisualText - - (no analyzer currently loaded)" below:

- -

- -

A New Analyzer dialog box appears.

- -

- -

Give the analyzer the name myAnalyzer.

- -

For location, give the following path C:\apps\myAnalyzerFolder

- -

For template, choose - Bare.

- -

A template is a starter or skeletal analyzer that you can copy and modify. - The Bare template analyzer is a minimal analyzer and with a single pass called - tokenize. The tokenize pass is discussed later in this tutorial.

- -

Click OK. You now see a VisualText splash - screen appear on the screen.

- -

After a few seconds, the VisualText title - bar reads "VisualText - myAnalyzer."

- -

When new analyzers are created, VisualText makes a directory structure - in the working directory. For this analyzer, VisualText created the following - directory structure in c:\apps:

- -

- -

This folder structure houses all the information - needed to run the analyzer, including folders for input and output, a - folder for the knowledge base, etc. The analyzer you just created, myAnalyzer, - refers to this entire directory structure. The set of related files and - directories under myAnalyzer is called a project - in VisualText and other development environments. For a detailed description - of an analyzer project, see Structure of - Analyzer Projects.

- -

One important file automatically generated - in the myAnalyzer folder is myAnalyzer.ana, - which organizes your project. When you load an analyzer, you do so by - opening the myAnalyzer.ana file.

- -

Exit and Load - Analyzer

- -

Now we exit the program.  (Note - that the Help should stay as-is, as a separate program.)  Launch - VisualText again to continue with Tutorial 1.

- -

Select File - > Exit from the Main Menu.

- -

To load the analyzer you - just created, invoke VisualText and select File - > Open Analyzer.

- -

- -

Use the Open dialog to navigate - to C:\apps\myAnalyzerFolder\myAnalyzer:

- -

- -

Select myAnalyzer.ana, - then select Open.

- -

This step loads the analyzer.

- -

Once an analyzer is open, be careful to - distinguish between File > Open... - which opens a file into the Workspace area, and File - > Open Analyzer, which enables you to close the current analyzer - project and browse to an .ana file of another project.

- -

Main Windows

- -

Three large windows comprise the analyzer - work area: the Log Window, located in the lower left corner of the VisualText - interface, the Tab Window located in the upper left portion of the interface - and the Find Window located in - the lower right corner. You can resize, dock and dismiss these windows - as you choose.  If you close a window and want to reopen it, use - the toggle buttons on the main menu toolbar:

- -

- -

The leftmost button toggles the Tab Window.

- -

The left hammer toggles the Log Window while - the right hammer toggles the Find Window.

- -

Log Window

- -

The Log Window gives you some basic information - about an analyzer when you open it. It also displays information about - rules as they are processed and applied to text you are analyzing.

- -

- -

Here it displays the time it took to load - the minimal analyzer (myAnalyzer) and its associated knowledge base (KB).

- -

When an error occurs or other information - needs to be reported to the user, it is displayed in the Log Window. For - typical messages displayed in the Log Window, refer to Log - Window Messages.

- -

Tab Window

- -

There are three tabs in the Tab Window: - Gram, Ana and Text.

- -

- - - -

VisualText is an integrated development - environment (IDE) for creating text analyzers. It gives you access to - all the machinery you need to create an analyzer in a single framework. - The Gram, Text and Ana tabs are essential parts of the machinery. Generally - speaking, you view and manage the input and output files through the Text - Tab; the text analyzer sequence, passes, and manually-edited rules in - the Ana Tab; and the automatically-generated rules in the Gram Tab.

- -

The Tab Window can be resized by dragging - the right edge of the window towards the left. When resizing the window - hides one of the tabs, you can use the toggle arrows to find it.

- -

Familiarize yourself with - each of the tabs in the Tab Window, by selecting the tab (Ana, Gram or - Text) and with the cursor in the window (not directly on the tab), right - clicking to bring up the popup menu.

- -

You'll be working with a few of these popup - menu items in this tutorial. But for a detailed list of the various functions - available in the Tab Window, see Popup Menus - in the Interface section of this guide.

- -

Click on the Gram - Tab and then on the Text Tab.

- -

Notice that nothing is contained in either - of these tabs. By default, these two tabs are left empty. As you add input - texts, analyze them, and sample information from the texts, you will create - data in these two areas.

- -

Now click on the Ana - Tab.

- -

Notice that it is not empty. - The Ana Tab comes with - a built-in default pass, the tokenize - pass.  A pass is a discrete step in the text analyzer, with its own - pass algorithm. The main pattern-matching - pass type in VisualText (called pat) - uses a pass file, which consists - of NLP++ code and rules. NLP++ is a general purpose programming language - that is integrated with both a grammar and a hierarchical knowledge base - management system (KBMS). This tight, three-way coupling is ideal for - developing text analysis applications. NLP++ code can manipulate the parse - tree and attach information to it, as well as access and modify knowledge - used during text analysis.

- -

The tokenize - pass algorithm is a system pass and does not have an associated pass file. - Tokenize executes first, traversing - the input text to convert it to a structure called a parse - tree.  As - you add new passes to the Ana Tab, they become part of the stepwise sequence - executed when the text analyzer runs on an input text. Each pass typically - updates the parse tree with patterns that it identifies. As we'll see, - NLP++ enables a pass to also manipulate the information in the parse tree - and the KB.

- -

Find Window

- -

The last of the Main Windows is the Find Window. It is used to display - search results when running a search on texts. Since we don't have any - text to search yet, we won't be able to use this window now.

- -

Add Text File

- -

Now we need some text to analyze.  - In VisualText you have the option to use existing text files or to create - new text files. For this tutorial, we'll create a text file from scratch.

- -

Click the Text - Tab.

- -

Then, with the cursor in the white area - of the Text Tab, right click in the window to bring up the Text Tab Popup - Menu.

- -

Select Add - > New Text File:

- -

- -

Name the file eg1 - (and don't include an ending such as .txt).

- -

Click OK.

- -

Notice that a file called eg1.txt - has been added to the Workspace and that the file has also been added - in the Text Tab. The default extension .txt is automatically added to - the file name. "input" - after the file name indicates the file is located in the input folder.

- -

In addition to text files, you can use binary - or formatted text files such as html as input.

- -

Your Tab Window and Workspace should look - like this:

- -

- -

Now, type the following text - in the eg1.txt file:

- -

The quick brown fox jumped over the lazy yellow dog.
-John gave the book to Mary.
-Susan kissed the little baby's cheek.

- -

An asterisk * - is added to the top of the file indicating the file has changes which - have not been saved.

- -

To save the file, put the - cursor in the body of the text file and right click to launch the Text - File Popup Menu:

- -

- -

Select Save.

- -

You can also save files in the Workspace - with the Save button located on the Main Toolbar.

- -

The file is written to the following directory:

- -

C:\apps\myAnalyzerFolder\myAnalyzer\input

- -

When you are supplying your own files, you - would copy them into this same directory.

- -

If you do add existing text files to the - directory, it will be necessary to refresh the Text Tab. You can refresh - by putting your cursor inside the Text Tab Window, right clicking and - selecting Refresh from the Text - Tab Popup Menu. Note that in this case you will get a dire-looking message - saying that "all knowledge associated with text files" will - be lost if you refresh. You can safely ignore that message and continue - to refresh. (It refers to attribute knowledge that one could add to text-file - concepts in the knowledge base.)

- -

Run Analyzer

- -

Now that we have some text, we are ready - to run the analyzer. Recall that the analyzer at this point is made up - of one default rule pass: tokenize.

- -

So that you can see each step in the analyzer - sequence, select Generate Logs - button on the Workspace Toolbar. - (When the button is depressed, it appears indented.) When you run an analyzer - with Generate Logs on, log files will be created so that you can get detailed - information about each step of the analyzer sequence, which can help you - debug your analyzer. You'll learn more about generating logs in later - tutorials.

- -

Select eg1.txt - in the Text Tab.

- -

Click the Run - button on the Workspace Toolbar to "run" the analyzer.

- -

Notice that information about the application - of rule passes is displayed in the Log Window. - Here it lets you know the name of the text being analyzed and the - date the file is processed.

- -

We recommend keeping the Verbose - Mode button TURNED OFF, as it has no useful function - at present. It is always a good idea to check the Log Window when creating - your analyzer since this is where error messages will show up.

- -

View Parse - Tree

- -

Now that we applied the tokenize pass to - the input file,  we can look at one output of the analysis, called - the parse tree.

- -

When the analyzer is run on a text, it first - converts the sequence of characters to "tokens" or "lexemes" - and places them under a root.  This - constitutes the initial parse tree.  Subsequent - passes modify the parse tree as they match new patterns within it.  A - parse tree is similar to the sentence diagramming taught in elementary - school where you break a sentence down into parts such as nouns, verbs, - noun phrases, verb phrases, subjects, objects, etc.  Each - pass in a VisualText analyzer takes as input the accumulated parse tree - built by all the prior passes. 

- -

There'll be a lot more information about - passes and trees later on in these tutorials, but for now, let's take - a look at the parse tree you just built for the eg1.txt file.

- -

Select eg1.txt - in the Workspace.

- -

Right click to bring up the - Text File - Popup Menu.

- -

Select View - > Parse Tree:

- -

- -

VisualText displays a parse tree for the - text file in a new window.

- -

- -

Notice the title of the window, 'eg1.txt: - _ROOT [1 tokenize 0-119]' and the graphic that precedes the title, the - parse tree graphic. The title lets you know that this is the parse tree - for the file eg1.txt and it corresponds to the first pass in our analyzer - sequence (1 tokenize) and also that 119 characters were tokenized. If - you don't have 119 this just means that there may have been some extra - spaces or carriage returns in your file.

- -

The parse tree consists of nodes. - Each line in the parse tree represents a single node. Our parse tree starts - with the node - _ROOT. The root of the parse tree covers the entire input text.

- -

Nodes 'contain' text and other nodes and - are said to be either terminal or non-terminal. A node is non-terminal - if it dominates other nodes and terminal if it does not. As you can see, - there is only one non-terminal node in our parse tree, _ROOT. All of the - elements in our sentences, right down to the punctuation are dominated - by _ROOT. The other nodes in our parse tree are terminal nodes. The terminal - nodes are like the 'leaves' or end point of the tree and do not dominate - anything else.

- -

Each terminal node contains the following: - the node name, the token type and the text from the file. The first terminal - node under ROOT is 'The::[alpha] The'. This means that the input text - has as its first terminal node 'The' and that the text 'The' is a sequence - of alpha characters. The next terminal node in the file is \_ which means - a white space, ie the white space after the word 'The'. (More on the literal - use of \ will be discussed in later tutorials.)

- -

As you may have found, you can distinguish - between terminal and non-terminal nodes by the symbols used in the parse - tree. The non-terminal nodes are indicated by an arrow and the terminal - nodes are indicated by a thin, rectangular bar. A non-terminal node will - be blue when selected and grey otherwise. A terminal node is red unless - it is selected at which time it will turn blue. You can collapse and expand - nodes by clicking on the + and - symbols.

- -

If you want to see the details of the analyzer - run, you can look at the log files located under the eg1.txt file in the - Text Tab Window. 

- -

Click the "X" - in the upper right-hand corner of the eg1.txt and parse tree files to - close them.

- -

Exit the analyzer by selecting - File > Exit.

- -

(In some versions) answer - No at the confirmation prompt - for saving the knowledge base.

- -

Summary

- -

In - this tutorial, we created an analyzer, gave it a name and a location, - selected a template, saved it, exited and reloaded it. We introduced the - Tab Window with the Text Tab, Ana Tab and Gram Tab. We looked at the Log - Window, then gave our analyzer some text to analyze. Finally, we analyzed - the text using the default system pass tokenize and viewed the resultant - parse tree.

- - - - diff --git a/Help/helps/Tutorials/tutorial2/tutorial2.html b/Help/helps/Tutorials/tutorial2/tutorial2.html deleted file mode 100644 index d14313a..0000000 --- a/Help/helps/Tutorials/tutorial2/tutorial2.html +++ /dev/null @@ -1,1480 +0,0 @@ - - - - - - - - - - -tutorial2 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tutorial 2 - Rules From Sample Data

- -

In Tutorial 1, you saw an analyzer with a single pass, the tokenize - pass. Recall that the tokenize pass is nothing more than a set of rules - and instructions that tell the analyzer what to do. In Tutorial 2, we - will learn more about generating rules in VisualText to create passes - in the analyzer sequence.

- -

VisualText supports two basic ways of generating rules in pass files: - hand-built and automatic.  In this tutorial, we introduce automatic - rule generation. Automatic rule generation results in the automatic creation - of passes in the analyzer sequence. In order to do this, we will also - learn about adding and sampling texts. Any text analyzer can be easily - enhanced by adding samples as shown in this tutorial. Here are the steps - for setting up this capability:

- - - -

After we have VisualText generate our rules, we will analyze our text - and inspect the results.

- -

Let's start with the analyzer we finished in Tutorial 1.

- -

Double click on the VisualText icon on your desktop.

- -

Load the analyzer you built in Tutorial 1, this time - by selecting File > Recent Analyzers - and selecting

- -

C:\apps\myanalyzerfolder\myanalyzer

- -

If you want to start with the ready-made analyzer in the docs folder, - unzip the contents of:

- -

C:\Program Files\TextAI\VisualText\docs\tutorial1\myAnalyzer1.zip

- -

and place the myAnalyzer folder - in the working directory:

- -

C:\apps\myAnalyzerFolder

- -

Create a Stub

- -

Our first step is to create a stub. - A stub is a placeholder for a - sequence of passes that will be generated automatically. You are setting - aside an area for the automatic rule generator to create the rules for - you.

- -

In the Tab - Window, click on the Ana Tab - and select the tokenize - pass. (The stub will be inserted after the pass you select - in this step.)

- -

Now go the Gram Tab, - move cursor into the window and right mouse click to bring up the Gram - Tab Popup Menu.

- -

Select Add > Stub.

- -

This will create a new stub in the sequence of passes after - tokenize. (You have to select the tokenize - pass in the Ana Tab to make sure the stub will be inserted in the correct - place.)

- -

- -

In the dialog box, name the stub words.

- -

- -

Click OK.

- -

The stub words has been created - in the Gram Tab.

- -

When a stub is selected in the Gram - Tab, it will appear in blue. Stubs which are not selected will - appear in green.

- -

- -

When you create analyzers, you can choose whatever name you like for - a stub. We called our stub words - because we are going to take the words - from the sentences in our eg1.txt file and use them as samples to generate - our rules.

- -

Add Rule Concepts

- -

Now that we have a words stub, - our next step is to create a set of rule - concepts under words. A rule concept is a concept that "owns" a set - of related samples, from which the rule generator will generalize, merge, - and generate rules.

- -

The rule - concepts we will create correspond to the classic syntactic categories - of noun, verb, determiner, adjective, etc.

- -

(Make sure the words concept selected.) Right click In - the Gram Tab, to bring up the - Gram Tab Popup Menu. 

- -

Select Add > Rule.

- -

- -

In the Dialog box, add the rule names det, - adj, noun, - verb, prep - like the image below:

- -

- -

Click OK.

- -

Your Gram Tab should look like this:

- -

- -

When you create analyzers, you can create any type of rule concept you - want or need.  For example, if you have text that contains urls or - email addresses and you want your analyzer to recognize them, you could - create a rule concept for urls and another one for email addresses. Being - able to create rule concepts as they are needed is one of the nice features - of VisualText.

- -

Add Folder Concept

- -

We only have - five rule concepts in this example. But as you begin to create analyzers, - your analyzer may require many more rules.

- -

In order to organize - rule concepts, VisualText uses a special concept called a folder - concept. Folder concepts - allow you to organize collections of rule concepts.

- -

Our rule concepts, - det, adj, noun, verb and prep can be classified as parts of speech. We'll - create a folder to organize our rule concepts.

- -

In - the Gram Tab, - click on the words - stub to highlight it.

- -

Right-click - to bring up the Gram Tab Popup Menu and select Add > Folder:

- -

- -

Name the folder pos - (for part of speech).

- -

- -

Click OK.

- -

You - should see the following structure:

- -

- -

Select each rule concept - one by one and drag it into the pos - folder.  (Easiest - to drag the bottom rule concept first, starting with prep.)

- -

The rule concepts should be under the pos - folder (under meaning below - and indented with respect to pos) - as in the image below.

- -

- -

Use the yellow up and down arrows on the Tab Toolbar - to move the concepts in the pos folder so that they look like the image - here.

- -

You can save your work with File - > Save Analyzer...

- -

...and answer Yes at - the confirmation prompt.

- -

Also Yes to save - the knowledge base, as the Gram Tab information resides in the knowledge - base.

- -

Sample the Data

- -

Now we are ready to start populating our concepts with samples from - eg1.txt. By "populate" we mean "fill in" or provide - each of the rule concepts with example text ("samples") of what - each rule concept is. The samples you add are called sample - concepts.

- -

CAVEAT!  It - is a good time to make sure your text file doesn't have any errors in - it. ONCE YOU START SAMPLING - DATA FROM A FILE, DO NOT EDIT THAT FILE.  The - Gram Tab depends on the offsets of text samples in files, so that it can - use the context of the samples to build proper rules.  Even - adding a blank line will invalidate the samples following it.  (Adding - material to the very end of such a file, or editing so as to preserve - exactly the same number of characters, are doable if needed.)  In - a "real" analyzer, we recommend putting files to be sampled - in specially designated folders and making them READ-ONLY.

- -

Once you have created eg1.txt, CLOSE - THE FILE.  VisualText - strips linefeed characters from text files when it reads them, and this - needs to happen before you start - highlighting samples.

- -

In the Text Tab, double click on the eg1.txt - file to open it in the Workspace. Make sure everything is correct in the - file.

- -

- -

First we'll add samples of nouns - to populate the noun rule concept.

- -

In the Gram Tab, - select the noun rule concept. - Notice that the arrow turns blue. 

- -

The first noun in our text file, is fox.

- -

Highlight - fox - and right click to bring up the popup menu.

- -

Notice how the noun rule concept - has been added to the bottom of the popup menu. It appears - in the popup menu because noun - has been selected in the Gram Tab - Window.

- -

Select noun.

- -

- -

Look at the noun rule concept in the Gram - Tab. The - noun fox has been added - as a sample concept under - the noun rule concept. Sample concepts are indicated by the little square - icon.

- -

- -

To add samples to the other rule concepts in the pos - folder, simply select the rule concept in the Gram - Tab and follow the same procedure.

- -

There is an alternative to selecting samples by selecting individual - rule concepts in the Gram Tab. - By selecting the pos folder instead, - all of the rule concepts in the pos folder will be listed when you right - click after highlighting a word in the text.

- -

Select the pos - folder in the Gram Tab.

- -

Select dog in - the text and right click to bring up the menu.

- -

Notice that all of the rule concepts, det, - adj, noun, - verb and prep - appear in the popup menu. It may be easier for you to sample text this - way.

- -

- -

Select noun.

- -

Continue adding samples to the rule concepts until all - the words in eg1.txt have been added to the appropriate rule concept.

- -

- - - - - - -
-

Note:  Make - sure you select baby and not baby's.

- -

You have just created a sample hierarchy. - (This is also referred to as the Gram - hierarchy since you can view it in the Gram Tab.) The sample hierarchy - organizes our rule concept samples.

- -

It is a good idea to get in the habit of saving your analyzer often. - Now in particular, since the creation of the words - stub has modified the analyzer sequence, is a good time to save the analyzer.

- -

To save the analyzer, select File - > Save Analyzer.  (And/or - KB > Save KB, as available.)

- -

Answer Yes at - the confirmation prompt.

- -

Since data has been added to the concepts in the Gram Tab, our Knowledge - Base (KB) has changed. When you save the analyzer, you are also saving - the Knowledge Base.

- -

Set Concept Properties

- -

In VisualText you can control the way rules are generated. You do this - by setting certain properties for rule concepts in the Gram Tab. 

- -

In this tutorial, we are concerned with setting only one of the properties, - the Closed attribute. The Closed attribute specifies that automatically - generated rules will consist of a closed set of words, as collected from - the samples you, the user, provided. By setting this attribute you are - controlling the way rules get generalized.

- -

(For more information about Gram concept properties, see Setting - Rule Generation Properties.)

- -

Select the words - stub in the Gram Tab.

- -

Right click to bring up the Gram - Tab Popup Menu.

- -

Choose Properties.

- -

- -

In the Gram Concept Properties dialog, uncheck - the Inherited box for the - Closed attribute.

- -

Now place a check in the On/Off - box.

- -

- -

Click Done when - you are finished.

- -

 

- -

Generate Rules

- -

Now let's have VisualText automatically generate some rules for us from - the sampled data.

- -

If we take a look at our analyzer sequence in the Ana - Tab, you will see the following:

- -

- -

Notice that our stub concept words - has a start, stub_words and an - end point, end_words. This is - called the stub region. We would - call this stub region the words stub - region. Remember that we created the stub region because we needed - a place for the automatically generated passes to be inserted. Once we - generate rules, the stub region will get filled in.

- -

From the Analyzer Menu, select Generate - Rules > All.

- -

- -

Now look at the stub region.

- -

Additional passes appear in our analyzer sequence. Each of the new passes - contains rules generated from the rule concepts we created earlier in - this tutorial. VisualText did this by taking the samples we provided in - each of the rule concepts.

- -

- -

Notice that the Log Window has also - changed.  It says Generating rules: - start then Generating rules: done. - This window is just telling you that - VisualText has started and completed the task of generating rules. If - there had been problems generating the rules, error messages would have - been displayed here.

- -

Let's take a - look at the rules VisualText created.

- -

Select the noun - pass.

- -

The link in front of 5 noun - turns blue and the text gets a grey box indicating that it is selected.

- -

Double click on the noun - pass to open the pass file.

- -

The noun pass file, automatically generated from the samples you added - to the Gram Tab, is displayed in the Workspace. This pass file (also called - rule file since it contains rules) - is stored in the spec folder of the myAnalyzer directory. The name of - the pass file is noun.pat. You'll - learn more about pass file names in Tutorial 3.

- -

- -

The form of the pass file is discussed in a later tutorial, too. But - generally what the above rule "says" is that every time the - analyzer runs across one of the words listed in parentheses (e.g. "dog", - "John", etc.), these words should be classified as "noun". - In Tutorial 3, you'll see how this automatically-generated rule can be - helpful.

- -

Run Analyzer & Inspect Parse Tree

- -

Now that we have some extra passes in our analyzer sequence, let's run - the analyzer to see how our parse tree changes.

- -

Double click eg1.txt - in the Text Tab to bring the file - to the front of the Workspace.

- - - - - - -
-

Note:  This - step is optional if eg1.txt appears in the current text panel within the - Tab Toolbar. This will let you know what text will be analyzed.

- -

Start the analyzer by either clicking on the Run button or by - selecting Analyzer > Run - from the Main Menu.

- -

- -

Now that we have run the analyzer, let's view the resulting parse tree.

- -

Click on the View Parse - Tree button. or select - the - eg1.txt in the Workspace, right - mouse click, and select View > Parse - Tree from the - Text File Popup Menu.

- -

Your parse tree should look like this:

- -

- - - - - - -
-

Note: - If you ran the analyzer with Generate Logs on, the title bar in your parse - tree may not match the image.  The - title bar will show the name of the selected pass in the Ana Tab.

- -

You can expand nodes labeled with a + by selecting them.  - Note that node labels (_noun, - _adj, etc.) have been added to - most lexical items in the file.  All non-terminal node labels start - with an underscore ('_'). 'fired' indicates that a rule matched and 'blt' - indicates that a node was built.

- -

Close the open files in the Workspace by selecting the - X in the upper right-hand corner - of the file.

- -

Exit the analyzer.

- -

From the Main - Menu, select File > Exit.

- -

Answer Yes at - the confirmation prompt for both - saving the analyzer and the knowledge base.

- -

Summary

- -

In this tutorial, we added sample data to - VisualText in order to generate rules automatically.  We created - a stub concept named words, some rule concepts for the classical syntactic - categories, and a folder concept to organize the rule concepts. We sampled text data for these - concepts and generated - rules. Finally, we ran - the analyzer against a text file and inspected the resulting parse tree. Adding samples - and automatically generating rules demonstrate a simple way to maintain - and enhance text analyzers with VisualText.  In Tutorial 3, we'll - show you how to create simple hand-built rules to parse the sentences - in the text we worked with in tutorials 1 and 2.
- 
- 
- 
- 
- 
- 
- 
- 

- - - - diff --git a/Help/helps/Tutorials/tutorial3/tutorial3.html b/Help/helps/Tutorials/tutorial3/tutorial3.html deleted file mode 100644 index 90edb76..0000000 --- a/Help/helps/Tutorials/tutorial3/tutorial3.html +++ /dev/null @@ -1,1794 +0,0 @@ - - - - - - - - - - -tutorial3 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tutorial 3 - Rules by Hand

- -

In Tutorial 2, we saw how to automatically generate rules by highlighting - samples in our text files. In this tutorial, we will use the other method - of generating rules, and that is to build them by hand. An in-depth text - analyzer will typically include rules made by both of these methods.

- -

In the last tutorial, we looked at the noun pass file that was automatically - generated by VisualText. In this tutorial, we will work with pass files - like this one but this time we will write the rules ourselves. In particular, - we will write rules for finding noun phrases (NP), verb phrases (VP), - prepositional phrases (PP), and possessive phrases (POSS).   

- -

We'll use the analyzer built in Tutorial 2.

- -

Launch VisualText.

- -

From the Main Menu, - select File > Recent Analyzers - and choose:

- -

C:\apps\myAnalyzerFolder\myAnalyzer

- -

If you want to start with the ready-made analyzer in the docs folder, - unzip the contents of:

- -

C:\Program Files\TextAI\VisualText\docs\tutorial2\myAnalyzer2.zip

- -

and place the myAnalyzer folder - in the working directory:

- -

C:\apps\myAnalyzerFolder

- -

Create New Pass in Analyzer Sequence

- -

Each 'bead' on the analyzer sequence is a pass. For the typical rule-based - pass, the analyzer engine traverses the input text (actually, the parse - tree representing the text) once before proceeding to the next pass.

- -

Click on the Ana Tab - and take a look at the pass sequence.

- -

Our analyzer right now has eight passes, namely tokenize, - the automatically generated passes det, - adj, noun, - verb and prep, - and the two stub region markers.

- -

Now we'll add a new pass following the words - stub region. In this case, we want to add our next pass to the end of - the current analyzer sequence.

- -

Select the end-of-stub marker end_words - and right-click to display the Ana - Tab Popup Menu. 

- -

Select Add > New - Pass.

- -

- -

In the Pass Properties - dialog, name the pass NP (for - 'noun phrase').

- -

- -

Leave Type: - as the default value, Pattern. (This is the pat - pass algorithm type mentioned previously.) Leave the Set Active Flag information - as is. (The Set Active Flag box is used to activate or inactivate a pass - in the analyzer sequence.)

- -

Click OK.

- -

VisualText creates a new pass in the analyzer sequence after end_words, - named NP.

- -

- -

Now, double click the NP - pass to open it in the Workspace.

- -

We are now ready to start hand building some rules in our pass file.

- -

 

- -

Edit Default Rules

- -

VisualText creates a pass file with some default structure to get you - started. This is just a starter template and it can be edited to create - the code and rules you need.

- -


 

- -

Now is a good time to review some definitions and learn a little more - about building rules.

- -

A pass - is one step in the analyzer sequence. It corresponds to one 'bead' in - the Ana Tab.  There are several - types of passes. The most prevalent type of pass is called pat (for 'pattern'), or a rule-based - pass. Each pat pass has an associated pass - file (also called a rule file). - The pass file that corresponds - to the noun pass for example, is called noun.pat. - When you hand-build rules, you are writing NLP++ (code and rules) within - the pass file itself. For a detailed description of the pass file syntax, - see About - Pass Files in the NLP++ section of this Help. Later - tutorials will also go into more detail.

- -

Let's take a closer look at our default NP pass file.

- -

Anything to the right of the pound sign, - #, is a comment and is not part of - the actual rule. The file header uses pound-sign comments to name and - describe the current pass.

- -

The actual rule file begins at the @CODE - marker.  This - is the beginning of a region called the CODE Region.  The - CODE Region optionally ends with an @@CODE marker. Code placed here executes - prior to any rule matching in the current pass. A pass may consist entirely - of code written within the CODE Region. In these tutorials, we won't be - editing the CODE Region.

- -

The @NODES marker tells the analyzer which - nodes to apply the subsequent rules to.  The line

- -

@NODES _ROOT

- -

means that the rules will be applied only to nodes 'dominated by' the - non-terminal node _ROOT.  - Recall from Tutorial 1 that the nodes dominated by _ROOT - include all of the words and characters in the eg1.txt file. You can see - the list of nodes that will be traversed by rules of this file by viewing - the parse tree for eg1.txt. This is what our parse tree looked like at - the end of Tutorial 1.

- -

- -

Next is the @RULES marker, which tells the - analyzer to expect a new RULES Region. - The RULES Region is where you write your rules.  If - the @RULES marker is not preceded by an action - region (e.g., POST), - then it implicitly tells the analyzer to use a default action for the - rules of this region.

- -

- -

A rule has the form 'X <-- Y Z ... @@', which means: 'nodes that - match the phrase Y Z ... - will get a new dominating node named X.' In the default rule, _xNIL is the suggested element and _xNIL (on the right hand side) is a phrase of - only one element. _xNIL - is a special rule element that - represents the empty token, or 'nothing', by convention. Another special - token, _xALPHA, matches any alphabetic token - in the parse tree.

- -

The @@ symbol - marks the end of a rule.

- -

Hand-edit the @RULES Region - of the NP pass file as below:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES _ROOT

- -

 

- -

@RULES

- -

_NP <-

- -

_det

- -

_xWHITE

- -

_adj

- -

_xWHITE

- -

_adj

- -

_xWHITE

- -

_noun

- -

@@

- - - - - - -
-

Note: You can copy and paste from the Help - documentation into your analyzer.  However, - note that formatting and colors may be lost when you paste the copied - text into the analyzer pass file.  Saving - the file after you have copied the text, closing the pass file and then - opening the pass file again can restore some of the formatting.  

- -

Make sure you use a lowercase x - when writing _xWHITE in your rules.

- -

Save the NP pass - file. (Though running the analyzer will automatically save the pass file, - it is a good idea to get into the habit of saving edited pass and text - files often.)

- -

The rule we've added says that a phrase consisting of a determiner followed - by two adjectives followed by a noun, with intervening blank spaces, will - suggest or reduce - to a noun phrase (NP). Such a match is searched for in the nodes directly - under the _ROOT node.

- -

The initial underscore character "_" is an NLP++ convention - marking nonterminal or internal - nodes, as opposed to terminal - (also called literal or leaf) nodes, which represent words, - numbers, and punctuation.

- -

You have just written a rule in VisualText standard rule format, - in which each element is on its own line. The Pass File Popup menu, - which provides handy tools for editing rules, assumes that rules are in - standard format. See Standard - Rule Format for more information.

- -

We have now finished creating our first hand-built pass file.

- -

Run Analyzer

- -

Now we are ready to run the analyzer again to see what kind of parse - tree we get after the addition of our NP pass file. For this tutorial - deselect the toggles for Generate Logs , Verbose Mode , and Page Mode .

- -

To run the analyzer sequence on the text,

- -

Select - eg1.txt in the Text - Tab and click the Run - button.
 

- -

Inspect Log Window

- -

When the analyzer is run, a report is generated in the Log - Window. The left side of the report will contain the name of the - analyzed file and the right side of the report will show the passes and - the total time needed to process the pass.

- -

View Parse Tree

- -

Now we can look at the resulting parse tree as we did in Tutorials 1 - and 2.

- -

Select eg1.txt - in the Workspace, right click and select View - > Parse Tree. Don't forget that you can also use the View - Parse Tree button.

- -

Notice how the new parse tree has inserted the node _NP - under _ROOT.

- -

Expand the first _NP - node. It contains the nodes _det _adj - _adj _noun corresponding to The quick brown and fox:

- -

- -

When the analyzer encountered the determiner ("the"), followed - by two adjectives ("quick", "brown") and a noun ("fox"), - the rule within the NP pass file was matched. The default reduce action - was executed, yielding the NP subtree whose root is highlighted above. - The same reduce action was executed to create the noun phrase 'the lazy - yellow dog.'

- -

 Close the parse tree display.

- -

 

- -

Adding Multiple Rules to a Pass

- -

Not all noun phrases have two adjectives like 'quick' and 'brown' in - the phrase 'the quick brown fox'. Some noun phrases may have only one - adjective, such as 'the quick fox' and some may not have any adjectives - at all, as in the noun phrase 'the fox'. We can modify the NP pass file - so that we can catch these types of cases, too.

- -

Modify the NP - pass file to add the three new rules to the RULES - Region as follows:

- -

@CODE

- -

G("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES _ROOT

- -

 

- -

@RULES

- -

_NP <-

- -

_det

- -

_xWHITE

- -

_adj

- -

_xWHITE

- -

_adj

- -

_xWHITE

- -

_noun

- -

@@

- -

_NP <- _det _xWHITE - _adj _xWHITE _noun - @@

- -

_NP <- _det _xWHITE - _noun @@

- -

_NP <- _noun @@

- -

This demonstrates multiple rules in a single pass. 

- -

There are four rules, all operating in the same context node _ROOT. - (We could have written the first rule, which now occupies 9 lines, on - a single line like the other rules in the file.) We have now expanded - the NP rules so that it can "catch" noun phrases that have one, - two or no adjectives contained in it.

- -

Run the analyzer again.

- -

  Click on the View - Parse Tree button to see the new parse tree.

- -

Notice how the new parse tree has inserted the _NP - for 'the book'.

- -

- -

Bleeding and Feeding

- -

The ordering of rules within a pass file, as well as the ordering of - passes in an analyzer sequence, can create, hide or destroy parse tree - structure available for later rules and passes. The analyzer engine attempts - to match rules as they are presented in a pass file. That is, it starts - from the first rule written in the file and goes down the list of rules. - Therefore, the longest rule expansion in the example above is tried first.  - If we had ordered the rules with the shortest expansion (_NP - <- _noun) first, the noun 'Fox' - would get an _NP node label, hiding - the 'structural description' (the right hand side of the rule) for application - of the longer expansion rules. In other words, we would short-circuit - the rules.

- -

When a pass creates a structural description used by later passes, it - is said to be 'feeding.' When it removes the structure, it is said to - be 'bleeding.' A large part of developing a text analyzer is figuring - out the right bleedings and feedings.

- -

Close the parse tree display.

- -

Create a Possessive Pass (Escape Characters)

- -

Now we will write a rule to handle possessives, as in baby's.

- -

Use the procedure you just used to create the NP pass to create a new - pass:

- -

Select the Ana Tab. - Left click on end_words in the - pass sequence then right-click to pull up the popup menu. (This will add - a new pass BEFORE the NP pass).

- -

Select Add > New - Pass.

- -

Name the pass POSS - (for 'possessive') and leave Type: - Pattern.

- -

Click OK.

- -

Your Ana Tab should look like this:

- -

- -

Double-click on the POSS - pass to open it in the Workspace and edit the - RULES Region so that it looks like - this:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES _ROOT

- -

 

- -

@RULES

- -

_adj <- _noun \' - s @@

- - - - - - -
-

Note: - There isn't a space after the backslash in the rule.

- -

The use of the backslash \ here is to 'escape' the apostrophe.  - Other escaped characters include all the punctuation characters as well - as space, tab ('\t'), and newline ('\n'). (To display blank spaces visibly - in the parse tree, VisualText uses '\_', but this is just an artifact.)

- -

Save the POSS - pass file.

- -

Now, let's run the analyzer.

- -

Select the text file and click the Run - button. 

- -

Right-click on eg1.txt - in the Workspace and select View > - Parse Tree to view the - parse tree for eg1.txt.

- -

Scroll down to the - phrase 'the little baby's cheek' to check the parse tree. Expand the parse - tree. It should look like this:

- -

- -

Close the parse tree display.

- -

Now let's add a PP (prepositional phrase), VP (verb phrase), and Sentence - (sentence) pass to the analyzer sequence. For these passes, go ahead and - add them one at a time to the end of the sequence. (Unlike the POSS rule - added above, where you inserted the POSS pass before the NP pass, the - passes below should be added one after the other at the end of the analyzer - sequence.)

- -

Here is a procedure to add these passes:

- -

To create a pass in the Ana Tab:

- - - -

To modify the pass file double click - on the pass in the sequence to open it in the Workspace and modify the - rules accordingly. Make sure to save your file after editing it.

- -

Your prepositional phrase pass file (PP) - should look like this:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES _ROOT

- -

 

- -

@RULES

- -

_PP <- _prep _xWHITE - _NP @@

- -

And the verb phrase pass file (VP) should look like this:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES - _ROOT

- -

 

- -

@RULES

- -

_VP <- _verb _xWHITE - _NP _xWHITE _PP - @@

- -

_VP <- _verb _xWHITE - _PP _xWHITE _NP - @@

- -

_VP <- _verb _xWHITE - _NP @@

- -

_VP <- _verb _xWHITE - _PP @@

- -

_VP <- _verb @@

- -

Finally, the sentence pass file named Sentence - should look like this:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES - _ROOT

- -

 

- -

@RULES

- -

_S <- _NP _xWHITE - _VP @@

- -

Take a moment to go through and "read" each of the pass files. - Each file specifies what nodes should be analyzed (in this case the nodes - under _ROOT) and further what - should be done (specified in RULES Region) - when a match is found.

- -

The resulting analyzer sequence should look like this:

- -

- -

If your analyzer sequence does not match this sequence, use the yellow - arrows on the Tab Toolbar to move the pass to the correct location in - the analyzer sequence.

- -

We are now finished hand-building our rules. By using these rules, our - analyzer will be able to find not only noun phrases, verb phrases, prepositional - phrases but even sentences.

- -

Inspect Final and Intermediate Parses

- -

Now, let's run the analyzer with the new pass sequence on eg1.txt. This - time before we run the analyzer we will use the Generate - Logs button. Turning the Generate Logs button on will allow us - to view the cumulative parse tree - for each pass in the sequence, called the intermediate - parse tree. You'll see how this works in a moment.

- -

Select the Generate - Logs button.

- -

Click the Run button. 

- -

Let's take a look at the parse tree for the Sentence pass.

- -

Select the Sentence - pass in the analyzer sequence.

- -

Click the View Parse - Tree button.

- -

The Parse Tree should look like this (NOTE: - The lines pass is obsolete):

- -

- -

The rule in our Sentence pass file says, if you find - a node with the label _NP followed - by a _VP, rename the node _S. As you can see, the analyzer was - able to identify our three sentences.

- -

Note that the name of the pass, Sentence, is - shown at the top of the tree display window. Also notice the number 13 - before Sentence. This let's you - know that you are looking at the intermediate parse tree for the 13th - pass (Sentence) in the analyzer - sequence.

- - - - - - -
-

Note: - If you don't see the parse tree for each of the passes, make sure the - Toggle Generate Logs button is selected before you run the analyzer. If - you run the analyzer with the logs turned off, you will see a zero in - the title bar of the parse tree.

- -

 

- -

If you select another pass in the analyzer sequence, - like VP, and then View - Parse Tree button, you - should see the following:

- -

- -

Notice that the parse tree window is labeled VP, - and that it displays the result of - applying the rules in the VP pass file. 

- -

The pass you select in the analyzer sequence determines which intermediate - parse tree is displayed. Being able to easily inspect the intermediate - parses is useful when you're trying to create and debug your analyzer.

- -

Close the open files in your Workspace.  (You - can use Window > Close All - from the main menu.)

- -

Exit the analyzer. From the Main Menu, select File - > Exit.  (If - asked to save the analyzer, click Yes).

- -

Save a copy of - c:\apps\myAnalyzerFolder\myAnalyzer as myAnalyzer3.  You - will need this in order to run Tutorial 6 later on.

- -

Summary

- -

In this third tutorial, we created new passes - for the analyzer sequence by hand, edited the default settings, ran the - modified analyzer, and viewed the resulting parse trees.  We then - created additional rules, one of them requiring an escape character \ - for handling apostrophes.  We then used - the Toggle Generate Logs button, ran the analyzer, and viewed - the intermediate parse trees.
- 
- 
- 
- 
- 
- 
- 
- 
- 
- 
- 

- - - - diff --git a/Help/helps/Tutorials/tutorial4/tutorial4.html b/Help/helps/Tutorials/tutorial4/tutorial4.html deleted file mode 100644 index bb8a719..0000000 --- a/Help/helps/Tutorials/tutorial4/tutorial4.html +++ /dev/null @@ -1,2261 +0,0 @@ - - - - - - - - - - -tutorial4 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tutorial 4 - Input/Output

- -

In this tutorial we create and read input files, add more rule concepts - and associate concepts with the input files. We extract text from specific - parts of the parse tree and write such information to an output file.  - We also take a closer look at rule syntax. Finally, we introduce some - navigation shortcuts and a useful debugging tool in VisualText.

- -

We'll start with the analyzer from Tutorial 3.

- -

Launch VisualText.

- -

From the Main Menu, select File - > Recent Analyzers and then select c:\apps\myanalyzerfolder\myanalyzer.

- -

If you want to start with the ready-made analyzer in the docs folder, - unzip the contents of:

- -

C:\Program Files\TextAI\VisualText\docs\tutorial3\myAnalyzer.zip

- -

and place the myAnalyzer folder - in the working directory:

- -

C:\apps\myAnalyzerFolder

- -

Add Top Level Folder

- -

In Tutorial 2, you learned how to sample words in a file in order to - generate rules. You took each word individually and manually placed it - in the appropriate concept, noun, verb, adjective, etc.

- -

Sometimes it is more convenient to add sample data to the rule generation - component of VisualText by importing a text file with data that is already - of a certain type. For example, suppose you had a list of company names - that you wanted to train an analyzer to recognize. To do this, you need - to tell the system where the input file is and then give it a label.

- -

To keep things organized, we'll store files containing our lists to - be sampled in a new folder called dictionary.

- -

Create a folder in the Text - Tab by right-clicking in the window and selecting Add - > Top Level > Folder:

- -

- -

Name the folder dictionary - and click OK.

- -

This is where we'll store the file containing our company names.

- -

Your Text Tab should look like - this:

- -

.

- -

Create Input Files for Storing Lists

- -

Since we don't have a text file of company names created yet, we will - have to generate one now.

- -

In the Text Tab, - select the new folder you just created, - dictionary.  

- -

Right click to bring up the popup menu, and select Add >  New Text File.

- -

Name - the new file companies and click OK.

- -

A - blank text file with the titlebar 'companies.txt - (dictionary)' will pop up in - your workspace. "dictionary" appears in the titlebar of the - file to indicate that your text file, companies - is part of the dictionary folder.

- -

Enter the following company names into your file:

- -

Lambsoft

- -

Fencetech

- -

IBM

- -

Cisco

- -

GM

- -

GE

- -

Save the file by right-clicking on companies.txt - and selecting Save from the menu.

- -

Close and double-click - on the file to open it.  (VisualText - converts carriage return plus line feed to carriage return, which will - affect the offsets and automated rule generation, so this step needs to - be done for each new or saved file.)

- -

The - file companies.txt is now contained in the dictionary - directory. (If you don't see it in the Text - Tab, choose Refresh from - the Text Tab Popup Menu)

- -


 

- -

Rather than creating and typing in the contents of the file in VisualText, - we could also have copied an existing file from some other source directly - into the dictionary folder:

- -

C:\apps\myAnalyzerFolder\myAnalyzer\input\dictionary

- -

Create New Concept for Input Data

- -

Now that we have this data saved in our dictionary folder, we're going - to create a new rule concept and read the data into the concept from this - new file. We'll follow the same procedure we used in Tutorial 2 when we - created the rule concepts for det, adj, noun etc. This time however, we - will populate the concepts with the data contained in the companies.txt - you just created.

- -

In the Gram Tab, select the pos - folder. (Make sure the pos folder - opens up.)

- -

Right click, and select Add - > Rule:

- -

- -

Type in the concept name companies - and click OK.

- -

The new rule concept is added to the pos - folder. (If the rule concepts in your pos folder are expanded, scroll - down to see the newly added rule concept, companies.)

- -

Your Gram Tab should look like - this:

- -

- -

Associate File with Concept

- -

Now we are ready to import the content of our companies.txt - file into our new concept companies.

- -

In the Gram Tab, select companies, - then right click and select Sample File > Import from the menu.

- -

- -

VisualText will warn you that importing samples will delete whatever - sample data you have in the concept rule.

- -

- -

Since we don't have any sample data that can be deleted, we don't have - anything to worry about.

- -

Click Yes to the warning popup.

- -

The Open dialog box will take us to the analyzer input directory.

- -

Open the dictionary - directory and select the companies.txt - file.

- -

- -

Click Open.

- -

The samples in companies.txt - have been added to our new concept.

- -

- -

While we built the companies.txt - file by hand, any list of phrases can be used in this manner, as long - as there is one phrase per line in the text file. Although we named the - concept companies and the input - text file companies.txt, this - name matchup is not required.

- -

Generate Rules

- -

Now we are ready to have VisualText generate our new rule.

- -

From the Main Menu, select Analyzer - > Generate Rules > All.

- -

- -

Click on the Ana Tab - and look at the analyzer sequence. A new 'bead' or link has been added - to our analyzer sequence as the last item in the words - stub region. The pass companies - has been added.

- -

- -

Now let's take a look at the rule in our new pass file.

- -

Double-click on the companies - pass to display the automatically generated rule:

- -

- -

Since companies was the last - "child" concept of pos - in the Gram Hierarchy, its pass - becomes the last pass in the words - stub. 

- -

If you didn't make companies - the last child concept when you created it, you can move it around in - the Gram Tab using the yellow - arrows. If you do move the pass around however, remember that you'll need - to regenerate the rules in order to update the analyzer sequence. (From - the Main Menu, select Analyzer > Generate - Rules > All).

- - - - - - -
-

Note: - The order of generated passes should not be changed via the Ana Tab.

- -

 

- -

More on Rule Syntax

- -

Let's take a closer look at the structure of the companies - pass file.

- -

First, look at the first line that has been commented out. It reads: - AUTOMATICALLY GENERATED! EDITS WILL BE LOST!. - This warning is letting you know that any edits you make to an automatically - generated pass will not 'stick' the next time you try to automatically - generate the rule. For example, you can't edit your companies - pass file and then try to automatically regenerate the file. Your - edits will be lost. The only - way to modify the automatically generated files is to add or edit samples in the Gram Tab and regenerate the - passes. You can, of course, copy the contents of a pass into a hand-built - pass (outside of a stub region) -- this is a way to have VisualText build - rules for you. Our other type of pass, the hand-built type, can always - be hand-edited.

- -

The next line in the comment section of the pass file, tells you where - the .pat file 'resides':

- -

c:\apps\myAnalyzerFolder\myAnalyzer\spec - is the directory, and the file name is companies.pat.  - All pass (.pat) files reside in the spec directory. 

- -

The last line in the comment section of our automatically generated - pass file contains a timestamp indicating when the pass file was created.

- -

Now let's take a look at the rest our pass file.

- -

Again, @NODES _ROOT means that the rules in - the pass file are applied to the list of nodes directly dominated by a - node named _ROOT.

- -

The line starting with the @PRE - marker indicates a region of constraints (or PRE Actions) that parts of - a rule (or rule elements) must satisfy in order to match. In this case, - the condition requires that the first element of each subsequent rule - match a capitalized word. Here you can see how VisualText performed some - rule generalization.  It looked at all the samples and deduced a - pattern in the text, that is, that all the words in the samples started - with capital letters and added this condition to the rule file.

- -

The <1,1> syntax means that the PRE action applies from element - 1 through element 1 of the phrase on the rule's right hand side.  - The first element of our rule is _xWILD. The - condition placed on this rule, is that whatever _xWILD - turns out to be, it must be a capitalized word. If we had <1,2> - instead of <1,1> then both the first and second elements in the - rule's phrase would need to be capitalized, and so on.

- -

The region marker @RULES - indicates where rules occur. We've introduced the basic syntax of a rule - earlier in this set of tutorials.

- -

Now, let's take a look at the rest of our pass file.

- -

The 'variable' _xWILD - is a wildcard match and matches anything within the constraints placed - on it.  In this case, the conditions of the wildcard match are given - following the wildcard in square brackets [ ].  The conditions min=1 - max =1 mean that _xWILD must match a minimum of one node - and a maximum of one node.  If max = 2, then the rule will match - one or two nodes of the parse tree. (max=0 means an unconstrained maximum. - min=0 means that the wildcard can optionally match zero elements.)

- -

The s - modifier stands for singlet.  It tells the rule matching engine to - search a node's descendants for a match, but stop looking 'down' in the - tree when a node has more than one child (i.e., it branches into more - than one node) or when a node has its BASE attribute set.  Singlet - and base allow syntactic and semantic features to be layered into the - parse tree.  If - a series of passes had applied that only operated on single elements, - building up a 'stack' of node labels for that element, this rule could - still match against a lower node in the stack of labels. In brief, the - singlet mechanism allows attributes to be placed and matched directly - within the parse tree. The subtree of attribute nodes is called a singlet chain.

- -

The final section of the conditions is the match list of tokens that - _xWILD must match, corresponding to the items - in the input sample file.

- -

Create Another Input Example

- -

Now, let's create some text that exercises the new companies - pass.  First, - let's create a new folder in the Text - Tab to hold our example file. Again, folders are used to help you - organize your text files.

- -

In the Text - Tab window, right click and select Add - > Top Level > Folder:

- -

Name the - folder examples. 

- -

Select the examples - folder.

- -

To create a new file, - right click and select Add > New Text File.

- -

Name the file eg2 - (again omitting the .txt extension).

- -

Notice that the folder name - "examples" is part of the titlebar of the eg2.txt file.

- -

Now, type the following text in the body of the text - file:

- -

Lambsoft jumped over Fencetech.
-Cisco gave GE the yellow book.

- -

Close and open the file to make sure VisualText converts it.

- -

Now let's run the analyzer on eg2.txt.

- -

Since the eg2.txt file is already selected in the Text Tab:

- -

Click the Run button.

- -

View the parse tree, by right clicking on eg2.txt in the Workspace, and selecting View > Parse Tree.

- -

- -

At first glance, it may look like our Sentence pass is not working here, - since we don't have any _S nodes. - (Remember that the rule in our Sentence pass says, if you find a NP followed - by a whitespace and then a VP, reduce them to the new node _S.) - In fact, the Sentence pass is working fine. It is just that the rules - in the Sentence pass file require a noun phrase before a verb phrase and - since the names in companies file are reduced to _companies - and not _noun, the conditions - set in the Sentence pass are not - met. We'll have to go in and create a new concept to get this done.

- -

Alter Existing Rules to Handle New Concept

- -

Now let's create a new pass in our analyzer sequence that will rename - a _companies node to be a _noun node. (Developer's - note: another way, not shown here, uses the Properties dialog in - the Gram Tab to add the layered label _noun when automatically generating - the rules for _companies.)

- -

In the Ana Tab, - select end_words. We want our - pass to be inserted before the - POSS pass.

- -

Right click and select Add - > New Pass.

- -

Name - the new pass company2noun.

- -

Leave Type as Pattern.

- -

Double-click the newly created company2noun - pass, and edit the RULES Region so that it looks like the following:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES - _ROOT

- -

 

- -

@RULES

- -

_noun <- _companies [s] - @@

- -

Now that we have this pass in our analyzer sequence, company names can - be treated as regular nouns. 

- -

Run the analyzer on eg2.txt. -

- -

Now, let's take a look at the new parse tree.

- -

(Select the Sentence - pass in the Ana Tab if the Generate - Logs toggle is on.)

- -

Select eg2.txt in - the workspace, right click and select View - > Parse Tree. Expand the first _S - and look at the parse tree.

- -

- -

That worked pretty well, except notice that in the second sentence, - the second noun phrase wasn't included. Instead of the sentence, 'Cisco - gave GE the yellow book.', our analyzer parsed 'Cisco gave GE' as a complete - sentence.  The structure of the sentence, 'Cisco gave GE the yellow - book.' is a noun phrase (Cisco) followed by a verb phrase (gave) followed - by two noun phrases, 'Cisco' and 'the yellow book.' It appears as if our - VP pass file does not account for "verb + NP + NP".

- -

Let's take a look at the VP - pass file.

- -

Double click on the VP - pass in the analyzer sequence. - Notice that our rules do not cover this case.

- -

- -

Let's edit the VP pass file to allow a verb phrase to be formed from - a "verb + NP + NP".

- -

Add the following rule as the first rule in the RULES Region:

- -

  _VP <- _verb _xWHITE _NP _xWHITE _NP @@

- -

Run the analyzer again on eg2.txt - with the Generate Logs - button on.

- -

Select eg2.txt - and view the parse tree.

- -

The parse tree should look like this:

- -

- -

(Select the Sentence pass in - the Ana Tab to see the parse tree.)

- -

Print Something To Output File

- -

One of the helpful features of VisualText, is the ability to extract - specific parts of the parse tree and write the extracted information out - to a file. Up until now, we have viewed our results using the parse tree. - Now we will learn how to direct the results to an output file. We can - do this by adding a POST Region to the relevant pass file.

- -

Double click on the company2noun - pass in the Ana Tab to - open up the pass file.

- -

Add the POST Region to the file so that it looks like - this:

- -

@CODE

- -

G("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES - _ROOT

- -

 

- -

@POST

- -

 "output.txt" - << "Found company: " - << N("$text") << "\n";

- -

 single();

- -

 

- -

@RULES

- -

_noun <- _companies [s] - @@

- -

You are already familiar with the NODES - and RULES part of the rule file. @POST is the marker - for the POST Region.  This - region contains code (the two lines between @POST - and @RULES) that will be executed if a rule - in the subsequent RULES Region matches.

- -

Let's take a closer look at the code in the POST Region.

- -

If you are familiar with C++ programming, you'll notice that the code - to write to an output file is similar.

- -

The first line of code can be broken down into two parts: what is to - the left of the output operator (<<) and what is to the right of - the operator. The left side of the operator specifies the name of the - output file, in this case 'output.txt', and the right side specifies what - should be written to the output file and how. We'll go over the code in - more detail below.

- -

The first item to the right of the output operator is "Found company: - ". This basically says, print the text "Found company: " - in the output file. Material enclosed in double quotes is printed literally.

- -

The next thing that gets printed to the output.txt is indicated by N("$text"). - This means 'get the text in the last node element' (i.e., the text contained - in _companies) and append it to - the output file.  (If there were more than one element in the right - hand side of the rule, you could specify which element to extract the - text from with a second argument to N, for example, the second element's text - could be extracted with the expression N("$text",2).) 

- -

Our last item, is "\n". This adds a newline. 

- -

The last item on our first line of code is the semicolon, analogous - to the statement terminator of other programming languages.

- -

The second line, single();, - instructs the rule engine to perform a single-tier, or default, reduction. - This is described below.

- -

Single

- -

(Advanced section. Primarily for the - programmer or computational linguist.)

- -

A reduction rule is one which takes one or more elements, such as the - sequence _NP _VP, and reduces - the matched phrase of nodes to one node, such as _S. - For historical reasons, the default reduce action is called a single-tier - reduction.  The default reduction can be superseded in the - POST Region.  That is, the rule - _S <- _NP - _VP will perform a single(); - reduction by default.  If you do write a non-empty POST Region, however, - you must specify the type of reduction you want, else the default is noop.  - noop - is an action that does nothing (i.e., "no op" or "no operation"). - Two other reduction types include group and excise.

- -

group - is used when you have a structural description such as 'A B C D' but you - only want to create a node from B and C, leaving A and D alone. A pass - such as

- -

@POST
-group(2,3,"_foobar");
-@RULES
-_xNIL <- A B C D @@

- -

creates a structure like A _foobar(B C) D.

- -

excise - deletes structure from the parse tree, and noop - causes no reduction to be performed. For a summary of the different types - of Post Actions available, see POST - Actions. For now, we need to include the default reduction type for - the purposes of the current example.

- - - - - - -
-

Note: - Omitting single();in - the POST Region is a common error. 

- -

 

- -

Inspect Output File

- -

Now that we have edited our company2nouns - pass file to print information to an output file, we are ready to run - the analyzer.

- -

Click on the Run - button. (You can also run the analyzer by right clicking on eg2.txt - in the Workspace and selecting, Run.)

- -

There are a couple of methods to view output files. The easiest is:

- -

Select the View Output - File button located on the Debug - Toolbar.

- -

The output.txt file will open up in the Workspace. The result looks - like this:

- -

- -

The output file output.txt is - written to a system-created directory. The path to the file is:

- -

C:\apps\myAnalyzerFolder\myAnalyzer\input\examples\eg2.txt_log\output.txt

- -

Another way to view the output file is from the Text - Tab Window.

- -

In the Text Tab, open the - examples folder. Expand - eg2.txt.

- -

Double click on the output.txt - file.

- -

- - - - - - -
-

Note: - The ana~.log files here let you know the analyzer was run with the Generate - Logs button on.

- -

Notice that the output file is placed into directories created by the - system corresponding to the input file name.  You can see this in - the titlebar of the output file. This makes working with multiple input - files convenient. (The other files in the folder are the intermediate - log files.)

- -

Note that the main output file, named output.txt - by default, is special and is bound to the View Output File button. If - you had named the output file something like foo.txt, - with a print statement like

- -

"foo.txt" << "found companies : " << N("$text") << "\n";

- -

then the View Output File button would not display the file foo.txt.

- -

To view such an output file, use the View - Dump Files button.

- -

This will list all the files with a .txt extension in the current log folder, - or default output folder associated with an input file. If you - were to have two output files created by:

- -

"foo.txt" << "found companies : " << N("$text") << "\n";
-"foo2.txt" << "found companies : " << N("$text") << "\n";

- -

then you could use the View Dump Files - pulldown list to choose - between them:

- -

- -

Navigate - Input Files

- -

There are many shortcuts in VisualText to help you navigate among files. - One of these navigation shortcuts uses the Previous - Text and Next Text buttons - on the Debug Toolbar.

- -

- -

For example, if you want to run the analyzer on the next file in the - currently selected folder in the Text Tab, you can use these buttons to - quickly get the file.

- -

First close all the open files in the Workspace.  (You - can use Window > Close All - in the main menu).

- -

Create another file in the examples - folder. (Double click to open the examples folder, right click and select, - Add > New Text File.)

- -

Name the file eg3.txt. In the eg3.txt file, add the following - text:

- -

John kissed the dog.
-Mary gave the book to John.

- -

Save eg3.txt.

- -

Select the Page Mode - Toggle button in order to enable the Next - Text and Previous Text - buttons.

- -

Notice how the two horizontal arrows to the right on the toolbar are - enabled:

- -

Next Text

- -

- -

Previous Text

- -

 

- -

Click on the Previous - Text button. Notice eg2.txt - is yellow in the Text Tab, and - that eg2.txt is the name of the - file displayed in the workspace.

- -

- -

Now click on the Next - Text button.  Notice - that eg3.txt in the Text - Tab is now selected (it turned yellow), and the file displayed - in the Workspace is eg3.txt.

- -

- -

Clicking the Previous - Text button operates the selection display in reverse.

- -

Highlight Matched Concepts

- -

Another one of VisualText's many shortcuts is a debugging tool used - to highlight concepts.

- -

The Toggle Highlight Mode button - highlights regions of text in an analyzed file that match - rules in a selected pass of the analyzer. This tool lets you quickly see - whether and where a pass has a rule that has applied.

- -

An alternative to using Toggle Highlight - Mode, is to inspect the parse tree, but this can be a slower process - (although it gives you more information). To use this highlighting feature, - you must have the Toggle Generate Logs - button selected. (Information on the Toggle Generate Logs button is given - below.)

- -

Let's see how the Toggle Highlight - Mode works.

- -

Select the Toggle Generate - Logs button. (Toggle - Generate Logs is next to Toggle - Highlight Mode on the toolbar.)

- -

Double click on the eg1.txt - to open it in the Workspace.

- -

Run the analyzer on eg1.txt. -

- -

Select the det - pass in the Ana Tab, then click - on the Toggle Highlight Mode button - :

- -

- -

Notice how the text matching the det - pass turns blue.

- -

Now select a later pass in the analyzer sequence, for - example the noun pass. Notice - how the text changes to now match the noun - pass file.

- -

If you prefer using pulldown menus, the same function can be accessed - by selecting Highlight Matches from either the Gram Tab or the Ana Tab - popup menus.

- -

Close the open files in the Workspace.

- -

(If available) save your analyzer. (File - > Save Analyzer.)

- -

(If available) answer Yes - at the confirmation prompt.  And - the same for saving the knowledge base.

- -

Exit the analyzer. From the Main Menu, select File - > Exit.

- -

Summary

- -

In this tutorial we created an input file - to store lists of items in a concept, created a new top concept for the - item and imported the file's samples into the concept.   We - generated the rules and ran the analyzer against the input data.  - We discussed some features of the automatically generated rules, then - created another hand-built rule to utilize the automatically generated - rules in the analyzer logic.  Finally, we printed some data to an - output file and explored some navigation and debug tools.

- - - - diff --git a/Help/helps/Tutorials/tutorial5/tutorial5.html b/Help/helps/Tutorials/tutorial5/tutorial5.html deleted file mode 100644 index cdd4f26..0000000 --- a/Help/helps/Tutorials/tutorial5/tutorial5.html +++ /dev/null @@ -1,2476 +0,0 @@ - - - - - - - - - - -tutorial5 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tutorial 5 - NLP++

- -

In the last several tutorials, you have learned how to create pass files - in VisualText. The pass files you have seen so far contain NLP++ code - and rules.

- -

NOTE: This tutorial does not discuss NLP++ user-defined functions and - local variables (L). -  The user - can find these in the Help sections dealing with NLP++.

- -

About NLP++

- -

NLP stands for Natural Language Processing. NLP++ is a general purpose - programming language combined with a grammar. Text Analysis International's - NLP++ allows you to concentrate on the conceptual aspects of building - a text analyzer. In other words, it simplifies the programming task involved - in building the analyzer.

- -

NLP++ is tightly bound to the parse tree used in VisualText-style text - analyzers, and also to the Conceptual Grammar knowledge base management - system, or KBMS. Conceptual Grammar is a knowledge representation framework - consisting of concepts, attributes and phrases, combined into knowledge - hierarchies and graphs. The hierarchy of samples in the Gram Tab is implemented - as a hierarchy within the Conceptual Grammar KBMS.

- -

A key feature of NLP++ is the ability to access and augment parse tree - information dynamically during text analysis. That is, it creates and - uses information during the text analysis process itself. NLP++ uses a - special syntax for variables that simplifies the manipulation of information - in parse trees. In this tutorial, we introduce five special variable types - and show how they can help us with two classic linguistic parsing problems - common to text analysis: extracting the head noun in a noun phrase, and - determining the subject and object of a sentence.

- -

Five Variable Types

- -

We'll briefly introduce the five special types of NLP++ variables here. - We'll go into more detail later on as we start to edit our rule files. - (For more information on variables see About - NLP++ Variables.)

- -

The five types of variables are: G - (global), N (node), - X (context), S (suggested), and L (local). The N, X, - and S variables - assume a rule context (that is, they apply only in regions where a rule - has matched, such as @POST).  The - G and L - variables are independent of the rule matching machinery.

- - - -

Some operations on these variables include: assignment, printing, arithmetic - and string manipulation. These operations will also be discussed below.

- -

In this tutorial, we will use the zipped version of the analyzer in - Tutorial 4. If VisualText has myAnalyzer from the myAnalyzerFolder accessed, - you'll have to close the analyzer. (Select Close - Analyzer from the File Menu.)

- -

First, navigate to:

- -

C:\apps\myAnalyzerFolder

- -

and rename the myAnalyzer folder - myAnalyzer4.

- -

Unzip the myAnalyzer4.zip - file located in:

- -

C:\Program Files\TextAI\VisualText\docs\tutorial4\myAnalyzer4.zip

- -

and place the myAnalyzer folder - in the working directory:

- -

C:\apps\myAnalyzerFolder

- -

Load with Recent Analyzers

- -

Here is a shortcut for loading an analyzer.

- -

Double-click the VisualText icon to launch VisualText.

- -

Select File > Recent - Analyzers and open:

- -

c:\apps\myanalyzerfolder\myanalyzer

- -

This will automatically load the .ana file in the myAnalyzer folder. 

- -

 

- -

Global Variables

- -

To illustrate the Global variable (G), - we will use the company2noun pass - file.

- -

In Tutorial 4, you added a POST Region directing your analyzer to create - an output file with the text "Found Company: company name". - We will add rules to count the total number of companies in a file and - write this information to the output file.

- -

Double click on the company2noun pass in the Ana Tab.

- -

Edit the company2noun - pass file so that it looks like this:

- -

@CODE

- -

  G("companyCount") = - 0; # Set to - 0 (the default).

- -

@@CODE

- -

 

- -

@NODES - _ROOT

- -

 

- -

@POST

- -

  # - Increment count of companies.

- -

  if(G("companyCount"))

- -

    G("companyCount") = G("companyCount") + 1;

- -

  else

- -

    G("companyCount") = 1;

- -

 

- -

  # - Print the text from the last node matched (_companies), and the count.

- -

  "output.txt" - << "Found company: " - << N("$text") << ". - Found "

- -

    << - G("companyCount") - << " companies so far\n";

- -

 

- -

  single(); - # Do the standard rule reduction.

- -

@RULES

- -

_noun <- _companies [s] - @@

- - - - - - -
-

Note: - Include the extra spaces as they appear in the pass file. This will make - your output file easier to read.

- -

Let's take a closer look at the rules you just added.

- -

There are three regions of our pass file, NODES, - POST and RULES. - It is generally easiest to start an analysis of a pass file by looking - at the RULES Region first.

- -

The rule in the RULES - Region

- -

_noun <- _companies @@

- -

tells the analyzer that any node in the parse tree named _companies is - rewritten with a - new dominating node _noun. The - new node, _noun is layered on - top of the existing node within the parse tree. The node _companies - attached to the company name "Fencetech", for example, will - have the node _noun layered on - top of it.

- -

The line @NODES - _ROOT specifies the context in which the rule is applied by the - rule engine. In this case, the rule is applied only to nodes which are - dominated by the node named _ROOT. - Recall that the nodes dominated by _ROOT - are underneath and indented to the right of _ROOT - in the parse tree.

- -

Finally, the POST Region contains code "actions" that will - be applied whenever there is a match in the RULES Region. The first part - of our POST Region,

- -

G("companyCount")

- -

creates a new global variable (G) - named companyCount, if one has - not been previously created.  There is no need to declare the variable - in advance. 

- -

The first four lines of the POST - Region track the count of companies in a file. The global variable companyCount is incremented by one for - every company found.

- -

The next part of the POST - Region specifies printing to an output file. The code prints the - company count to the output file "output.txt". It also specifies - a format for printing to the output file. Generally, where and how something - should be printed is controlled by the output operator <<. In this - example, we output to the file 'output.txt' in the following format:

- -

"Found Company: Company Name. Found n companies so far."

- -

Note that expressions to the right of the output operator can be variables - such as G("companyCount"), literal text surrounded - by double quotes (e.g. "Found company: ") - or formatting characters ("\n" for - newline or "\t" for a tab space). Note also that it is possible - to chain together a series of operations within a single NLP++ statement, - as we did here using multiple output operators.

- -

The last part of the POST Region, - single(); as we - learned in Tutorial 4, is an instruction to the rule engine to perform - a standard (or "single-tier") reduction.

- -

Node Variables

- -

We - first saw the - variable N - in Tutorial - 4. The node variable is associated with nodes that match elements - on the right hand side of rules and is written in the following way

- -

N("variable-name", position-on-right-hand-side-of-rule).

- -

A related form is, for example, N(3), which evaluates to the first parse - tree node matching rule element 3.

- -

The positions are from 1 to N, and if there is only 1 then the position - reference can be omitted as it has been in our rule statement. The $ prefix - in the argument N("$text"), - indicates this is a special NLP++ variable name, and is reserved. There - are a number of reserved variables that return useful information. See - the Special NLP++ Variables - for a complete list of special reserved variables.

- -

N("$text") - in our statement extracts the text under the last node of the right hand - side of the rule, i.e., whatever is in _companies. - In short, the expression

- -

"output.txt" << N("$text")

- -

extracts whatever is in _companies - and puts it into the output.txt file.

- -

The text could also be assigned to another variable associated with - the node, such as:

- -

N("string",1) = N("$text",1);

- -

and that text could be sent to an output file with the expression:

- -

"output.txt" << N("string",1);

- -

Note the difference between the node variables N("string") - and N("$text"). - Variables prefixed with the $ are special reserved variables that act - like functions.

- -

Highlight eg2.txt - in the Text Tab and run the analyzer. -

- -

Double click on the output.txt - file in the examples folder to see the results. It - should look like this:

- -

- -

If there are any errors in formatting, for example not enough spaces - between text, open the company2noun - pass file and edit accordingly. Run the analyzer again until output.txt - resembles the one above.

- -

Now, let's look at a neat way to simplify the counting code in the POST Region. We will do this by using the '++' - operator.

- -

Edit the company2noun - pass file to simplify it as shown here:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES - _ROOT

- -

 

- -

@POST

- -

#increment count of companies

- -

++G("companyCount");

- -

 

- -

# print out the text from the node of the - last

- -

# matching rule element and the count

- -

    "output.txt" - << "Found company: " - << N("$text") << ". - Found "

- -

 << - G("companyCount") - << " companies so far\n";

- -

   

- -

#do a single tier reduce

- -

single();

- -

 

- -

@RULES

- -

_noun <- _companies [s] - @@

- -

Notice that the "if" and "else" statements have - been replaced by the single statement: "++G("companyCount"). The  ++ - operator means 'increment the value of the variable by one'.  To - decrement something, you can use the decrement operator, --.  When - a variable is first used, its value is set to 0 by default. 

- -

In addition to the increment and decrement operators, NLP++ also supports - the standard arithmetic operators, add(+), substract(-), multiply(*), - and divide(/), as well as the logical operators AND(&&), OR(||), - and NOT(!). One of the text operators include the concatenation operator - (+). You'll see how the + operator works in the next section. For a complete - list of NLP++ operators see the Operators - and Expressions topic.

- -

 

- -

Build A List from the Input

- -

Suppose we wanted to build a list of the companies that have been mentioned - in our text file. To illustrate one way this can be done, we will make - an output file which concatenates the names of the companies in a file - and puts them together on a single line.

- -

Now edit the company2noun - file so that it looks like the following:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES - _ROOT

- -

 

- -

@POST

- -

# - increment count of companies

- -

++G("companyCount");

- -

 

- -

# print out - the text from the node of the last

- -

# matching rule element and the count

- -

"output.txt" - << "Found company: " - << N("$text") << ". - Found "

- -

  << - G("companyCount") - << " companies so far: ";

- -

 

- -

# concatenate the company names

- -

if - (G("outputString"))

- -

  G("outputString") = G("outputString") + ", - " + N("$text");

- -

else

- -

  G("outputString") = N("$text");

- -

 

- -

# print out - the list of companies found so far

- -

"output.txt" - << G("outputString") << "\n";

- -

 

- -

# do a single - tier reduce

- -

single();

- -

 

- -

@RULES

- -

_noun <- _companies [s] - @@

- -

 

- -

Here's some pseudo code that describes how to concatenate the company - names:

- -

If outputString has been declared, then
-    concatenate it with the text from the first node element
-otherwise, (if it has not been declared)
-    initialize it with the text from the first node element

- -

Select eg2.txt - and run the analyzer.

- -

Open output.txt - and look at the results. You should get the following:

- -

- -

The above output shows the results of applying the - NLP++ string catenation operator ("+") to build lists.
 

- -

Variables of  Suggested and Context Nodes

- -

Now, let's take a look at our third variable S, the suggested variable. S denotes a variable associated with the - left hand side of a rule (the element to the left of the <-), the suggested node or the structural change. - The suggested node is built if a reduce action such as single - is executed, else it is not built. (If the suggested - node is not built, then its S variables are removed - when the rule actions have finished executing.)

- -

Earlier in this tutorial we said that we would be extracting the 'head' - noun of a noun phrase. The S variable is going to help us do that. - But first, let's review what a head noun is.

- -

A head noun is a technical - linguistic term, like subject, - or indirect object. Head - noun refers to the main, or 'head' word in a noun phrase.  - For example, in the phrase 'The man from Philadelphia', 'man' is the head - noun, and determines a number of parameters associated with the noun phrase, - such as whether it is single or plural. Whether a noun phrase is singular - or plural is important for number agreement with the verb. For example, - 'The man - from Philadelphia is...' - vs. 'The men - from Philadelphia are...' - Since the agreement between the head noun and the verb takes place over - unpredictably long distances, some process is needed to extract the value.

- -

Double click on the - NP pass - in the Ana Tab.

- -

Modify the NP - pass to look like this:

- -

 

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES _ROOT

- -

 

- -

@POST

- -

  # the variable named - "head" in the suggested node

- -

  # - gets the text of the LAST node matched by a rule below.

- -

  S("head") = N("$text");

- -

  single();

- -

@RULES

- -

_NP <-

- -

_det

- -

_xWHITE

- -

_adj

- -

_xWHITE

- -

_adj

- -

_xWHITE

- -

_noun

- -

@@

- -

_NP <- _det _xWHITE - _adj _xWHITE _noun - @@

- -

_NP <- _det _xWHITE - _noun @@

- -

_NP <- _noun @@

- -

Basically, what we are doing here is extracting the 'head' noun of a - noun phrase and putting it in a variable named "head". Since - we are using the S - variable, 'head' is associated with the left hand side of the rule, the - _NP.

- -

The POST Region applies to all four rules in the subsequent RULES Region. - Since N("$text") - without a rule element number fetches the text of the node matching the - LAST element of a rule, this same action can apply to all four rules, - since they all happen to end with a _noun - element.

- -

The net effect of the rules is to extract the text from the _noun - element on the right hand side of each rule, and write the text to a variable - named head which is associated - with the suggested node, named _NP, on the left hand side of each rule. - Each parsed noun phrase will now have an associated head noun. A global - variable wouldn't work in this instance, because there would only be one, - and we need a separate variable 'head' for each noun phrase.  S variables - can also serve as "local" variables, since they are visible - only in the context of the current suggested node of a rule. (Whereas - global variables are available anywhere, anytime.)

- -

Select the eg1.txt - file in the Text Tab and run the - analyzer. (Select the Generate Logs - button, so that you can see the intermediate parse tree information.)

- -

Open eg1.txt - in the Workspace and right click on the text to view the resultant parse - tree. Click the Toggle Node Variables - button to see the variable name (head) and the value assigned - to the variable for each noun phrase.

- -

- -

Now is a good time to present some other labels you will see attached - to nodes in parse tree displays. In addition to fired - and blt, you may also see the - labels dsem, un - and b attached to nodes. These - labels are summarized in the following table.

- - - - - - - - - - - - - - - - - - - - - - - -
-

Label

 -

Description

-

dsem

 -

The node has some 'semantic' variables set.

-

un

 -

The node is unsealed. A sealed node does not - expose its internal structure to the rule engine, so additional rules - cannot apply and alter it. If a node is set to unsealed, its structure - is available for taking part in later rules.

-

b

 -

"b" is for "base" and - tells the rule matcher to stop looking down singlets in this node. A singlet - is a node label that dominates (contains) another single node. For example, - in (NP (N John)) - there is only one node, labeled N, - having the text John, - contained in the node NP. - The NP node is a singlet, and - so is the N. The N - is available (exposed) to match the structure description (right hand - side) of later rules. If the NP - node's base attribute is set, the N - will not be available to later rules.

-

sem

 -

Used for internal semantic data structures - that the user doesn't have access to.

- -

Now that we are extracting the head nouns from noun phrases, let's extract - the head noun of the direct object from the elements of the verb phrase - and associate it with the verb phrase node. 

- -

Double click the VP - pass in the Ana Tab and edit it - so that it looks like this:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES _ROOT

- -

 

- -

@POST

- -

  S("directObject") = N("head",5);

- -

  S("indirectObject") = N("head",3);

- -

  single();

- -

@RULES

- -

_VP <- _verb _xWHITE - _NP _xWHITE _NP - @@

- -

 

- -

@POST

- -

  S("directObject") = N("head",3);

- -

  single();

- -

@RULES

- -

_VP <- _verb _xWHITE - _NP _xWHITE _PP - @@

- -

 

- -

@POST

- -

  S("directObject") = N("head",5);

- -

  single();

- -

@RULES

- -

_VP <- _verb _xWHITE - _PP _xWHITE _NP - @@

- -

 

- -

@POST

- -

  S("directObject") = N("head",3);

- -

  single();

- -

@RULES

- -

_VP <- _verb _xWHITE _NP @@

- -

 

- -

@RULES

- -

_VP <- _verb _xWHITE - _PP @@

- -

_VP <- _verb @@

- -

 

- -

The line

- -

S("directObject") = N("head", 5);

- -

handles sentences like "John gave Mary the book" where "the - book" is the direct object, i.e. the 5th element, _NP. - The rule takes the value of a variable named head - associated with the 5th element (an NP) of the right hand side of the - following syntactic rule, and copies it to a variable named directObject associated with the suggested node - on the left hand side of the syntactic rule (the VP).  The value - of the variable head was assigned - in the preceding NP pass.

- -

The multiple POST and RULES - sections are matched to extract the correct arguments (5, 3, 5, 3), i.e. - the element corresponding to _NP - in each rule.  The last two rules do not have direct objects, so - the value of the variable directObject is left - alone (i.e., it will default to zero). Note that the lone @RULES specifier resets the subsequent rules to - have the default POST Action, i.e. - the implicit single reduce action.

- -

Run the analyzer on the eg1.txt - and view the resultant parse tree.

- -

 

- -

Find Grammatical Roles: X Variables

- -

Having extracted head nouns and figured out which NPs are direct objects, - we're ready to complete the analysis of major syntactic categories. Also, - we're ready to introduce the last of the five special variable types in - NLP++, the context, - or X - variables. To do this, we need to add another pass to our analyzer sequence.

- -

Click on the Sentence - pass in the Ana Tab. Right click - and select Add > New Pass.

- -

Name the new pass, roles.

- -

  Leave Type as Pattern and click OK.

- -

The purpose of the roles pass is to extract the subject and object - of a sentence.

- -

Double-click the roles pass - and edit so that it looks like this:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@PATH - _ROOT _S

- -

 

- -

@POST

- -

  X("subject",2) = N("head");

- -

  noop();

- -

@RULES

- -

_xNIL - <- _NP @@

- -

 

- -

@POST

- -

  X("object",2) = N("directObject");

- -

  # noop();

- -

@RULES

- -

_xNIL - <- _VP @@

- -

The structure of this pass file is a little different than the other - pass files we have been editing in this tutorial. Let's take a closer - look at each section of the file.

- -

@PATH Selector

- -

The line

- -

@PATH _ROOT _S

- -

is different from what we've seen before. Earlier, we only made references - to the SELECT Region

- -

@NODES _ROOT 

- -

The @PATH - _ROOT - _S syntax instructs the rule matching engine to look only at a - path from the root of the parse tree, _ROOT, - down to its direct child _S when - trying to match the rules of the current pass.  This means that not - any old NP will match, but only NPs that are dominated by _S, - which in turn are dominated by _ROOT.  - Similarly, not any old VP will match, but only those dominated by _S dominated by _ROOT.

- -

X(variable-name,position) - is the prototype of the fourth variable type. The line

- -

X("subject",2) = N("head");

- -

assigns to the second context node in the @PATH list, that is, - _S, a variable named subject. - The rule engine copies into that variable the value of the variable named - head - which is associated with the (default) first element of the syntactic - rule (i.e., _NP). 

- -

The line noop(); means 'no operation', and means that - no reduction should take place in the syntactic rule.  That is, do - not place an _xNIL - on top of the _NP.  This - is a method for finding a node to match without modifying the parse tree. 

- -

The second rule set matching against a _VP - extracts the value of directObject - associated with the _VP and copies - it into a variable named object - associated with the context node _S. - Note that we don't need to explicitly add a noop - action. Once a POST Region is non-empty, the default single - action is displaced. The commented noop - action reminds us that there is no implicit reduce action in a non-empty - POST Region.

- -

 

- -

Print Out the Results

- -

Finally, let's create another pass file to print out the subject and - object roles for each of the sentences in eg1.txt.

- -

In the Ana Tab, - highlight the roles pass.

- -

Right click and select Add - > New Pass.

- -

Name the new pass - printRoles. - Leave Type - as Pattern - and click OK.

- -

Double-click the printRoles pass - and edit so that it looks like this:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES - _ROOT

- -

 

- -

@POST

- -

  "output.txt" - << "sentence: " << - phrasetext() << - "\n";

- -

  if - (N("subject"))

- -

    "output.txt" - << "subject: " << - N("subject") - << "\n";

- -

  if - (N("object"))

- -

    "output.txt" - << "object: " << - N("object") - << "\n";

- -

  "output.txt" - << "\n";

- -

@RULES

- -

_xNIL - <- _S @@

- -

This pass looks for a node labeled _S - dominated by _ROOT. If it finds - one, it prints out the text of all the nodes that matched the right hand - side, using the predefined function phrasetext. - Then it prints the value of _S's - variables named subject and object. - Finally, it terminates the output for each sentence with a blank line.

- -

Run the analyzer on eg1.txt. 

- -

Open the output.txt - file located under eg1.txt. Output.txt - should look like this:

- -

- -

If your results are different, go back to the last three pass files - we edited, printRoles (pass 17), - roles (pass 16) - and VP (pass 14) - to make sure the code is the same.

- -

Now open the examples - folder and run the analyzer on eg2.txt.

- -

The results in the output.txt - file should look like this:

- -

 
- 

- -

Notice that the information generated from the printRoles - pass has been appended to the output.txt file.

- -

Close the open files in the Workspace.

- -

Exit the analyzer (File - > Exit.) and save whatever is asked to be saved.

- -

Summary

- -

In this tutorial, we introduced five types of NLP++ variables:

- - - -

One way to think of these four types of variables is to remember that - a global variable can be anywhere, but with only one instance. For the - other three, imagine an abstract NLP++ pass:

- -

@NODES - X

- -

@RULES

- -

S - <- N ... @@

- -

Variables of type X - are associated with the context of a rule, variables of N - are on the right side and variables of type S - on the left.

- -

Operations on the variables introduced were:

- - - -

We also introduced a new type of context selector, - the PATH selector, which specifies a more - constrained context for the application of rules than the NODES selector introduced earlier.

- -

 

- - - - diff --git a/Help/helps/Tutorials/tutorial6/tutorial6.html b/Help/helps/Tutorials/tutorial6/tutorial6.html deleted file mode 100644 index f368986..0000000 --- a/Help/helps/Tutorials/tutorial6/tutorial6.html +++ /dev/null @@ -1,1676 +0,0 @@ - - - - - - - - - - -tutorial6 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tutorial 6 - More on Rules

- -

In this tutorial we will cover more details of rule construction.  - We will review rule element conditions, special rule elements and some - code region rule conditions.

- -

The analyzer we will start with in the this tutorial, is the one that - you created in Tutorial 3. If you are taking Tutorial 6 directly after - Tutorial 5, you will have to close myAnalyzer - before going through the steps below. (Select Close - Analyzer from the File Menu.)

- -

Rename the myAnalyzer - folder in:

- -

C:\apps\myAnalyzerFolder

- -

to myAnalyzer5.

- -

If you saved myAnalyzer3 at the end of Tutorial 3, you - can copy that and rename it to myAnalyzer.

- -

OR, if available, do the following. (If not available, - you'll need to go through Tutorials 1, 2, and 3, executing the yellow - arrow instructions.)  Unzip - the myAnalyzer3.zip located in - the release directory:

- -

C:\Program Files\TextAI\VisualText\docs\tutorial3\myAnalyzer3.zip

- -

and place the myAnalyzer folder - in the working directory:

- -

C:\apps\myAnalyzerFolder

- -

Recursion

- -

Launch VisualText and open the myanalyzer.ana - file located in

- -

C:\apps\myAnalyzerFolder\myAnalyzer

- -

or select File > Recent Analyzers - > c:\apps\myanalyzerfolder\myanalyzer

- -

In the Ana Tab, - double-click on the NP pass to - open it in the Workspace.

- -

Here are the NP rules created by the end of Tutorial 3:

- -

- -

Our current NP pass file will work for sentences in which we have a - maximum of two adjectives preceding a noun. However, for sentences in - which there are more than three adjectives in succession, such as ''He's - a big brown friendly dog." the rules in our NP file are insufficient. - Of course, if we wanted to, we could add another rule to our NP pass file - to account for this. But since there are many possibilities, our rule - file would get extremely bulky. A more efficient way to handle cases like - this is to use a 'recursive' pass.

- -

The recursive algorithm will apply the rules within each context node - (i.e., each node found by a NODES or PATH selector) until no more matches are found. - Let's create a recursive pass now to see how this works.

- -

We will create a new pass just after the POSS - pass in the analyzer sequence.

- -

In - the Ana Tab, click - on POSS, right-click and select - Add > New Pass. - (You'll see why we are placing a new pass file at this point in the analyzer - sequence in a moment.)

- -

 

- -

Set Pass Properties

- -

In the Pass Properties - dialog, name the pass recursiveAdj

- -

For Type, select Recursive. - Click OK.

- -

Notice the R - next to the new pass in the Ana Tab. - This lets you know that it as a recursive - pass. 

- -

Double click recursiveAdj - pass (pass #10) to open it in the Workspace. 

- -

Edit the pass file so that it looks like this:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES - _ROOT

- -

 

- -

@RULES

- -

_adj <- _adj _xWHITE - _adj @@

- -

Your Ana Tab should look like this:

- -

- -

This rule will recursively join adjacent adjectives (separated by a - single space), so that the dominating _adj - built by one instance of this rule matching becomes element one the next - time this rule is matched.  The recursion stops when no rules match. 

- -

The recursiveAdj pass is placed after the POSS pass - so that possessives can be collected into lists of adjectives. It precedes - the NP pass so that adjective lists can be collected into noun phrases. - This illustrates that multi-pass text analyzers are sensitive to the ordering - of passes.

- -

The rule matcher looks for context (that is, the dominating node, in - this case _ROOT ) and then tries - to match rules in the phrase of children under the context nodes.  - In this case, the rule matcher finds _ROOT, - then finds the first instance of '_adj - _xWHITE _adj'.  It reduces this to a single - _adj and then continues looking - within the _ROOT context node - for another match of the pattern.  It uses the newly created _adj node as the first element of the - rule and searches for an _xWHITE _adj to complete the rest of the rule.  - As you will see below in the tree structure, the leftmost '_adj - _xWHITE _adj' becomes the most deeply embedded - phrase.

- -

When the recursive pass can no longer match rules in a phrase, the rule - matcher tries to find (i.e. select - via NODES - or PATH) - another context node that matches. If there is another string of adjectives - later, the same algorithm applies.  When this finishes, the rule - matcher looks for another context node.  In this case, there is only - one context node _ROOT in the - parse tree, so the algorithm stops here.  In a slightly different - situation, where the context node was not _ROOT - but rather a node name that appeared more than once in the parse tree, - the recursive algorithm would apply to each of those context nodes. (This - selection mechanism is also used by the nonrecursive pat - algorithm.)

- -

What happens with a noop rule, - i.e., one that does consume a phrase of nodes to build a new dominating - node? The recursive algorithm pretends that a noop rule has failed and - keeps looking in the phrase for a "real" rule match, i.e., one - that modifies the parse tree. Had the noop rule match been counted a success, - the recursive algorithm would fail to terminate (i.e., it would be in - an infinite loop) whenever it matched a noop rule. (There are several - other ways to get into an infinite grammar loop, as you may discover!)

- -

Simplify NP Rule

- -

Look at the sequence of passes in our analyzer. Notice that recursiveAdj - occurs before the NP pass. Since - the analyzer goes from pass to pass along the chain, instances of multiple - adjectives occurring together will be handled by the recursiveAdj - pass making some of the rules in our NP pass file redundant.

- -

Edit the NP pass - file so that the first rule (the rule highlighted below) is deleted from - the file:

- -

- -

To test our recursive adjective rule, let's create a new text file with - lots of adjectives.

- -

In the Text Tab - window, right click and select Add > - New Text File.

- -

Name the file eg2.txt. Add the following sentence:

- -

John gave the quick little brown fox's baby's yellow book to Mary.

- -

Click the Run - button to run the analyzer.

- -

Right click on eg2.txt - in the Workspace and select View > - Parse Tree.

- -

Under the second _NP - node, expand the _adj node to - see the effect of the recursive algorithm:

- -

- -

A Better List

- -

There's a way to build a more efficient and compact list, such as the - list of adjectives that we have built up to this point. Also, to reflect the difference between a lone adjective - and an adjective list, we'll create a new nonliteral, - or abstract name, called _adjlist.

- -

Edit the recursiveAdj - pass file to look like:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES - _ROOT

- -

 

- -

@RULES

- -

#_adj - <- _adj _xWHITE _adj @@

- -

_adjlist [base] - <- _adj @@

- -

 

- -

@POST

- -

  listadd(1,3, - "false");

- -

@RULES

- -

_xNIL - <- _adjlist _xWHITE - _adj @@

- -

The first rule (_adjlist - [base] - <- _adj) builds an adjective list from one adjective. The second - rule (_xNIL <- adjlist _xWHITE - _adj) uses a special reduce action called - listadd in order to extend an adjective list with - one adjective node. Rather than building a new list node, listadd places the new _adj - node under the list of nodes already collected under the _adjlist - node. The arguments to listadd are

- -

listadd(<element - number of the list node>, <element number of the new element to - add to list>, <true/false to also collect nodes between the list - node and new element node.)

- -

The base attribute - in the first rule keeps the recursive algorithm from finding _adj - within _adjlist and getting stuck - in an infinite loop. It's not strictly necessary in the rule as written, - but would be necessary if written as

- -

_adjlist [base] - <- _adj [s] @@

- -

where the element - action s (for singlet) leads to finding _adj even when it is buried - under _adjlist.

- -

If not enabled, select the Toggle - Generate Logs button on the toolbar.

- -

Run the analyzer on eg2.txt.

- -

Select the recursiveAdj - pass in the analyzer sequence. Then right click on eg2.txt - in the Workspace and View > Parse - Tree. - Notice how all of the adjectives - are now part of _adjlist.

- -

- -

This version of the adjective list is more practically useful than the - first form we built. It's more compact, easier to read, faster and easier - to traverse. The philosophy underlying NLP++ is to support similar engineering - of text analyzers in enhanced ways wherever feasible.

- -

To integrate _adjlist - into the analyzer, edit the first _NP - rule in the NP pass file to use - _adjlist instead of _adj.

- -

The first rule in your NP pass file should look like this:

- -

_NP <- _det _xWHITE _adjlist - _xWHITE - _noun @@

- -

 

- -

Optional Rule Elements

- -

You just learned how to create a recursive pass file. We made the recursiveAdj - pass file so that we could account for multiple adjectives such as 'the - big brown friendly dog' without having to write individual rules to cover - the number of different possible ways adjectives can be used in sentences. - One other way to optimize rules is by using the optional element modifier [opt]. - Though it is possible to specify every variant of the same basic rule, - we can cut down on the number of rules in a pass file by using [opt]. Some might argue that optional rule elements - make the rules less comprehensible. Nevertheless, they can be useful for - development, and in some situations may even help to clarify the rules.

- -

For example, the first rule of our NP pass file has two optional elements, - _adjlist (an adjective) and _xWHITE (a white space). - You can rewrite the rule to indicate that these two elements are optional - like this:

- -

NP <- _det _xWHITE _adjlist - [opt] _xWHITE - [opt] _noun @@

- -

Edit the first rule of your NP - pass file using the [opt] element modifier.

- -

While we are at it, go ahead and eliminate the second - rule in the NP pass file since - the rule is now redundant with the first rule. That is, the rule _NP <- _det - _xWHITE _noun - is now contained or covered by the first rule.

- -

Your NP pass file should look like this:

- -

- -

Optional elements that place conditions on rule elements such as [opt] must be contained in square brackets. See - Phrase Element - Modifiers for more information on [opt] -  and other - modifiers that can be used with rule elements.

- -

Special Rule Elements - Variables

- -

We've already seen some of the conditions that the automatic rule generator - uses.

- -

For example, in the verb rule below:

- -

- -

there's a list of conditions in the square brackets.

- -

 _xWILD - is a special rule element, and very useful.  It behaves like - any other rule element, such as _NP - or _verb, in the sense that the - text of that element can be extracted, or the node operated on, but special - rule elements have properties that make writing rules easy.

- -

The rule element _xWILD is an unrestricted - wildcard and will match zero or more tokens.  For example, in a rule - like

- -

_foo <- four _xWILD seven @@

- -

a phrase of nodes in the parse tree corresponding - to 'four five six seven' would match and the new node _foo would be created.

- -

In the rule

- -

_verb <- _xWILD [min=1 max=1 s match=("gave" "jumped" "kissed")] @@

- -

what _xWILD can match is specified - in a match list. The list of possible matches for _xWILD - are gave, jumped - and kissed.

- -

_xWILD is a unique rule element in that what - it matches depends on the element following it in a rule. By contrast, - _xANY (which - matches any single node) in the following rule

- -

_ABC <- PQR _xANY [min=1 - max=5] - XYZ @@

- -

will greedily gobble up any - XYZ node that it encounters, possibly failing though it "could" - have matched. Only _xWILD knows how to "back - up" to make such a rule match succeed.

- -

 

- -

Match List

- -

If the match list were absent, this rule would match any single node.  - Items in the match list can be literals, as is the case in this rule, - but they can also be nonliterals, such as _NP - or _noun.  - min=1 - means match a minimum of 1 element. min=0 - would mean the current element is optional, and is the same as using opt. max=1 - means match a maximum of 1 element. max=0 - means the current element can match an indefinite number of nodes.  - In addition to the match list, there is also a fail - list, which uses the same syntax ('fail=("X", "Y", - "Z")') but means fail to match the rule if any of the listed - items are present in the input. The element modifiers fail and match are - further discussed in the Phrase - Element Modifiers topic.

- -

Special Rule - Elements

- -

Here is a list of the special rule elements in NLP++.

- -

 

- - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Rule Element

 -

Description

-

_xWILD

 -

Matches any nodes and any number of nodes. - An intelligent matcher that is "back up aware."

-

_xANY

 -

Matches any single token. A dumb or greedy - matcher that can't "back up" the way _xWILD - can.

-

_xALPHA

 -

Matches an alphabetic - token.

-

_xNUM

 -

Matches a numeric token.

-

_xPUNCT

 -

Matches a punctuation - token.

-

_xWHITE

 -

Matches a whitespace - token, including newline.

-

_xBLANK

 -

Matches a whitespace - token, excluding newline. Equivalent to _xWILD [match=(\ - \t)]

-

_xCAP

 -

Matches an alphabetic - with uppercase first letter.

-

_xEOF

 -

Matches the end of file.

-

_xSTART

 -

Matches if at the start - of a phrase in the parse tree.

-

_xEND

 -

Matches if at the end - of a phrase in the parse tree.

- -

We'll use _xALPHA in our next example.

- -

Pass Regions

- -

In the previous tutorials, you were introduced to some of the different - regions of a pass file, namely POST and RULES. We also briefly saw what the PRE Region does in Tutorial 4. These different - regions are used to control the application of rules in a pass file. Here - we will summarize these three regions and also introduce a new region, - the CHECK Region. - These are the main regions used in pass files.

- - - -

Each region has a set of begin and end markers for - it: @PRE and @@PRE, @CHECK and @@CHECK, @POST and @@POST, @RULES and @@RULES. You may - wonder why we have not been using the end markers in the pass files we - have been editing. The use of end markers is optional when writing pass - files. Some people like using the end markers to make - the pass file easier to read. This is a matter of personal preference.

- -

 In the next example, we will use the PRE Region to control the - application of a rule. The PRE Region along with CHECK, POST and RULES - make up the Grammar Region of a pass file. See the Grammar - Region topic for more information.

- -

 

- -

Finding Unknown Acronyms - PRE Region

- -

The PRE Region places conditions - on individual rule elements and operates during the rule matching of a - rule. Suppose we want to add to our system the capability to categorize - unknown words.  Given the sentence, 'The XZY - took the QRS - to LMNO" - we might infer that the capitalized phrases, XZY, - QRS and LMNO - were nouns.  (Basically, any uppercase word of a certain size is - likely to be an acronym, and a noun.) We can set up the conditions in - the PRE Region so that words in all - capital letters are identified as nouns. To illustrate this, we will have - to create another pass in the analyzer sequence. We'll place it before - the POSS pass, since we're building nouns that can potentially be used - in possessive constructs (e.g., "IBM's software").

- -

In the Ana - Tab, create a new pass - just before the - POSS pass. (Select end_words. Right click, Add > New Pass)

- -

Name the pass, Acronyms. - Leave Type: as Pattern and click OK.

- -

Double click the Acronym - pass and edit so that it looks like this:

- -

@CODE

- -

L("hello") = 0;

- -

@@CODE

- -

 

- -

@NODES - _ROOT

- -

 

- -

@PRE

- -

  <1,1> - uppercase();

- -

  <1,1> - lengthr(3,10);

- -

@RULES

- -

_noun <- _xALPHA - @@

- -

Let's look at the conditions in the PRE Region.

- -

<1,1> uppercase(); - means 'from the first element to the first element of the rule, the elements - must be all uppercase in order for this rule to match.' 
-<1,1> lengthr(3,10); means from the 'first element - to the first element, the element must be a minimum of three characters - long and a maximum of 10.' 

- -

In other words, if _xALPHA - (the first element of our rule) is composed of uppercase letters, and - is between three and ten characters long, create a dominating node named - _noun.

- -

Both uppercase and lengthr are special NLP++ functions - and are called "PRE Actions". You'll learn more about these - later.

- -

Running the analyzer on the sentence

- -

TAI gave VISUALTEXT to the WORLD.

- -

produces a parse tree like:

- -

- -

Notice how TAI, WORLD and VISUALTEXT have been identified as nouns.

- -

The PRE Region handles only special - functions (called PRE Actions). It - does not yet understand NLP++ expressions and statements.

- -

Close the files in the workspace and exit the analyzer.

- -

Summary

- -

In this tutorial we introduced the recursive - pass-algorithm and showed that it could handle an indefinitely long list - (e.g., of adjectives).  We looked at other features of rule construction, - such as making elements optional [opt].  - We reviewed special rule elements, showing _xWILD and _xALPHA in action, and finished - with some conditions placed on rule elements in the PRE - Region.
- 
- 
- 
- 
- 
- 

- - - - diff --git a/Help/helps/Unicode_Test.htm b/Help/helps/Unicode_Test.htm deleted file mode 100644 index c95d997..0000000 --- a/Help/helps/Unicode_Test.htm +++ /dev/null @@ -1,57 +0,0 @@ - - - - - - - - - - -Unicode_Test - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Unicode_Test

- -

NOTE: - Some users find a Microsoft library missing when they run VisualText Unicode. -  support@textanalysis.com - can help with any problems.

- -

Unicode_Test is an analyzer supplied with the release, at Program Files/TextAI/VisualText/apps. -  It provides - sample input files that can be loaded and processed to produce a tokenized - parse tree for various test languages.  VisualTextU.exe, - the version of VisualText that supports Unicode, should be invoked in - order to run this analyzer.

- -

 

- - - - diff --git a/Help/helps/Updating_Analyzers_and_KBs.htm b/Help/helps/Updating_Analyzers_and_KBs.htm deleted file mode 100644 index 33bd800..0000000 --- a/Help/helps/Updating_Analyzers_and_KBs.htm +++ /dev/null @@ -1,317 +0,0 @@ - - - - - - - - - - -Updating Analyzers and KBs - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Updating Analyzers and KBs

- -

This topic applies to the latest version of VisualText.  You - may want to read the topic Compiled - and Standalone Analyzers for background.

- -

In case you need to merge code from your old analyzers, we recommend - using WinDiff (available with - MS Visual Studio 6.0).  Also, - note the new include files referenced in the prog and user projects, as - well as code changes that compile with both Unicode and non-Unicode configurations - (see the API header file Program Files\TextAI\include\Api\my_tchar.h - as well as studying the new code files).  Finally, - note that the prog, kb, run, and user projects now all include Unicode_Release - configurations which may need to be compiled for Unicode VisualText.

- -

WARNING: BEFORE - uninstalling an older VisualText and installing a later version, one should - do steps [1] and [2] below.

- -

LIMITATION: - At present, we have no simple method for converting a Unicode KB to an - ANSI KB.  You - may need to manually edit out all Unicode characters.

- -

[1] Get - interpreted KB.  For - each analyzer you want to update, you'll need to make sure it has an up-to-date - INTERPRETED form of the knowledge base.  This - is because compiled KBs (kb.dll) will not load with the new VisualText.

- -

Simplest is to click the Save KB button (or select Save - KB from the KB menu).  This - assures that you generate the latest interpreted version of a knowledge - base, whether you use a compiled analyzer or not.

- -

[2] Back - up analyzers.  Prior - to upgrading the VisualText version, make a safe copy of your analyzers. -  This will - give you a fallback in case something goes wrong during the update process.

- -

[3] Install - new VisualText.

- -

Follow the instructions provided to you with the latest - version of VisualText.

- -

[4] Create - new analyzer.  In - general, the best way to update an analyzer is to create a new version - from scratch, then merge parts of the old analyzer into the new.

- -

Rename the old analyzer, e.g. c:\apps\myAnalyzer to c:\apps\myAnalyzer_saf

- -

Create a new analyzer of the same name by invoking VisualText, - selecting File > New Analyzer, and selecting the Bare template.

- -

If you will use a compiled KB, check the box at File - > Preferences > Analyzer Dependent > Load Compiled KB. (Don't - worry, if the analyzer doesn't find a compiled KB it will still load the - interpreted KB.)

- -

Exit VisualText.

- -

Next we'll merge the old analyzer into the new, one piece at a time.

- -

[5] Merge - KB.  

- -

Copy the interpreted kb from your old analyzer, e.g., - at myAnalyzer_saf\kb\user.

- -

Paste this user folder into the new analyzer at myAnalyzer\kb.  (Yes - to overwrite everything).

- -

[6] Merge - Run.  The - analyzer's \run folder is automatically - managed by VisualText, so don't muck around in here!  Once - it's updated, you can manually compile the analyzer in here. (Professional - version only.)

- -

[7] Merge Prog.  Prog - is a sample driver program that comes with an analyzer. (Usable in Professional - version only.)

- -

If you use the Prog driver, then follow instructions - in the Compiled - and Standalone Analyzers topic.

- -

If you've made your own modifications, you'll need to - manually merge \myAnalyzer_saf - files into \myAnalyzer.

- -

Compile the Release or Unicode_Release configuration, - as appropriate.

- -

[8] Merge - User Project.  An - analyzer's user project is where extensions to the NLP++ language can - be added, third party code can be integrated, and so on.

- -

If you haven't modified the user proejct, then you're - done.

- -

If you have modified the user project, you'll need to - merge your code from myAnalyzer_saf\user - with myAnalyze\user.

- -

Compile the Release or Unicode_Release configuration, - as appropriate.

- -

NOTES:

- -

[1] In 2.0.1.x and later, various analyzer DLLs have been moved and - renamed.  In - most cases, you'll find them in the \bin - folder of a project, with names to distinguish Unicode from non-Unicode - versions.

- -

[2] Non-Unicode analyzers should be amenable to loading in the Unicode - VisualText.  However, - Unicode analyzers will not load correctly in the ANSI VisualText.

- -

[3] We recommend using a particular analyzer with either ANSI VisualText - or Unicode VisualText, not both.  Once - you start placing Unicode characters in analyzers and knowledge bases, - you can't then run them in the ANSI VisualText.

- -

[4] We recommend naming analyzers according to whether they are Unicode - or ANSI.

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Analyzer_Sequence/Hand-Built_Pass_Files/Adding_passes.htm b/Help/helps/Using_VisualText/Analyzer_Sequence/Hand-Built_Pass_Files/Adding_passes.htm deleted file mode 100644 index 2de32ff..0000000 --- a/Help/helps/Using_VisualText/Analyzer_Sequence/Hand-Built_Pass_Files/Adding_passes.htm +++ /dev/null @@ -1,365 +0,0 @@ - - - - - - - - - - -Adding passes files - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Adding Passes

- -

When you add a pass to the analyzer sequence you are also creating the - pass file that is associated with the pass. You can associate existing - pass files from another analyzer project with a new pass in the analyzer - sequence.  

- -

VisualText also comes with a set of library - passes that can be added to the analyzer sequence.  Library - passes are pre-built passes that contain code to recognize telephone numbers, - email addresses, dates, money and other common patterns. Some library - passes are available as VisualText analyzer templates.  

- -

To add a new pass - to the analyzer sequence:

- -
    - -

  1. In the Ana Tab, - click on the pass in the sequence where you want to insert the pass. The - new pass will be added AFTER the pass you select.

    2. - -

  2. With the mouse in the Ana Tab, right-click and - select Add > New - Pass.

    3. - -

  3. In the Pass Properties - dialog box, give the pass file a name. (For pass names with multiple words, we - recommend using an underscore between the words.)  For - Type, - select Pattern for a pattern based pass or Recursive for a recursive one. - By default the pass is set with the Active flag, meaning the pass will be processed - the next time the analyzer is run. (To inactivate the pass, see Inactivating - a pass.)

    4. - -

  4. Click OK to finish creating the pass. The newly pass - is added to analyzer sequence.

    5. -
- -

To associate - an existing pass file with a pass in the analyzer sequence:

- -
    - -

  1. In the Ana Tab, - click on the pass in the sequence where you want to insert the pass. The - new pass will be added AFTER the pass you select.

    2. - -

  2. Right-click and select Add - > Pass from File.

    3. - -

  3. Navigate to the directory where the pass file - exists.  (For - example, c:\apps\myAnalyzer\spec.)

    4. - -

  4. Select the pass - file and click Open.  The - pass is added to the analyzer sequence.  The - name of the pass defaults to the name of the existing pass file.

    5. -
- -

To add a library - pass to the analyzer sequence:

- -
    - -

  1. In the Ana Tab, - click on the pass in the sequence where you want to insert the pass. The - new pass will be added AFTER the pass you select.  If - a library pass contains more than one pass in the sequence, all passes - are added to the Ana Tab.  The - number following the library pass name indicates the total number of passes - in the sequence.

    2. - -

  2. Right-click and select Add - > Library Pass.

    3. - -

  3. Select - the pass from the menu.

    4. -
- - - -- - - - - - -
-

Note: -  You must - change the Pass Properties on some library passes in order for them to - work correctly.  By - default, the Pass Property Type is set to Pattern.  Pattern - should be changed to Recursive - for the following:

-

 

- - --- - - - - - - - - - - - - - - - - - - - -
-

Library Pass

 -

Pass Name

-

Date Plus Numerics

 -

RecursiveCharPatterns

-

NumericLists

-

HTML

 -

HTML_TagStructure

-

Xml

 -

XML_DeclarationConstructors2

-

XML_TagStructure

-

XML DTD

 -

XML_DTD_DeclarationConstructors2

-

 

- -

 

- -

 

- - - - - - -
-

Tip: - You can add pass files to TextAI\VisualText\data\spec directory located - in the install directory to create your own set of library of passes. -

- -

 

- -

See Also

- -

Editing - Pass Files

- -

Moving - Passes

- -

Analyzer - Templates

- -

 

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Analyzer_Sequence/Hand-Built_Pass_Files/Editing_pass_files.htm b/Help/helps/Using_VisualText/Analyzer_Sequence/Hand-Built_Pass_Files/Editing_pass_files.htm deleted file mode 100644 index a0cd072..0000000 --- a/Help/helps/Using_VisualText/Analyzer_Sequence/Hand-Built_Pass_Files/Editing_pass_files.htm +++ /dev/null @@ -1,112 +0,0 @@ - - - - - - - - - - -Editing pass files - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Editing Pass Files

- -

To edit a pass file associated with - a pass in the analyzer sequence:

- - - - - - - - -
-

Tip: - You can also open the pass file by selecting the pass in the sequence, - right-clicking and selecting Edit - > Pass File.

- - - - - - -
-

Note: - For help writing rules, refer to the NLP++ section.

- -

See Also

- -

Moving - Passes

- -

Inactivating - Passes

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Analyzer_Sequence/Hand-Built_Pass_Files/Moving_passes.htm b/Help/helps/Using_VisualText/Analyzer_Sequence/Hand-Built_Pass_Files/Moving_passes.htm deleted file mode 100644 index eed090e..0000000 --- a/Help/helps/Using_VisualText/Analyzer_Sequence/Hand-Built_Pass_Files/Moving_passes.htm +++ /dev/null @@ -1,134 +0,0 @@ - - - - - - - - - - -Moving passes - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Moving Passes

- -

Passes, whether they are a part of a stub region or not, can be moved - around in the analyzer sequence.

- -

To move a pass in the analyzer sequence:

- -
    - -

  1. In the Ana Tab, - select the pass you want to move.

    2. - -

  2. Click on the yellow up - or down arrows on the Tab Toolbar, to move the pass to the desired - location.

    3. -
- - - - - - -
-

Note:

-
    - -

  • Once the analyzer sequence has been changed, you - will have to run the analyzer again for the change in the sequence to - take.

    • - -

  • Passes in stub regions can only be moved around - in the stub region. Once the order of the passes in the stub has been - changed you must regenerate the rules before running the analyzer again.

    • - -

  • You can move a stub - region by selecting the stub (either the start stub bead or the end of - stub bead) and using the up and down arrows.

    • -
- -

 

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Analyzer_Sequence/Opening_Pass_Files.htm b/Help/helps/Using_VisualText/Analyzer_Sequence/Opening_Pass_Files.htm deleted file mode 100644 index 8105421..0000000 --- a/Help/helps/Using_VisualText/Analyzer_Sequence/Opening_Pass_Files.htm +++ /dev/null @@ -1,83 +0,0 @@ - - - - - - - - - - -Opening Pass Files - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Opening Pass Files

- -

To open a pass file associated with - a pass in the analyzer sequence:

- - - - - - - - -
-

Tip: - You can also right-click and select Edit - Pass File from the pulldown menu.

- -

See Also

- -

Editing Pass - Files

- - - - diff --git a/Help/helps/Using_VisualText/Analyzers/Closing_analyzers.htm b/Help/helps/Using_VisualText/Analyzers/Closing_analyzers.htm deleted file mode 100644 index 2c3420b..0000000 --- a/Help/helps/Using_VisualText/Analyzers/Closing_analyzers.htm +++ /dev/null @@ -1,88 +0,0 @@ - - - - - - - - - - -Closing analyzers - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Closing Analyzers

- -

To close an analyzer:

- - - - - - - - -
-

Tip: - You can also close the currently loaded analyzer by selecting Open Analyzer, navigating to another - analyzer project and opening the .ana file. You can also close an analyzer - by selecting an analyzer from the drop-down list under Recent - Analyzers.

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Analyzers/Creating_analyzers.htm b/Help/helps/Using_VisualText/Analyzers/Creating_analyzers.htm deleted file mode 100644 index f6c12ca..0000000 --- a/Help/helps/Using_VisualText/Analyzers/Creating_analyzers.htm +++ /dev/null @@ -1,82 +0,0 @@ - - - - - - - - - - -Creating Analyzers - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Creating Analyzers

- -

To create an analyzer:

- -
    - -

  1. From the File - menu, select New Analyzer.

    2. - -

  2. In the New Analyzer - dialog, give the analyzer a name. - This name will be used to organize the analyzer project. VisualText automatically - assigns a .ana extension to - the name to mark it as the analyzer project file. You use the .ana file - to open the analyzer project.

    3. - -

  3. Specify the drive and folder where you want the - analyzer project and its files to be created in Location. - The application directory defaults - to c:\apps. (You can change the default application directory in the Preferences - dialog.)

    4. - -

  4. For Template, - select one of the templates under Type. -  (For more - on templates, see Analyzer - Templates.)

    5. - -

  5. Click OK - to finish creating the analyzer project.

    6. -
- -

 

- -

See Also

- -

Opening analyzers
-Structure of analyzer projects
-Setting VisualText Preferences

- - - - diff --git a/Help/helps/Using_VisualText/Analyzers/Opening_analyzers.htm b/Help/helps/Using_VisualText/Analyzers/Opening_analyzers.htm deleted file mode 100644 index 19801df..0000000 --- a/Help/helps/Using_VisualText/Analyzers/Opening_analyzers.htm +++ /dev/null @@ -1,142 +0,0 @@ - - - - - - - - - - -Opening analyzers - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Opening Analyzers

- -

There are several ways to open an analyzer project.

- -

To open an analyzer, using Open Analyzer:

- -
    - -

  1. From the File - menu, select Open Analyzer.

    2. - -

  2. In the Open - dialog, navigate to the drive and the folder containing the .ana file.

    3. - -

  3. Select the .ana - file and click Open.

    4. -
- -

To open an analyzer, using Recent Analyzers:

- -
    - -

  1. From the File - menu, select Recent Analyzers.

    2. - -

  2. Select the analyzer name from the drop-down list. - (Analyzer names in the drop-down list do not contain the .ana extension.)

    3. -
- - - - - - -
-

Note: If you already - have an analyzer open, using Open Analyzer or Recent Analyzers will close the currently loaded analyzer. If the current - analyzer has not been saved, you will be prompted to save the analyzer - before the analyzer is loaded.

- - - - - - -
-

Tip: You can also set - the preferences to open the last analyzer loaded when the program is launched - by checking "Auto Load Last Analyzer" in the General Tab of - the Preferences box.

- -

See Also

- -

Load - Compiled KB Preference

- -

Saving - Analyzers

- -

Closing - Analyzers

- - - - diff --git a/Help/helps/Using_VisualText/Analyzers/Saving_analyzers.htm b/Help/helps/Using_VisualText/Analyzers/Saving_analyzers.htm deleted file mode 100644 index 4772fda..0000000 --- a/Help/helps/Using_VisualText/Analyzers/Saving_analyzers.htm +++ /dev/null @@ -1,62 +0,0 @@ - - - - - - - - - - -Saving analyzers - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Saving Analyzers

- -

To save an analyzer:

- -
    - -

  1. Select Save Analyzer - from the File menu. Save Analyzer - saves the analyzer project and the associated Knowledge Base.

    2. - -

  2. Answer Yes - to the prompt asking if you want to save the analyzer.

    3. -
- -

See Also

- -

Closing analyzers
-Opening analyzers

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_Label_Concepts.htm b/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_Label_Concepts.htm deleted file mode 100644 index 93e72e7..0000000 --- a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_Label_Concepts.htm +++ /dev/null @@ -1,89 +0,0 @@ - - - - - - - - - - -Adding Label Concepts - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Adding Label Concepts

- -

Label concepts hold subsamples of larger concepts.  For - example, the area code or exchange of a telephone number can be a subsample - of a telephone number.  

- -

To add a label concept to a rule concept:

- -
    - -

  1. In the Gram Tab, - select the rule concept where you want to add the label concept.

    2. - -

  2. Right-click and select Add - > Label.

    3. - -

  3. Give the label concept a name - in the dialog box.

    4. - -

  4. Click OK - to finish adding the label concept.

    5. -
- - - - - - -
-

Note: - Label concepts can only be added to rule concepts.  

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_folders.htm b/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_folders.htm deleted file mode 100644 index 3977c6c..0000000 --- a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_folders.htm +++ /dev/null @@ -1,130 +0,0 @@ - - - - - - - - - - -Adding folders - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Adding Folder Concepts

- -

Folder concepts are used to organize multiple types of samples.

- -

To add a folder concept to a stub:

- -
    - -

  1. In the Gram Tab, - select the stub concept where you want to insert the folder concept.

    2. - -

  2. Right click and select Add - > Folder.

    3. - -

  3. Give the folder concept a name - in the dialog box.

    4. - -

  4. Click OK - to finish adding the folder concept.

    5. -
- - - - - - -
-

Note: Folder concepts can be added to stub - concepts.

- -

 

- - - -- - - -
-

Tip:

-
    - -

  • Add multiple folder - concepts at the same time separating the names with commas.

    • - -

  • Use - the yellow up and down arrows on the Tab Toolbar to move your folder concepts.

    • -
- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_rule_concepts.htm b/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_rule_concepts.htm deleted file mode 100644 index 624f954..0000000 --- a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_rule_concepts.htm +++ /dev/null @@ -1,150 +0,0 @@ - - - - - - - - - - -Adding rule concepts - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Adding Rule Concepts

- -

Rule concepts own a set of related samples, from which the rule generator - will generalize, merge and generate rules.

- -

To add a rule concept to a stub:

- -
    - -

  1. In the Gram Tab, - select the stub where you want to add the rule concept.

    2. - -

  2. Right click and select Add - > Rule.

    3. - -

  3. Give the rule concept a name - in the dialog box.

    4. - -

  4. Click OK - to finish adding the rule concept.

    5. -
- - - - - - -
-

Note: Rule concepts can be added to stub - and folder concepts.

- -

 

- - - - - - -
-

Tip:

-
    - -

  • Add - multiple rule concepts at the same time separating the names with commas.

    • - -

  • Use - the yellow up and down arrows on the Tab Toolbar to move your rule concepts.

    • - -

  • Organize your rule concepts by using folders - to the Gram Tab.

    • -
- -

 

- -

See Also

- -

Setting Concept Properties

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_samples_to_rule_concepts.htm b/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_samples_to_rule_concepts.htm deleted file mode 100644 index cbcc8a1..0000000 --- a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_samples_to_rule_concepts.htm +++ /dev/null @@ -1,181 +0,0 @@ - - - - - - - - - - -Adding samples to rule concepts - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Adding Samples to Rule Concepts

- -

There are several ways to add samples to a rule concept. You can add - samples to rules by highlighting example text from a text file in the - Workspace, or by importing lists of samples from an existing text file.

- -

To add samples using a text file in - the Workspace:

- -
    - -

  1. Double click on the text file - in the Text Tab to open it in - the Workspace.

    2. - -

  2. In the Gram Tab, - select the rule concept you want - to add samples to. The arrow in front of the rule concept turns blue.

    3. - -

  3. In the text file, highlight the word - (or phrase) that corresponds - to the rule concept.

    4. - -

  4. Right click and select the rule - concept name listed in the pulldown menu. The sample is added below - and indented to the right of the rule concept.

    5. -
- - - - - - -
-

Tip: - If you are using folders to organize your rule concepts, select the folder instead of the individual rule concept to see all the - rule concepts in the pulldown menu when you highlight text.

- -

Adding samples already contained in a file is a fast and easy way to - add samples to a rule concept. This is particularly helpful when you have - lists of names or other such data.

- -

To add samples by importing a text - file:

- -
    - -

  1. In the Gram Tab, - select the rule concept you want - to add samples to. The arrow in front of the rule concept turns blue.

    2. - -

  2. Right-click and select Sample - File > Import.

    3. - -

  3. VisualText will warn you that importing samples - from a file will delete any samples you may have for the rule concept. - Since you don't have any samples for the rule concept yet, click Yes. The warning is for those who already - have samples for the rule concept.

    4. - -

  4. In the Open - dialog, navigate to the directory containing the file of samples.

    5. - -

  5. Select the file - you want to import and click Open. - The samples are added below and indented to the right of the rule concept.

    6. -
- - - - - - -
-

Note: Each line of your text file must - contain one sample only.

- - - - - - -
-

WARNING: - Make sure all edits to be made from a text file you are sampling have - been made to the file BEFORE you - start sampling. Once you start sampling the file, you can't edit it.

- -

 

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_stub_regions.htm b/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_stub_regions.htm deleted file mode 100644 index 9269168..0000000 --- a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Adding_stub_regions.htm +++ /dev/null @@ -1,104 +0,0 @@ - - - - - - - - - - -Adding stub regions - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Adding Stub Concepts

- -

The stub concept is the primary organizer of user-highlighted samples.

- -

To add a stub concept:

- -
    - -

  1. In the Ana Tab, - select the pass in the analyzer sequence where you want the automatically - generated rules to be inserted.

    2. - -

  2. In the Gram Tab, - right click and select Add > - Stub.

    3. - -

  3. Give the stub concept a name - in the dialog box.

    4. - -

  4. Click OK - to finish adding the stub concept.

    5. -
- - - - - - -
-

Note: -

-
    - -

  • The newly added stub - concept is added to the analyzer sequence in the Ana Tab as a stub - region. The stub region is indicated by stub_name - and ends with end_name.

    • - -

  • You can add a stub - region directly to the Ana Tab by highlighting the pass where you want - the region to be inserted, right-clicking and selecting Add - > Stub from the context menu. -  The stub - concept is added to the Gram Tab.

    • -
- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Running_the_Rule_Generator.htm b/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Running_the_Rule_Generator.htm deleted file mode 100644 index 5083d19..0000000 --- a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Running_the_Rule_Generator.htm +++ /dev/null @@ -1,123 +0,0 @@ - - - - - - - - - - -Running the Rule Generator - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Running the Rule Generator

- -

To run the rule generator:

- -
    - -

  1. From the Analyzer - menu, select Generate Rules.

    2. - -

  2. Select All to - generate all rules as specified by samples in the Gram - Tab.

    3. -
- - - - - - -
-

Note: - The Dirty and Unlocked options are used - to generate rules for rule concepts that have the Dirty or Unlocked attribute - set. For more information on generating rules using Dirty or Unlocked, - see Generating - Rules.  See - Setting Rule Generation - Properties for more information on other attributes.

- - - - - - - -
-

WARNING: The edits - to automatically generated pass files will not be saved the next time - you try to regenerate the rules.  The - only ways to permanently modify automatically generated files are (1) - to add or edit samples in the Gram Tab and regenerate the rules, (2) to - move or copy the files out of a stub region.

- -

 

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Setting_concept_properties.htm b/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Setting_concept_properties.htm deleted file mode 100644 index 91b681d..0000000 --- a/Help/helps/Using_VisualText/Automatically_Generating_Pass_Files/Setting_concept_properties.htm +++ /dev/null @@ -1,62 +0,0 @@ - - - - - - - - - - -Setting Concept Properties - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Setting Concept Properties

- -

To set concept properties for rule - concepts:

- -
    - -

  1. In the Gram Tab, - select the rule concept.

    2. - -

  2. Right click and select Properties - from the pulldown menu. The Gram Concept Properties dialog is opened.

    3. - -

  3. Check or uncheck the boxes for the appropriate - attribute. (See Setting - Rule Generation Properties for a description of the attributes.)

    4. - -

  4. Click Done - to finish setting the properties.

    5. -
- - - - diff --git a/Help/helps/Using_VisualText/Hand_Building_Pass_Files/Inactivating_a_pass.htm b/Help/helps/Using_VisualText/Hand_Building_Pass_Files/Inactivating_a_pass.htm deleted file mode 100644 index 4ee15e0..0000000 --- a/Help/helps/Using_VisualText/Hand_Building_Pass_Files/Inactivating_a_pass.htm +++ /dev/null @@ -1,91 +0,0 @@ - - - - - - - - - - -Inactivating a pass - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Inactivating Passes

- -

To inactivate a pass in the analyzer - sequence:

- -
    - -

  1. Select the pass - in the Ana Tab.

    2. - -

  2. Right-click and select, Properties.

    3. - -

  3. In the Pass Properties dialog, uncheck the Active box. The current pass number - is indicated in the From: To: windows.

    4. - -

  4. Click OK - to finish inactivating the pass. The pass will show a break in the chain, - indicating that it has been inactivated.

    5. -
- - - - - - -
-

Note: - You can also inactivate other passes in the sequence at the same time. - Check the To End box to indicate - all passes from the current pass forward should be skipped. Check All to indicate ALL passes in the analyzer - sequence should be skipped.

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Importing_Databases_to_KB/Importing_Word_Lists_to_the_Dictionary_Hierarchy.htm b/Help/helps/Using_VisualText/Importing_Databases_to_KB/Importing_Word_Lists_to_the_Dictionary_Hierarchy.htm deleted file mode 100644 index 1a3074e..0000000 --- a/Help/helps/Using_VisualText/Importing_Databases_to_KB/Importing_Word_Lists_to_the_Dictionary_Hierarchy.htm +++ /dev/null @@ -1,81 +0,0 @@ - - - - - - - - - - -Importing Word Lists to the Dictionary Hierarchy - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Importing Word Lists to the Dictionary Hierarchy

- -

The Knowledge Base comes with a specialized concept called dict - (for dictionary) which stores information about words.  

- -

There are several ways to incorporate a word list into the dict hierarchy. -  One is - to create a simple analyzer to read in and add the list of words to the - hierarchy.  You - can also add a pass to an existing analyzer project in order to add the - word list.  The - pass can look like

- -

@POST
- L("text") = strtolower(N("$text"));
- addword(L("text"));
-@RULES
-_xNIL <-
- _xALPHA
- @@
 

- -

Note that you can copy a knowledge base folder to a minimal analyzer - having the pass above, run the analyzer on your word file, then Save KB. -  Then you - can copy the KB folder back to your main analyzer.  In - this way, knowledge bases can be conveniently moved around and modified - with data sources as they become available.  Some - analyzers can thus be dedicated to building or augmenting knowledge bases. -  (Things - are somewhat more complex with a compiled knowledge base.  Mainly, - you want to modify the analyzer preferences to load the interpreted knowledge - base, when making upgrades as described here.)

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/ParseTrees/ViewingConceptTree.htm b/Help/helps/Using_VisualText/ParseTrees/ViewingConceptTree.htm deleted file mode 100644 index 51a38a2..0000000 --- a/Help/helps/Using_VisualText/ParseTrees/ViewingConceptTree.htm +++ /dev/null @@ -1,154 +0,0 @@ - - - - - - - - - - -Viewing Concept Trees - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Viewing Concept Trees

- -

The concept tree shows you which nodes were built for a specific pass - in the analyzer sequence.  

- -

To view nodes created in a parse tree:

- -
    - -

  1. In the Ana Tab, right-click on a pass.

    2. - -

  2. Select View - > Concept Tree.

    3. -
- - - - - - -
-

Note: - To view concept trees, you must run the analyzer with the Generate - Logs button on.

- -

 

- - - - - - -
-

Tip: - You can also view a concept tree by selecting the View Concept Tree button - on the Debug Toolbar.

- -

See Also

- -

Viewing Parse - Trees

- -

Viewing - Intermediate Trees

- -

Viewing - Partial Trees

- -

Viewing - Parse Tree Matches

- -

Viewing - Final Trees

- - - - diff --git a/Help/helps/Using_VisualText/ParseTrees/ViewingFinalTrees.htm b/Help/helps/Using_VisualText/ParseTrees/ViewingFinalTrees.htm deleted file mode 100644 index 197a31e..0000000 --- a/Help/helps/Using_VisualText/ParseTrees/ViewingFinalTrees.htm +++ /dev/null @@ -1,98 +0,0 @@ - - - - - - - - - - -Viewing Final Trees - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Viewing Final Trees

- -

You can view the final parse tree after all the passes in the analyzer - sequence have been executed.

- -

To view the final parse tree:

- -
    - -

  1. In the Ana Tab, - right-click on a pass.

    2. - -

  2. Select View - > Final Tree.

    3. -
- -

See Also

- -

Viewing Parse - Trees

- -

Viewing - Concept Trees

- -

Viewing - Intermediate Trees

- -

Viewing - Partial Trees

- -

Viewing - Parse Tree Matches

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/ParseTrees/ViewingIntermediateTrees.htm b/Help/helps/Using_VisualText/ParseTrees/ViewingIntermediateTrees.htm deleted file mode 100644 index 312287a..0000000 --- a/Help/helps/Using_VisualText/ParseTrees/ViewingIntermediateTrees.htm +++ /dev/null @@ -1,158 +0,0 @@ - - - - - - - - - - -Viewing Intermediate Trees - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Viewing Intermediate Trees

- -

You can select a specific pass in the analyzer sequence and view the - cumulative parse tree for that pass. This is helpful when you want to - see how the each pass in the analyzer sequence is operating.

- -

To view an intermediate parse tree:

- -
    - -

  1. In the Ana Tab, select the pass in the analyzer sequence.

    2. - -

  2. Right-click and select View > Entire - Tree.

    3. -
- - - - - - -
-

Note: - To view intermediate parse trees, you must run the analyzer with the Generate Logs button on.

- - - - - - -
-

Tip: - You can also select the pass and - click on the View Parse Tree button in the Debug Toolbar.

- -

See Also

- -

Viewing Parse - Trees

- -

Viewing - Concept Trees

- -

Viewing - Partial Trees

- -

Viewing - Parse Tree Matches

- -

Viewing - Final Trees

- - - - diff --git a/Help/helps/Using_VisualText/ParseTrees/ViewingParseTree.htm b/Help/helps/Using_VisualText/ParseTrees/ViewingParseTree.htm deleted file mode 100644 index 48f7dfe..0000000 --- a/Help/helps/Using_VisualText/ParseTrees/ViewingParseTree.htm +++ /dev/null @@ -1,183 +0,0 @@ - - - - - - - - - - -Viewing Parse Trees - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Viewing Parse Trees

- -

There are several different ways to view parse trees of an analyzed - text.

- -

To view a parse tree using the Ana - Tab Popup menu:

- -
    - -

  1. In - the Ana Tab, select a pass in the analyzer sequence.

    2. - -

  2. Right - click and select View > Entire Tree.

    3. -
- -

To - view a parse tree using the Text File Popup menu:

- -
    - -

  1. Open the text file - in the Workspace.

    2. - -

  2. Right-click - and select View > Parse - Tree.

    3. -
- -

 

- - - - - - -
-

Note: -  Running - the cursor over lines in the parse tree display will display the lines - in a tooltip box.  Lines - which are too long to be displayed in full, are indicated with an ellipsis -  mark (...).  Clicking - on the tooltip with the left mouse button will display the entire truncated - text in a dialog box.  To - clear the dialog box, click on the "x" in the upper right-hand - corner.

- -

 

- - - - - - -
-

Tip: - You can also view the parse tree of the selected pass - by clicking on the View Parse Tree icon.

- -

See Also

- -

Viewing - Final Trees

- -

Viewing - Concept Trees

- -

Viewing - Intermediate Trees

- -

Viewing - Partial Trees

- -

Viewing - Parse Tree Matches

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/ParseTrees/ViewingParseTreeMatches.htm b/Help/helps/Using_VisualText/ParseTrees/ViewingParseTreeMatches.htm deleted file mode 100644 index 0d0e9d4..0000000 --- a/Help/helps/Using_VisualText/ParseTrees/ViewingParseTreeMatches.htm +++ /dev/null @@ -1,121 +0,0 @@ - - - - - - - - - - -Viewing Parse Tree Matches - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Viewing Parse Tree Matches

- -

You can select a specific pass in the analyzer sequence and view the - parse tree for that pass only.

- -

To view a partial parse tree:

- -
    - -

  1. In the Ana Tab, select the pass in the analyzer sequence.

    2. - -

  2. Right-click and select View > Tree - Matches

    3. -
- - - - - - -
-

Note: - To view the parse tree for a single pass, you must run the analyzer with - the Generate Logs button on.

- -

See Also

- -

Viewing Parse - Trees

- -

Viewing - Concept Trees

- -

Viewing - Intermediate Trees

- -

Viewing - Partial Trees

- -

Viewing - Final Trees

- - - - diff --git a/Help/helps/Using_VisualText/ParseTrees/ViewingPartialTrees.htm b/Help/helps/Using_VisualText/ParseTrees/ViewingPartialTrees.htm deleted file mode 100644 index 8de9d15..0000000 --- a/Help/helps/Using_VisualText/ParseTrees/ViewingPartialTrees.htm +++ /dev/null @@ -1,105 +0,0 @@ - - - - - - - - - - -Viewing Partial Trees - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Viewing Partial Trees from Text Files

- -

You can select a specific part of text in an analyzed file and view - the parse tree for that line of text only. This is helpful when you want - to see how the rule matcher worked on a particular part of text.

- -

To view a partial parse tree:

- -
    - -

  1. In - the Text Tab, - double click on the processed text to open it in the Workspace.

    2. - -

  2. Highlight - the text.

    3. - -

  3. Right-click and select - View > Tree - of Selected. The parse tree for the highlighted text opens in the - Workspace.

    4. -
- -

See Also

- -

Viewing Parse - Trees

- -

Viewing - Concept Trees

- -

Viewing - Intermediate Trees

- -

Viewing - Parse Tree Matches

- -

Viewing - Final Trees

- - - - diff --git a/Help/helps/Using_VisualText/Running_Analyzers/Running_Analyzer_on_Multiple_Files.htm b/Help/helps/Using_VisualText/Running_Analyzers/Running_Analyzer_on_Multiple_Files.htm deleted file mode 100644 index c4b4a8b..0000000 --- a/Help/helps/Using_VisualText/Running_Analyzers/Running_Analyzer_on_Multiple_Files.htm +++ /dev/null @@ -1,122 +0,0 @@ - - - - - - - - - - -Running Analyzer on Multiple Files - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Running Analyzer on Multiple Files

- -

You can run an analyzer on multiple files contained in a tree of folders.

- -

To run an analyzer - on a tree of folders:

- -
    - -

  1. Select the top level folder - in the Text Tab.

    2. - -

  2. Click on the Run - button in the Workspace Toolbar.

    3. -
- - - - - - -
-

Note: - Log files generated for the text are placed under (indented and to the - right of) the text file.

- - - - - - -
-

Tips: - You can also select Run from the Analyzer - Menu.

- -

 

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Running_Analyzers/Running_Analyzer_on_Single_File.htm b/Help/helps/Using_VisualText/Running_Analyzers/Running_Analyzer_on_Single_File.htm deleted file mode 100644 index 8f42ead..0000000 --- a/Help/helps/Using_VisualText/Running_Analyzers/Running_Analyzer_on_Single_File.htm +++ /dev/null @@ -1,167 +0,0 @@ - - - - - - - - - - -Running Analyzer on Single File - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Running Analyzers

- -

You can run the analyzer on files one at a time or you can run the analyzer - on multiple files contained in a tree of folders.

- -

To run an analyzer on a single file:

- -
    - -

  1. Select - the file in the Text Tab.

    2. - -

  2. Click - the Run button on the Workspace - Toolbar.

    3. -
- -

 

- -

To run an analyzer on a tree of folders:

- -
    - -

  1. Select the top level folder - in the Text Tab.

    2. - -

  2. Click the Run - button on the Workspace Toolbar.

    3. -
- - - - - - -
-

Note: -

-
    - -

  • Log files generated - for the text are placed under (indented and to the right of) the text - file.  To - generate log files, the Toggle Generate Logs button must - be selected before the analyzer - is run.

    • - -

  • A red checkmark on a - file in the Text Tab indicates the file has been analyzed.

    • -
- - - - - - -
-

Tip: - You can also select Run from the Analyzer - Menu.

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Text_Files/About_Text_Files.htm b/Help/helps/Using_VisualText/Text_Files/About_Text_Files.htm deleted file mode 100644 index e5ff348..0000000 --- a/Help/helps/Using_VisualText/Text_Files/About_Text_Files.htm +++ /dev/null @@ -1,57 +0,0 @@ - - - - - - - - - - -About Text Files - - - - - - - - - - - - - - - - - - - - - - - - - - - -

About Text Files

- -

This "section" describes ways to manage text files, primarily - within the Text Tab.

- -

Plain text files are usually suffixed with .txt, while HTML files are - suffixed with .htm or .html and are displayed with a globe icon in the - text tab.

- -

When a text in the Text Tab has been analyzed, its icon gets a red check - mark.  A - plus sign appears to its left, indicating that log files, dump files, - and output files can be viewed under it.

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Text_Files/Adding_existing_text_files_to_a_project.htm b/Help/helps/Using_VisualText/Text_Files/Adding_existing_text_files_to_a_project.htm deleted file mode 100644 index 540cb17..0000000 --- a/Help/helps/Using_VisualText/Text_Files/Adding_existing_text_files_to_a_project.htm +++ /dev/null @@ -1,69 +0,0 @@ - - - - - - - - - - -Adding existing text files to a project - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Adding Existing Text Files to a Project

- -

If you already have text files to be analyzed, you can add them directly - to the analyzer input directory for processing.

- -

To add files to the analyzer project:

- -
    - -

  1. Using the file explorer, navigate to the location - of your text files.

    2. - -

  2. Select the files - you want to add to the analyzer project.

    3. - -

  3. Navigate to the input - folder of the analyzer project.

    4. - -

  4. Copy the - files into the input folder.

    5. - -

  5. To see the added files in the Text Tab, right-click - in the Text Tab and select Refresh.

    6. -
- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Text_Files/Adding_new_text_files_to_a_project.htm b/Help/helps/Using_VisualText/Text_Files/Adding_new_text_files_to_a_project.htm deleted file mode 100644 index d6c9a1f..0000000 --- a/Help/helps/Using_VisualText/Text_Files/Adding_new_text_files_to_a_project.htm +++ /dev/null @@ -1,105 +0,0 @@ - - - - - - - - - - -Adding new text files to a project - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Adding New Text Tiles to a Project

- -

There are several ways to add new text files to an analyzer project.

- -

To add a text file using the Text Tab:

- -
    - -

  1. In your analyzer project, click on the Text - Tab.

    2. - -

  2. With the mouse in the Text Tab, right-click and - select Add > New - Text File.

    3. - -

  3. In the dialog box, give the text file a name. - You do not need to add the .txt extension to the file.

    4. - -

  4. Click OK to finish creating the text file. The newly - created text file opens up in the Workspace. (The text file is stored - in the input - folder of the analyzer project.)

    5. -
- -

To add a text file using File > - New:

- -
    - -

  1. From the File - menu, select New. A blank text - file with a default VisualText title opens up in the Workspace. (You don't - have to be in the Text Tab.)

    2. - -

  2. To save the text file in your analyzer project, - right-click on the text file and select Save - As. Navigate to the location - in the input - directory for the current analyzer project where you would like to save - the text file.

    3. - -

  3. In the Save As - dialog, give the text file a name.

    4. - -

  4. Click Save - to finish creating the text file.

    5. - -

  5. To see the - file in the Text Tab, right-click in the Text Tab and select, Refresh.

    6. - -

  6. You will be - prompted with a warning saying that refreshing the Text Tab will cause - all knowledge associated with the text files in the Text Tab to be lost. - Answer Yes - if you are not concerned with losing any information.

    7. -
- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Text_Files/Deleting_text_files.htm b/Help/helps/Using_VisualText/Text_Files/Deleting_text_files.htm deleted file mode 100644 index 9ff9626..0000000 --- a/Help/helps/Using_VisualText/Text_Files/Deleting_text_files.htm +++ /dev/null @@ -1,77 +0,0 @@ - - - - - - - - - - -Deleting text files from a project - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Deleting Text Files from a Project

- -

To delete a text file from your project:

- - - - - - - - -
-

Tip: - Delete log files from the Text Tab to save space in your analyzer project.

- - - - diff --git a/Help/helps/Using_VisualText/Text_Files/Navigating_Among_Text_Files.htm b/Help/helps/Using_VisualText/Text_Files/Navigating_Among_Text_Files.htm deleted file mode 100644 index e2eefc9..0000000 --- a/Help/helps/Using_VisualText/Text_Files/Navigating_Among_Text_Files.htm +++ /dev/null @@ -1,123 +0,0 @@ - - - - - - - - - - -Navigating Among Text Files - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Navigating Among Text Files

- -

You can navigate among the different files in the Text Tab using the - Page Mode feature.  This - feature allows you to move from one file to the next file and is handy - when you want to quickly get to the next file listed in the Text Tab. -  

- -

To move from one file to the next in the - Text Tab:

- -
    - -

  1. Double-click on a - file to open it in the Workspace.

    2. - -

  2. Select the Page Mode Toggle    to - activate the yellow Previous Text and Next Text arrows.   

    3. - -

  3. Use the Previous - Text and Next Text arrows to open up files in the Workspace.  

    4. -
- - - - - - -
-

Note: - If a parse tree is displayed for the file, using the Previous Text and - Next Text arrows will display the parse tree display for the next text - in the list.

- -

 

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Text_Files/Opening_Text_Tab_Files.htm b/Help/helps/Using_VisualText/Text_Files/Opening_Text_Tab_Files.htm deleted file mode 100644 index dd47c55..0000000 --- a/Help/helps/Using_VisualText/Text_Files/Opening_Text_Tab_Files.htm +++ /dev/null @@ -1,90 +0,0 @@ - - - - - - - - - - -Opening Text Tab Files - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Opening Text Tab Files

- -

To open a file in the Text Tab:

- - - - - - - - -
-

Tip: - If the file icon in the - Text Tab is already highlighted, you can also click on the View Current - Text button on the Workspace Toolbar to open the file.

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Working_with_Debugging_Tools/DumpFiles.htm b/Help/helps/Using_VisualText/Working_with_Debugging_Tools/DumpFiles.htm deleted file mode 100644 index 75d219c..0000000 --- a/Help/helps/Using_VisualText/Working_with_Debugging_Tools/DumpFiles.htm +++ /dev/null @@ -1,182 +0,0 @@ - - - - - - - - - - -Using Dump Files - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Using Dump Files

- -

To debug your analyzer it is often helpful to output information to - dump files.  Dump - files are created by inserting a line of code in the desired pass file - indicating where to output the information.   

- -

To create a dump file:

- -
    - -

  1. In the Ana Tab, right-click on a pass file to open - it in the Workspace.

    2. - -

  2. In the pass file, add a printing statement using - the "<<" operator in the following form: "dump_file_name.txt" - << content_to_print.  Below - is a rule from the corporate analyzer that dumps the matched rule text - into a dump file called "anaphora":

    3. -
- -

- -

To view a dump file:

- -
    - -

  1. On the Debug - Toolbar, click on the arrow next to the View - Dump Files   icon.  

    2. - -

  2. Select - a dump file from the pulldown - menu.

    3. -
- -

 

- - - - - - -
-

Note:

-
    - -

  • Dump - files are created and housed in a system-created directory with a name - corresponding to the name of the text file in the project directory.

    • - -

  • The blue arrow icon on the Debug Toolbar is reserved for output files with the - name "output.txt" only.

    • - -

  • Dump files that do - not include a .txt extension will not show up in View Dump Files pulldown - menu.

    • - -

  • See the NLP++ section - for other ways to redirect data to output files.

    • -
- - - - diff --git a/Help/helps/Using_VisualText/Working_with_Debugging_Tools/PageMode.htm b/Help/helps/Using_VisualText/Working_with_Debugging_Tools/PageMode.htm deleted file mode 100644 index 144ed93..0000000 --- a/Help/helps/Using_VisualText/Working_with_Debugging_Tools/PageMode.htm +++ /dev/null @@ -1,127 +0,0 @@ - - - - - - - - - - -Using Page Mode - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Using Page Mode

- -

Page Mode allows you to automatically browse multiple windows simultaneously. - For example, if you've run the analyzer on a folder, you can then open - a window for the input file and output file, then browse among all the - input-output pairs using the left and right arrows on the Workspace Toolbar. - If a parse tree view is open, you can also browse the parse tree for each - pass of a single text analysis.

- -

To browse between multiple files in - an input folder:

- -
    - -

  1. Select - the Toggle Page Mode button on the Debug Toolbar.  This - activates the yellow Previous Text and Next Text buttons.

    2. - -

  2. Open - a file from an input directory.

    3. - -

  3. Select - the yellow Next Text button to - view the next file in the input directory.  Select - the Previous Text button to view - the last file.

    4. -
- -

To - update parse tree displays for passes in the analyzer sequence:

- -
    - -

  1. Select - the Toggle Generate Logs button - and run the analyzer.

    2. - -

  2. Click - on the View Parse Tree button - to display the parse tree.

    3. - -

  3. Select - the Toggle Page Mode button.

    4. - -

  4. In - the Ana Tab, click on a pass to update the parse tree display - in the Workspace.  

    5. -
- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Working_with_Debugging_Tools/Using_highlighting.htm b/Help/helps/Using_VisualText/Working_with_Debugging_Tools/Using_highlighting.htm deleted file mode 100644 index 1e80f4c..0000000 --- a/Help/helps/Using_VisualText/Working_with_Debugging_Tools/Using_highlighting.htm +++ /dev/null @@ -1,112 +0,0 @@ - - - - - - - - - - -Using highlighting - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Using Highlighting

- -

You can highlight regions of text in an analyzed file that match rules - in a selected pass of the analyzer sequence using Highlight Mode. To use - this highlighting feature, you must run the analyzer with Toggle Generate - Log button selected.

- -

To highlight regions in a text matching - a pass:

- -
    - -

  1. Select the Toggle - Generate Log button. -

    2. - -

  2. In the Text Tab, - double click on a text file to - open it in the Workspace.

    3. - -

  3. Run the analyzer.

    4. - -

  4. In the Ana Tab, - select a pass.

    5. - -

  5. Select the Toggle - Highlight Mode button. The text matching the pass - is highlighted in the text file. If a node has been created, the highlighting - is blue. If a node has matched, the highlighting is green.

    6. -
- - - - - - -
-

Note: - You can also highlight text in a file that match selected rule concepts - in the Gram Tab. To do this select the rule concept in the Gram Tab.

- -

 

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Working_with_Knowledge_Base/Opening_the_Phrase_Tool.htm b/Help/helps/Using_VisualText/Working_with_Knowledge_Base/Opening_the_Phrase_Tool.htm deleted file mode 100644 index 693305a..0000000 --- a/Help/helps/Using_VisualText/Working_with_Knowledge_Base/Opening_the_Phrase_Tool.htm +++ /dev/null @@ -1,150 +0,0 @@ - - - - - - - - - - -Opening the Phrase Tool - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Opening the Phrase Tool

- -

The Phrase Tool allows you to view the phrase of a concept in the Knowledge - Base.  You - can delete, add or edit nodes and their attributes using the Phrase Tool. -  Concepts - in the Knowledge Base with phrases are indicated by a yellow icon in the - Knowledge Base hierarchy.  

- -

To view the phrase for a concept in the Knowledge - Base:

- -
    - -

  1. In the KB - Editor, navigate to the concept - in the hierarchy with a yellow icon.  

    2. - -

  2. Right-click on the - icon and select Phrase - Tool.  The - Phrase Tool opens up in the Workspace.  

    3. -
- -

To delete a node from a phrase:

- -
    - -

  1. Select the node - in the list of nodes displayed in the Phrase Tool.

    2. - -

  2. Click the Delete - button.  

    3. - -

  3. At the confirmation prompt, answer Yes.

    4. -
- -

To add a node to a phrase:

- -
    - -

  1. In the Phrase Tool, click the Add - button.

    2. - -

  2. In the dialog box, enter a name - for the node and click OK. The - new node is added to the phrase.

    3. -
- -

To add attributes to a node:

- -
    - -

  1. Select the node - in the list of nodes displayed in the Phrase Tool.

    2. - -

  2. Click on the Attributes - button.  The - Attribute Editor is launched.

    3. - -

  3. To add an attribute to node, click Add - Attribute.  In - the Name column, provide a name - and hit the return key.  At - this time you can provide a value - for the name.  

    4. -
- -

To change the order of the nodes in - the phrase tool:

- -
    - -

  1. Select the node - in the list of nodes.

    2. - -

  2. Use the blue - arrow buttons to move the node to the desired location.

    3. -
- -

See Also

- -

Attribute - Editor

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/Using_VisualText/Working_with_Knowledge_Base/Using_the_KB_Editor.htm b/Help/helps/Using_VisualText/Working_with_Knowledge_Base/Using_the_KB_Editor.htm deleted file mode 100644 index 7847e0f..0000000 --- a/Help/helps/Using_VisualText/Working_with_Knowledge_Base/Using_the_KB_Editor.htm +++ /dev/null @@ -1,204 +0,0 @@ - - - - - - - - - - -Using the KB Editor - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Editing the Knowledge Base with the KB Editor

- -

The KB Editor displays the hierarchy of concepts in the Knowledge Base. -  You can - edit the Knowledge Base using NLP++ Functions or using the KB Editor. -  

- -

To open the KB Editor:

- - - - - - - - -
-

Tip: - You can also  access - the KB Editor from the KB Menu.

- -

To move a concept in the Knowledge - Base hierarchy:

- -
    - -

  1. Select the concept - you want to move.

    2. - -

  2. Using - the Move Concept Up and Move Concept Down button on the Workspace toolbar, move the concept in the - hierarchy.

    3. -
- - - - - - -
-

Tip: - You can also  move - concepts using the Move Concept Up or Move Concept Down options in the - KB Menu.

- -

 

- -

 

- - - - - - -
-

Note: -

-
    - -

  1. A - concept can only be moved within its own subhierarchy.  For - example, you cannot move a child concept so that it becomes the parent - concept.

    2. - -

  2. A - concept should not be added to the dict hierarchy using the KB Editor. -  To add - a concept to the dict hierarchy, use NLP++ functions, or the add feature - on the Attribute Editor.

    3. -
- -

To edit the attributes of a concept - in the KB:

- -
    - -

  1. Select a concept - in the KB.

    2. - -

  2. Right-click and select Attributes - from the menu.  This - will bring up the Attribute Editor.  See - Attribute - Editor Description for details on the Attribute Editor buttons.

    3. -
- - - - diff --git a/Help/helps/Using_the_Attribute_Editor.htm b/Help/helps/Using_the_Attribute_Editor.htm deleted file mode 100644 index f235206..0000000 --- a/Help/helps/Using_the_Attribute_Editor.htm +++ /dev/null @@ -1,101 +0,0 @@ - - - - - - - - - - -Using the Attribute Editor - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Using the Attribute Editor

- -

You can edit the attributes of a concept or node in the Knowledge Base - using the Attribute Editor.  The - Attribute Editor is particularly handy when creating or working with a - lexicon as you can edit lexical items directly.

- -

To access the Attribute Editor:

- - - - - - - - -
-

Note: - You can also access the Attribute Editor from the following:

-
    - -

  • In - the KB Editor, Gram Tab, Ana Tab, and Text Tab, right-click on a concept - and select Attributes.

    • - -

  • In - an open text file, highlight a word and select Dictionary - > KB Attributes from the menu.  If - the word exists in the dict hierarchy, the Attribute Editor will be launched.

    • -
- -

See Attribute - Editor Description for details on the Attribute Editor buttons.

- -

 

- - - - diff --git a/Help/helps/Using_the_KB_Editor.htm b/Help/helps/Using_the_KB_Editor.htm deleted file mode 100644 index 5f2c028..0000000 --- a/Help/helps/Using_the_KB_Editor.htm +++ /dev/null @@ -1,49 +0,0 @@ - - - - - - - - - - -Using the KB Editor - - - - - - - - - - - - - - - - - - - - - - - - - - - -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/VB_to_VT.htm b/Help/helps/VB_to_VT.htm deleted file mode 100644 index 8a75bb1..0000000 --- a/Help/helps/VB_to_VT.htm +++ /dev/null @@ -1,172 +0,0 @@ - - - - - - - - - - -VB_to_VT - - - - - - - - - - - - - - - - - - - - - - - - - - - -

VB_to_VT

- -

A Visual Basic .NET driver program that demonstrates how to call a VisualText - analyzer.  It - uses the new VTAPI_VB.dll library built to provide a simple programmer's - API to VisualText.  The - main code file is repeated here for convenience.  All - the API functions are invoked in this sample.

- -

'''''''''''''''''''''''''''''''''''''
-'VB_TO_VT = Driver to call a VisualText analyzer from a Visual Basic program.
-'ASSUMES:
-'  Global - environment variable APPS = d:\apps
-'  Minanalyzer - available in d:\apps.  (Copy - from c:\Program Files\TextAI\VisualText\apps)
-'''''''''''''''''''''''''''''''''''''

- -

Module Module1

- -

  Private - Declare Function makeVTRUN_VB Lib - "VTAPI_VB.dll" () As Integer
-  Private - Declare Function makeNLP_VB Lib "VTAPI_VB.dll" - (ByVal ana As String, ByVal vtrun As Integer) As Integer
-  Private - Declare Function cleanup_VB Lib "VTAPI_VB.dll" - (ByVal nlp As Integer, ByVal vtrun As Integer) As Integer
-  Private - Declare Function setbatchstart_VB - Lib "VTAPI_VB.dll" (ByVal flag As Boolean, ByVal nlp As Integer) - As Integer
-  Private - Declare Function analyzeFile_VB Lib - "VTAPI_VB.dll" (ByVal inf As String, ByVal ob As String, ByVal - osiz As Integer, ByVal datum As String, ByVal compiled As Boolean, ByVal - nlp As Integer) As Integer
-  Private - Declare Function analyzeBuffer_VB - Lib "VTAPI_VB.dll" (ByVal ib As String, ByVal ob As String, - ByVal osiz As Integer, ByVal datum As String, ByVal compiled As Boolean, - ByVal nlp As Integer) As Integer

- -

   Sub - Main()

- -

      'NAME - OF ANALYZER
-     Dim - analyzerName As String = "minanalyzer"

- -

      'RUNTIME - ENVIRONMENT HANDLES.
-     Dim - vtrun As Integer = 0
-     Dim - nlp As Integer = 0

- -

      'SET - UP PARAMETERS FOR ANALYZER.
-     Dim - infile As String = "D:\apps\minanalyzer\input\misc\xyz.txt"
-     Dim - inbuf As String = "This is an input buffer"
-     Dim - outsiz As Integer = 512
-     Dim - outbuf As String = Space(outsiz)
-     Dim - datum As String = "Parameters to pass to analyzer"

- -

      'Todo: - After you compile the analyzer in VisualText, you can turn on this flag.
-     Dim - compiled As Boolean = False

- -

      'INITIALIZE - VISUALTEXT RUNTIME SYSTEM (Onetime call)
-     '(Handles - multiple analyzers)
-     vtrun - = makeVTRUN_VB()

- -

      'INITIALIZE - CURRENT ANALYZER RUNTIME (Onetime call)
-     nlp - = makeNLP_VB(analyzerName, vtrun)

- -

      '''''''''''''''''''''''''''''''''''''''''''''''''
-     'Analyzer - calls below can be placed into a loop.
-     '''''''''''''''''''''''''''''''''''''''''''''''''

- -

      'Inform analyzer - of first file in a batch.
-     'Note: - Analyzer automatically resets flag to false.
-     setbatchstart_VB(True, - nlp)

- -

      'RUN ANALYZER ON - FILE.
-     analyzeFile_VB(infile, - outbuf, outsiz, datum, compiled, nlp)
-     Console.WriteLine("outbuf: - " & Trim(outbuf))

- -

      'RUN - ANALYZER ON BUFFER.
-     analyzeBuffer_VB(inbuf, - outbuf, outsiz, datum, compiled, nlp)
-     Console.WriteLine("outbuf: - " & Trim(outbuf))

- -

      '''''''''''''''''''''''''''''''''''''''''''''''''
-     'CLEAN/DELETE - THE RUNTIME ENVIRONMENT (Onetime call)
-     cleanup_VB(nlp, - vtrun)

- -

   End - Sub
-End Module

- -

 

- - - - diff --git a/Help/helps/VisualText_Basics/About_Automated_Rule_Generation.htm b/Help/helps/VisualText_Basics/About_Automated_Rule_Generation.htm deleted file mode 100644 index 2c075d3..0000000 --- a/Help/helps/VisualText_Basics/About_Automated_Rule_Generation.htm +++ /dev/null @@ -1,208 +0,0 @@ - - - - - - - - - - -About Automated Rule Generation - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

About Automated Rule Generation

- -

VisualText allows you to enhance text analyzers by merely highlighting - and categorizing samples from text. Automated rule generation (or RUG, for short) creates passes and rules - for you automatically, so that you don't have to worry about the details - of rule writing. It also makes analyzers much more maintainable, because - when you modify an analyzer and then invoke RUG, it automatically - updates rules according to changes made earlier on in the analyzer. That - is, the developer doesn't need to worry that changes to rules in one pass - will invalidate dependent rules in subsequent passes.

- -

RUG is best used for word and phrase-level constructs that are modular - and well defined, e.g., telephone numbers. More generally, RUG can be - used for the things that an analyzer deals with explicitly. For example, - a linguistically sophisticated analyzer that identifies the components - of sentences (noun phrases, verb phrases, etc.) can exploit RUG to identify - clausal and sentential patterns.

- -

RUG is less useful for completely open ended or nonproductive constructs. - For example, RUG might not be very useful for storing and identifying - many instances of full names, such as "John Smith."

- -

As you develop your text analyzer, you will learn how best to organize - the samples that you want to collect within the Gram Tab.

- -

Controlling Automated Rule Generation

- -

A caveat in using RUG is that sometimes rules will not be as general - as you want, or they may be more general than you want. The Gram - Concept Properties dialog can be used to help control some of these - behaviors. Updates to this Help documentation will help detail some settings, - but experimentation may also be required.

- -

Automated Rule Generation and the Gram Tab

- -

The Gram Tab - is where you set up concepts and samples for the automated rule generation - process.  

- -

Basic Process of Automated Rule Generation

- -

VisualText automatically generates rules by taking samples you provide, - making generalizations from the samples, and then creating the rules. - An outline of the process of automatic rule generation is given below. -  Each - step is detailed in the Automatically Generating Pass Files section.

- -
    - -
  1. -

     Create a stub, (e.g., - called phrases).

    - -

    Select a pass in the Ana Tab at the location where you want a set - of automatically generated passes to be placed. Then, either add a stub concept in the Gram Tab or a stub region in the Ana Tab, since adding - one always adds the other automatically. The stub concept is the primary - organizer of user-highlighted samples. The stub region is reserved for - the automatically generated passes and rules. When RUG is invoked, it - automatically inserts or overwrites passes within the stub regions.

    - -

    We recommend that the ordering of stub concepts (in the Gram Tab) - and their corresponding stub regions (in the Ana Tab) be kept in tandem, - for clarity.

    -
    2. - -
  2. -

    Create a folder concept, another level of organizing user samples - (e.g., contact information).

    - -

    Folder concepts organize multiple types of samples.

    -
    3. - -
  3. -

    Create a rule concept, which - represents a set of closely related samples (e.g., telephone number).

    - -

    Rule concepts "own" a set of related samples.

    -
    4. - -
  4. -

    Create a label concept, - which represents a set of subsamples (e.g., area code).

    - -

    Label concepts "own" a set of related subsamples. Label - concepts and subsamples are optional. For example, if you don't care about - subdividing a telephone number, then you need not create a label concept - for the area code.

    -
    5. - -
  5. -

    Select a concept in the - Gram Tab (under which you want to add text samples).

    - -

    This will dynamically create a menu of rule and label concepts, - with which the user may categorize samples.

    -
    6. - -
  6. -

    Highlight a sample within - a text file in the workspace.

    - -

    Use the Text Tab to find and open a text file, then highlight a - text sample within it.

    -
    7. - -
  7. -

    Right-click and select a rule or label concept.

    - -

    For example, if a full telephone number sample is highlighted, select - the rule concept for telephone numbers. If you highlight a part of the - full sample, such as an area code, then select the label concept for area - codes.

    -
    8. - -
  8. -

    Invoke RUG.

    - -

    When done adding samples and subsamples, from the Main Menu select - Analyzer > Generate Rules > All. RUG will create new passes, as - appropriate, in the stub regions of the Ana Tab. (You could also use the - Generate button, but we recommend that Generate Rules > All be used - as far as possible.)

    - -

     

    -
    9. -
- -

See Also

- -

Adding - Stub Regions

- -

Adding - Rule Concepts

- -

Adding - Folder Concepts

- -

Adding - Samples to Rule Concepts

- -

Adding - Label Concepts

- -

Setting - Concept Properties

- -

Running - the Rule Generator

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Basics/About_KB_Command_Files.htm b/Help/helps/VisualText_Basics/About_KB_Command_Files.htm deleted file mode 100644 index c5a7fed..0000000 --- a/Help/helps/VisualText_Basics/About_KB_Command_Files.htm +++ /dev/null @@ -1,480 +0,0 @@ - - - - - - - - - - -About KB Command Files - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

About KB Command Files

- -

On examining the interpreted knowledge base in the kb\user folder of - an analyzer project, you will find files with human-readable commands:

- - - --- - - - - - - - - - - - - - - - - - - - -
-

main.kb

 -

Invokes command files for loading the knowledge base.

-

hier.kb

 -

Adds the concepts in the knowledge base hierarchy

-

word.kb

 -

Adds the words in the dictionary

-

phr.kb

 -

Adds phrases to concepts

-

attr#.kb

 -

Adds attributes to concepts and phrase nodes

- -

 

- -

When an interpreted KB is read as part of loading an analyzer, the commands - in these four files are executed in the order given above.  (This - ordering guarantees that concepts are added to the kb before they are - used in phrases, and that all concepts and nodes are present before attaching - attributes to them.)  Similarly, - the gui.kb file at the top-level directory of an analyzer project defines - the analyzer sequence and the input file hierarchy, as loaded into the - knowledge base.

- -

All these .kb files execute a batch command language for adding and - editing the knowledge base.  A - typical command looks like

- -

add hier "concept" "sys"

- -

add hier <path> specifies - that a concept is to be added to the hierarchy, as specified by a path - of concept names that starts at the root of the knowledge base hierarchy, - called "concept."  Thus, - the command above says to start at the root, named "concept," - and to add a concept called "sys" under the root, if not already - present.  Another - example is

- -

add word "xyz"

- -

The add word command above adds - the word "xyz" under its appropriate index-concept in the special - dictionary hierarchy within the KB.

- -

The command ind phrase is used - to add a phrase of nodes to a concept.  Each - concept can only have one phrase, with any number of nodes:

- -

ind - phrase

- -

<concept>         ; - The concept to which the phrase is being added.

- -

<concept 1>       ; - Concept referenced by 1st node of phrase.

- -

<concept - 2>       ; - Concept referenced by 2nd node of phrase.

- -

...

- -

end ind

- -

An example that adds an attribute to a concept is

- -

ind - attr

- -

"concept" - "gram"

- -

0

- -

"concept" - "sys" "dict" "a" "c" "co" - "constrain"

- -

pst

- -

"true"

- -

end - ind

- -

The command ind attr is a multi-line - command for adding an attribute.  Its - general form is

- -

ind attr

- -

<concept>         ; - The concept to which an attribute is being added.

- -

<concept index> -   ; - 0 means attr is to be added to the concept itself.  1 - = add attr to 1st

- -

                  ; - node of concept's phrase, etc.

- -

<key concept> -     ; - The attribute key (or name), specified by a concept in the kb.

- -

<value type> -      ; - The type of value:

- -

                  ; - (pst = string, pnum = integer, pfloat = float pcon = concept)

- -

<value>           ; - The value itself, according to its type.

- -

...

- -

<value type N> -    ; - Can optionally add second value type, second value, and so on.

- -

<value N>

- -

end ind

- -

 

- -

Each command file should - end with the quit command alone - on a line.  Comments - are indicated by a semicolon, as above.

- -

 

- -

Note that command files - can be executed from an analyzer pass file, using the NLP++ take - function (also available as an API function in include/API/consh/cg.h). -  take(filename) - executes the commands in the specified .kb file.  This - is one way in which knowledge can be layered or imported into a knowledge - base conveniently.  In - addition, there's a take command, - as in

- -

 

- -

take "kb\\user\\main.kb"

- -

 

- -

which accepts both an absolute - and relative filename as argument, and executes the commands within the - given filename.  Relative - filenames assume the current application project as the root directory.

- -

 

- -

A variant to add attributes - to a dictionary word concept has been implemented, called ind - wattr.  Its - format is identical with ind attr, except that instead of a concept, a - word may be specified for convenience.

- -

 

- -

ind - wattr

- -

<word>            ; - A string.  The - dictionary word to which an attribute is being added.

- -

0                 ; - This line is kept merely to match the format of ind attr.

- -

<key concept> -     ; - The attribute key (or name), specified by a concept in the kb.

- -

<value type> -      ; - The type of value:

- -

                  ; - (pst = string, pnum = integer, pfloat = float pcon = concept)

- -

<value>           ; - The value itself, according to its type.

- -

...

- -

<value type N> -    ; - Can optionally add second value type, second value, and so on.

- -

<value N>

- -

end ind

- -

 

- -

 

- -

Additional batch commands - exist, but the most important ones have been mentioned here.

- -

 

- - - - diff --git a/Help/helps/VisualText_Basics/About_Parse_Trees.htm b/Help/helps/VisualText_Basics/About_Parse_Trees.htm deleted file mode 100644 index 61e499f..0000000 --- a/Help/helps/VisualText_Basics/About_Parse_Trees.htm +++ /dev/null @@ -1,189 +0,0 @@ - - - - - - - - - - -About Parse Trees - - - - - - - - - - - - - - - - - - - - - - - - - - - -

About Parse Trees

- -

VisualText analyzers build and use a single parse - tree (or tree, for short) - which is a data structure that tracks the patterns that have matched within - the input text. Building parse trees is similar to the sentence diagramming - taught in elementary school. The first pass of every analyzer tokenizes - the text, that is, it converts a stream of characters to a parse tree - in which alphabetic, numeric, whitespace, and punctuation are grouped - into units called nodes. Below - is a tokenized parse tree as displayed in the Parse Tree View.

- -

 

- -

- -

 

- -

Parse Tree Terminology

- -

Each line in the display above corresponds to one node of the parse - tree. The root, or top-level node - of the parse tree, is named _ROOT and appears at the top of the display. - The remaining nodes are children - of the root, and the root is their parent. - Generations, or levels, of nodes - in the parse tree are indicated by indentation. The phrase - or list of children are all in - a sibling relationship to each - other. In general, a parse tree has more levels than shown above. We speak - of ancestors and descendants to describe the relationships between nodes - at different levels of the parse tree.

- -

A token, literal, - leaf, or terminal - node, represents a literal text. Tokens are drawn with red dashes, as - above. A nonterminal (or abstract) node is one with descendants. - We say that a nonterminal node dominates - its descendant nodes. Nonterminal nodes are drawn with a grey pentagon. - The root node is the only nonterminal in the above tree. When selected, - a node's color changes to blue.

- -

Terminal Nodes

- -

A terminal node is a token and represents words, numbers, whitespace, - and punctuation. The format for terminal nodes is as follows:
-
-nodename::(node_type info1 info2.. ("var1" "val")...) - text-associated-with-node

- -

The node type identifies the type of token. It is one of alpha - (for alphabetic), num (for number), - white (for whitespace), or punct (for punctuation). Information - about the node follows the node type. Flag settings such as unsealed - (or un), built, - and fired may be shown in abbreviated - form. Variables and values attached to the node may be displayed on the - node line as well. Each variable and value-list is a parenthesized list - of the form ("varname" "value1" "value2" - ... ).

- -

Nonterminal Nodes

- -

A nonterminal node is a node with descendants. An underscore is normally - the first character in the name of a nonterminal. The format for nonterminal - nodes in the parse tree is as follows:

- -

_nodename::(node info1 info2) text-associated-with-node

- -

For example, a non-terminal node _company - created when the rule matcher encountered the company name "Text - Analysis International" in a text might look like:

- -

_company::(node un) Text Analysis International

- -

The information following the node type is as described for terminal - nodes above.

- -

Distinguishing Terminal and Nonterminal Nodes

- -

You can distinguish between terminal and nonterminal nodes by the symbols - used in the parse tree display.

- -
    - -

    denotes a nonterminal node, and

    - -

    denotes a terminal node.

    -
- -

When a node is selected in a parse tree view, the icon for the selected - node turns blue. A nonterminal is blue when selected and grey otherwise. - A terminal is blue when selected and red otherwise. You can collapse and - expand nodes by clicking on the + and - symbols.

- -

Final Parse Tree

- -

The final parse tree is the - tree produced by fully analyzing a text with a text analyzer.

- -

Intermediate Parse Tree

- -

VisualText enables you to view the parse tree - after each pass has executed. Each such intermediate - parse tree records the patterns found in the input text by the - analyzer passes up to and including the current pass.

- -

Concept Tree

- -

The concept tree is a parse - tree that contains only nodes that have been built for the selected pass - in the analyzer sequence.  The - concept tree view lets you quickly check which nodes were created in order - to step through your analyzer.  

- - - - diff --git a/Help/helps/VisualText_Basics/About_the_Knowledge_Base.htm b/Help/helps/VisualText_Basics/About_the_Knowledge_Base.htm deleted file mode 100644 index 4f409e9..0000000 --- a/Help/helps/VisualText_Basics/About_the_Knowledge_Base.htm +++ /dev/null @@ -1,168 +0,0 @@ - - - - - - - - - - -About the Knowledge Base - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

About the Knowledge Base

- -

The Knowledge Base Management System - (KBMS) used in VisualText is - based on a knowledge representation framework developed by Text Analysis - International, called Conceptual Grammar (CG). - CG represents knowledge in terms of concepts, phrases, and their attributes. - Everything "hangs" off a hierarchy of concepts. Because CG can - represent concepts, phrases, attributes, and their interrelationships, - it allows diverse types of knowledge to be stored flexibly within a single - framework. See also the Introduction - to the Knowledge Base.

- -

The KB - Editor and the Attribute - Editor can be used to manipulate portions of the KB.

- -

KB Location

- -

The KB is stored under an application's - kb folder. For example, for the tutorials, the folder

- -

C:\apps\myAnalyzerFolder\myAnalyzer\kb

- -

has a folder called user. - The user - folder stores data specific to the analyzer in question. The KB for an - analyzer is defined by the four files in the kb\user folder: attr.kb, - hier.kb, - phr.kb - and word.kb.

- -

KB Objects

- -

The KB consists of a hierarchy of concepts. - Attributes can be associated with - each concept . An attribute consists of a key - (or slot, or name) - and zero or more values. A value - can be a string, an integer, or another concept. Each attribute can have - multiple values, each with a distinct type.  (Note - that a value, or val, - is a type of knowledge base object.)

- -

A concept can also have an associated phrase - of nodes. A node is similar to - a concept in most respects, except that a node is not placed directly - in the knowledge base hierarchy. Rather, nodes serve as "proxies" - or references to concepts that reside in the hierarchy. Phrases can be - used to implement idioms, patterns, samples, rules -- in short, any sequential - information.

- -

Advanced Notes: An - attribute key itself is actually a concept, rather than a string, so that - multiple attributes with the same "name" can coexist peacefully - within the same concept. To add further flexibility (or complexity, depending - on your view), multiple attributes with identical keys can reside within - a single concept. For many applications, referring to attribute keys by - name is just fine, so most of the implemented NLP++ functions that you'll - encounter refer to attributes by name.

- -

Knowledge Base Functions

- -

See the Table of - KB Functions and page-per-function in the Knowledge Base Functions - section.

- -

Saving the Knowledge Base

- -

The knowledge base can be changed manually by using the KB Editor.  Analyzers - can also be programmed to update the knowledge base automatically, for - example to accumulate information across a set of documents.

- -

Some analyzers are geared to load up a relatively empty knowledge base - and then populate it temporarily.  Some - analyzers are built primarily to grow a large knowledge base.  When - exiting VisualText, the user is asked to save the changes to the knowledge - base or discard them.  Unless - your purpose is specifically to build or modify a knowledge base, then - you can opt not to save the modified knowledge base.  Some - distinctions to be aware of:

- -

Save KB - Saves the knowledge base into command files.
-Compile KB - Saves the knowledge base into a file system.
-Load interpreted KB - read kb command files into the knowledge base on - loading an analyzer.
-Load compiled KB - load or use the compiled kb file or file system.

- -

QDBM-Based Knowledge Base

- -

With the 2.1.0.0 version and later, VisualTextQ - provides a knowledge base manager based on the QDBM freeware.  This - supports growing very large knowledge bases in VisualText, by virtue of - replacing the old kb.dll file with the QDBM file system.  With - this variant, loading the interpreted version reads the knowledge base - command files into the QDBM database.  Saving - the knowledge base writes out command files, as before.  Compiling - the knowledge base saves off a copy of the QDBM database.

- -

Loading the compiled knowledge base copies - the QDBM database to a temporary folder (kb\tmp) for use in the current - session.  If - the session is closed without saving the kb, then the changes made in - the tmp area are discarded.  If - the kb is "compiled", then the tmp area is merely copied to - the permanent store (kb\qdbm).

- -

In this way, the same knowledge base management - commands work as before, but under the hood the QDBM system replaces the - kb.dll file for the compiled version of the knowledge base.  As - always, BEFORE UPGRADING to the new VisualText version, it is important - to save any knowledge base that is important -- this will create command - files that will automatically load in any newer version of VisualText. -  For questions - about kb management, please contact our support.

- - - - diff --git a/Help/helps/VisualText_Basics/About_the_analyzer_sequence.htm b/Help/helps/VisualText_Basics/About_the_analyzer_sequence.htm deleted file mode 100644 index 848e19a..0000000 --- a/Help/helps/VisualText_Basics/About_the_analyzer_sequence.htm +++ /dev/null @@ -1,241 +0,0 @@ - - - - - - - - - - -About the analyzer sequence - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

About the Analyzer Sequence

- -

Passes

- -

The analyzer sequence is a series of steps or passes, - each containing its own pass algorithm. As the analyzer is run over the - text, each pass is taken in the order it occurs in the sequence and executes - the code and rules that are contained in it. The Ana - Tab manages the analyzer sequence.

- -

Analyzer passes use and modify a common data structure, called a parse tree.  They - may also post and use shared information in the knowledge base.  VisualText - analyzers can be very fast because they build a single - parse tree per text.

- -

System and User Passes

- -

There are two types of passes in the analyzer sequence: system - or built-in passes and user passes. - The first pass in the analyzer sequence is tokenize - by default, which is a system pass that converts the input text to a parse - tree. The tokenize pass, like all system passes, cannot be modified.

- -

User passes employ either the pattern - (or pat) and the recursive - (or rec) algorithm.  Both - employ a pass file, in which the - user places NLP++ code and rules.

- -

NOTE: - VisualText does not provide a way to add system passes to the analyzer - sequence.  You - can do this manually by exiting VisualText and editing the analyzer.seq - file in the project's spec folder. -  VisualText - 2.5.x.x and onward provides a new tokenize pass called dicttok - that uses a knowledge base to automatically put part-of-speech information - on the parse tree. (Grab the three kb.* files from bin - folder of TAIParse, available at our website, and put them in a new analyzer - project's bin folder).

- -

If you need a customized tokenization pass to - handle OCR output (already exists, commercially available in conjunction - with http://www.XIEO.info)), convert HTML to plain text, or another specialized - input format, TAI can help.  Contact - us at support@textanalysis.com, 1-877-235-6259 USA toll free.

- -

dicttok also puts a CTRL - attribute on control characters to help manage control characters (e.g., - using feature-based matching), as in the code snippet below.

- -

@PRE
-<2,2> vareq("CTRL",128); -  # - Match ASCII 128 =
-@POST
-"output.txt" << "Found - ctrl = " << N("$text",2) << "\n";
-@RULES
-_xNIL <-
-
-_xCTRL
-_xALPHA
-@@

- -

will match the character () in the following text and output it.

- -

       “small

- -

 

- -

Ordering of Passes

- -

Since each pass in the analyzer typically builds upon the last one (via - a common parse tree), careful attention should be given to the ordering - of each pass in the sequence. When a pass creates a structure that is - used by later passes, the pass is said to be feeding them. If it removes or hides - information from successive passes, it is said to be bleeding. - One nice feature of VisualText is that you can create an analyzer sequence, - see how it works and easily change the sequence around to suit your needs. - Since the ordering of rules within - a pass file can also create, hide or destroy parse tree structure available - for later rules and passes, careful attention should also be given to - the ordering of rules within the pass file.

- -

Pass Types

- -

Pattern - The pattern algorithm - is the workhorse pattern-matching algorithm type used in VisualText. For - each node in the currently selected phrase, each rule within the pass - file is tried in turn. When a rule has matched, the pattern algorithm - executes its associated code and actions, then moves onto the node following - those that matched the current rule. (Notes: [1] when rules have a trigger, - they are only tried if the triggered rule element will successfully match. - [2] The next node to be tried after a rule match may be selected using - the lookahead element action.)

- -

Recursive - The recursive algorithm - type implements a standard recursive-grammar algorithm. The nodes of the - current phrase are in effect traversed repeatedly, as long as a rule in - the current pass file matches. When no rules match in a phrase, then the - next phrase in the parse tree is selected, till no more phrases remain - to select. (Notes: [1] noop rule - matches are counted as failures - by the Recursive algorithm, so that they won't cause an infinite loop.)

- -

Adding Passes to the Analyzer Sequence

- -

There are several ways to add passes to the analyzer sequence. One way - is to create or add the pass yourself and the other way is to have VisualText - create the pass for you. An analyzer will generally contain both types - of passes.

- -

Hand Building Passes

- -

To create the pass yourself, you write NLP++ code and rules in the pass - file used by the pass. It is also possible to add an existing pass file - from another analyzer directly into your project. When you add your own - passes, you use the Ana Tab. For more information on adding passes to - the analyzer sequence yourself, see Adding - Pass Files.

- -

Automatically Generating Passes

- -

When VisualText automatically creates a pass, it takes samples that - you provide, makes generalizations from those samples and then creates - the passes and the associated pass files. The automatically created passes - are placed in a stub region.  The - process of adding samples is called sampling. - Automatically generated passes can be edited, but any edits will be lost - the next time the automatic rule generation (RUG) is invoked. To make - changes, you will have to go through the sampling process again. The Gram - Tab is used for sampling. For more information on the automatic rule generation, - see About Automatic Rule - Generation.

- -

Inactivating a Pass in the Sequence

- -

As you add new passes to the Ana Tab, they become part of the stepwise - sequence executed when the text analyzer runs on an input text. It is - possible to have the analyzer skip - a pass in the analyzer sequence by setting the pass properties for the - pass. When you inactivate a pass, you tell the analyzer to ignore the - code and rules included in the pass file associated with the pass. Skipping - passes can be used as a tool to debug your analyzer. For more information, - see Inactivating - Passes.

- -

Icons Used in the Analyzer Sequence

- - - - - - diff --git a/Help/helps/VisualText_Basics/Compiled_and_Standalone_Analyzers.htm b/Help/helps/VisualText_Basics/Compiled_and_Standalone_Analyzers.htm deleted file mode 100644 index 5051064..0000000 --- a/Help/helps/VisualText_Basics/Compiled_and_Standalone_Analyzers.htm +++ /dev/null @@ -1,882 +0,0 @@ - - - - - - - - - - -Compiled and Stanalone Analyzers - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Compiled and Standalone Analyzers

- -

This section is concerned with configuring and using text analyzers - outside the VisualText environment. It covers:

- - - -

We assume that an analyzer has been built using VisualText, and that - Microsoft Visual C++ is available to the developer.

- -

Linux and Gnu C++. -  Analyzers - built with VisualText can be configured to run on Linux with Gnu C++. -  TAI periodically - upgrades the port to the VisualText runtime libraries.  The - libraries are not currently bundled with the VisualText release, and you - may inquire about them by contacting us at info@textanalysis.com .

- -

Overview

- -

The interpreted text analyzer - constructed with VisualText consists partly of the definition files within - the application's spec folder. - The spec folder contains the analyzer sequence file (analyzer.seq) - that defines the order of text-analysis passes, as well as the pass files - (.pat files) that specify the - rules and code to be executed at each pass. The analyzer engine and knowledge - base management system (KBMS) are implemented as DLLs that execute the - interpreted knowledge within the spec folder and within the kb - folder, respectively. In addition, the app's user - project is available for adding new NLP++ functionality to the core functions - provided with VisualText.

- -

When the user compiles the analyzer, C++ code is generated, placed in - the app's run folder, and compiled - using MS Visual C++. When the user compiles the knowledge base, the app's - kb folder is similarly updated.

- -

C++ code files are also provided for a default standalone analyzer. - The user creates an empty project and adds the C++ files at the application's - top level folder (main.cpp and - so on). The default standalone analyzer exemplifies the way in which a - VisualText application can be embedded in larger programs.

- -

The following sections provide further details about these topics.

- -

Standalone Analyzer

- -

Analyzer templates (such as Bare) come with prebuilt Microsoft Visual - C++ project files prog.dsw and - prog.dsp, which contain most of - the project settings described below.  You - should be able to just modify the line in main.cpp and compile this project - to have a standalone version of your analyzer.

- -

Instructions for creating and running the standalone analyzer:

- -

1. Make sure that your app folder (e.g., C:\apps\chatter) has the latest - versions of the code files, by comparing with the files in the directory - VisualText\Templates\Bare. The files include main.cpp, Uanalyze.cpp, Uanalyze.h, - and possibly others.

- -

2. Use MS Visual C++ to create a standalone application in the app folder - (e.g., an empty Win32 Console Application). Use the name of the app folder - to name the project and create the project within the app folder.

- -

3. Add the code files in the app folder to the project.

- -

4. Edit the Project - Settings:
-
-In Project Settings > C/C++ tab > - Category: Preprocessor > Additional include directories: add - the filler $(VISUALTEXT)\include\Api

- -

In Project - Settings > C/C++ tab > Category: Code Generation > Use run-time - library: select

- -

    (in - Release config) Multithreaded

- -

    (in - Debug config) Debug - Multithreaded

- -

In Project - Settings > C/C++ tab > Category: General > Debug info: - select

- -

    (in - Release config) None

- -

    (in - Debug config) Program - Database

- -

 

- -

In Project - Settings > Link tab > Category: Input > Additional library path: - add the filler $(VISUALTEXT)\lib

- -

In Project - Settings > Link tab > Object/library modules: append the - text
-  (in release config:) libconsh.lib - lite.lib libprim.lib

- -

  (if - available: in debug config:) libconshd.lib - lited.lib libprimd.lib

- -

 

- -

In Project - Settings > Debug tab: you may want to set up an input text file - to process for testing purposes.

- -

5. In main.cpp, edit the name of the application. e.g., appname - = "chatter";  This - name must conform to the name of the top-level folder of the analyzer - project.

- -

6. Build the program in the standard - fashion within the MS Visual C++ environment.

- -

Command Line Arguments

- -

Invoking a standalone analyzer (built as above) looks like

- -

analyzer-executable [/INTERP][/COMPILED][/IN - infile] [/OUT outfile] - [/DEV] [infile [outfile]]

- -

The various arguments are

- - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

ARGUMENT

 -

DESCRIPTION

-

/INTERP

 -

Run the interpreted analyzer (this is the default).

-

/COMPILED

 -

Run the compiled analyzer (run.dll).

-

/IN infile

 -

Specify the input file to be analyzed

-

/OUT outfile

 -

Specify the output file [SEE NOTES]

-

/DEV

 -

Developer's verbose mode (e.g., output intermediate logs).

-

[infile [outfile]]

 -

Unmarked args are input file followed by output file.

-

 

 -

 

- -

Analyzer and Knowledge Base APIs

- -

The code files in the top level folder of an analyzer project demonstrate - API calls to load and run the analyzer and knowledge base.  The - API header files are located by default in

- -

C:\Program Files\TextAI\VisualText\include\API

- -

and its subfolders.  The - main API header files are

- -

API\lite\nlp.h, which specifies - function calls for creating an NLP object to manage an analyzer and run - it on input text files or buffers, and

- -

API\consh\cg.h, which specifies - function calls for creating a CG object to manage, access, and modify - the knowledge base.

- -

The header documentation in nlp.h is repeated here for reference.

- -

THE MAIN API FUNCTIONS - ARE:

- -

 

- -

NLP              = - Create the NLP object to manage analyzer.

- -

make_analyzer    = - Create and load a text analyzer.

- -

 

- -

API CALLS FOR NORMAL - RUNNING OF ANALYZER ON A TEXT:

- -

 

- -

analyze          = - Run analyzer on a single text.

- -

 

- -

API CALLS FOR STEPPING - THROUGH ONE PASS AT A TIME:

- -

 

- -

apiInitAnalyzer  = - Initialize stepwise interp analysis of a text.

- -

apiStepAnalyzer  = - Execute next pass in the analysis of a text.

- -

apiCleanAnalyzer = - Clean up stepwise analysis of a text.

- -

 

- -

 

- -

We describe the call to analyze(), which is the main function that you'll - call outside VisualText to analyze a text.  Note - that an example of the use of the API calls NLP, - make_analyzer, and analyze - is found in the main.cpp file of any analyzer project.

- -

void - analyze(

- -

   char - *input,         // - INPUT FILE PATH.

- -

   char - *output,        // - OUTPUT FILE PATH.

- -

   char - *appdir,        // - ANALYZER APPLICATION TOP DIR.

- -

   bool - flogfiles,      // - OUTPUT INTERMEDIATE LOGS.

- -

   char - *outdir = 0,    // - DIRECTORY FOR OUTPUT.

- -

   char - *inbuf = 0,     // - IF ANALYZING A BUFFER, SUPPLY IT HERE

- -

   long - len = 0,        // - LENGTH OF INPUT BUFFER

- -

   bool - compiled=false, // RUN COMPILED ANALYZER

- -

   ostream - *os = 0,     // - STANDARD OUTPUT STREAM

- -

   char - *outbuf = 0,    // - OUTPUT BUFFER TO BIND TO cbuf STREAM

- -

   long - outlen = 0,     // - MAXIMUM LENGTH OF cbuf OUTPUT BUFFER

- -

   char - *datum          // - STRING OF PARAMETERS TO ANALYZER [NEWLY DOCUMENTED 2.4.0.1]

- -

   );

- -

 

- -

You have a choice of - specifying input from a file or a buffer.  If - inbuf and len - are zero, then analyze() assumes that input comes from the input - file argument.

- -

 

- -

Output is more flexible - in that during the analysis of a text, you can send output to a file, - a user-supplied stream, and a user-supplied buffer.  output - specifies the default output file, outbuf - and its length outlen specify - an output buffer, and os specifies - an output stream.  In - NLP++, you may also send outputs to additional files (see discussion of - output statements).

- -

 

- -

The compiled - flag, if true, specifies that you are running a compiled version of the - analyzer (run.dll, see below).  If - false, then you are running the same interpreted analyzer definition that - you have developed within VisualText.

- -

 

- -

OUTPUT - FILE. The output argument leads nowhere in current-day NLP++. (That - is, there's no built-in NLP++ function such as cbuf() that writes to it.) -  But you - can manually add it to the DATUM argument to pass it to the analyzer.

- -

 

- -

DATUM - STRING. This arg has been available for many 2.x VisualText versions, - but hasn't been documented prior to 2.4.0.1.  It - is a string that gets passed to the G("$datum") global variable - of the called analyzer.  Typically, - one can separate values with a pipe character ('|') in the datum string, - and use something like

- -

    L("arr") - = split(G("$datum"),"|");

- -

to read each parameter into an array element. -  One can - use the parameters to set up output files, assign tasks, and pass information - to the called analyzer.

- -

Compiling the Analyzer

- -

Instructions for compiling the analyzer and running compiled vs interpreted - mode in the standalone analyzer. (Currently VisualText doesn't provide - a way to run the compiled analyzer.) Can only compile the analyzer during - load of VisualText, at present. (Could also compile it from a C++ program, - if desired.)

- -

1. Invoke VisualText.

- -

2. Depress the toggle for compiling the analyzer:  Compile Analyzer

- -

3. Load the analyzer to be compiled. C++ code gets generated and compiled - (in the app\run folder) during analyzer load. NOTE: - In version 2, you will need to go to the run folder of your analyzer and - manually upgrade and compile the project.  For - example, in version 2.3 and later, files are generated for each pass of - an analyzer.

- -

Running - the compiled analyzer

- -

Examine the API call for creating an analyzer, in main.cpp of the current - application folder:

- -

nlp->make_analyzer(seqfile, appdir, develop, 0, compile-on-load, - compiled);

- -

The compiled flag will load - the compiled analyzer if true. (Note: the interpreted analyzer is ALWAYS - loaded as well.) You can also specify that code be generated when making - the analyzer, by setting the compile-on-load argument to true.

- -

To run the compiled analyzer, set compiled - to true in the call to analyze, as below.

- -

nlp->analyze(infile,outfile,appdir,true,0,line,0,compiled);

- -

 

- -

Compiling the KB

- -

Unlike the analyzer, which can only be compiled during load, the knowledge - base can be compiled at any time by clicking the compile kb button:  Compile KB

- -

NOTE: In version 2, you will need to go to the - kb folder of your analyzer and manually upgrade and compile the project, - after the above.  You - may also need to add all the .cpp and .h files in that folder to the kb - project in order for it to compile.

- -

The user can save an interpreted version of the KB with the Save KB - button at anytime as well: Save KB

- -

Compiling very large KBs or loading interpreted KBs can take a long - time.  A - hint for managing very large KBs is to load the compiled version, an option - that can be set in the File > Preferences menu item.  Then, - the NLP++ functions kbdumptree - and take - can be used to dump and load subhierarchies of the kb that are being worked - on.  (Also - available is: in the KB Editor window, select a concept whose tree is - to be dumped or reloaded.  Right-click - and select Modular > Dump to dump the subtree to a .kb command file, - or Modular > Load to browse to a .kb file to be loaded.)

- -

The knowledge base can also be compiled from a C++ program.  In - the main.cpp file of the standalone app, add the API call

- -

cg->genKb();

- -

This will generate and compile C++ code for the knowledge base, creating - a kb.dll file in kb\Debug or kb\Release, as appropriate.

- -

Extending NLP++ with the User Project

- -

The user C++ project created - with a new analyzer provides an example of a user-defined NLP++ function, - in the Ucode.cpp file. An interpreted as well as compiled variant of the - function is exemplified. Methods for processing the function's arguments, - as well as accessing the parse tree and knowledge base, are illustrated.

- -

One way to define a new NLP++ function is within the @DECL region, which - must be first in a pass file, if present.  Another - way is to create a C++ function in the User project, as exemplified with - the function called testfn in - the file user\Ucode.cpp.  The - assumption is that a user-defined function must participate in NLP++ expressions, - and so it must process arguments in the same way that other NLP++ functions - process arguments.  It - must also return values by a special mechanism involving a retVal - object, as exemplified in the testfn example.  From - a pass file, the function must be called with a user:: prefix.  E.g., -  user::testfn().

- -

Just after an analyzer is loaded, the function ucodeIni() - is called, and just before an analyzer is closed, ucodeFin() - is called.  These - functions are available in the User project and may be filled with user-defined - initializations and cleanups.

- -

UPDATING ANALYZERS

- -

To update your analyzers to Version 2.0, you - will mainly need to regenerate and recompile the code in your projects. -  First, - compare your code with that in the top level, kb folder, run folder, and - user folder of the Bare analyzer in c:\Program Files\TextAI\VisualText\Templates\Bare. -  You may - need to add the code lines

- -

#include - <windows.h>
-#include <tchar.h>
-#include <my_tchar.h>
-#include <streamclasses.h>

- -

to your StdAfx.h file, or

- -

#include "vtapi.h"

- -

as appropriate, in order to succesfully compile. -  Then, regenerate - run and kb code, if you're using the compile version of the analyzer and - kb.  Similarly, - recompile the user project for your analyzer, especially if you are using - functions there or adding new ones to extend NLP++.

- - - - diff --git a/Help/helps/VisualText_Basics/DebuggingOverview.htm b/Help/helps/VisualText_Basics/DebuggingOverview.htm deleted file mode 100644 index 2903aca..0000000 --- a/Help/helps/VisualText_Basics/DebuggingOverview.htm +++ /dev/null @@ -1,413 +0,0 @@ - - - - - - - - - - -Debugging Overview - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Working with Debugging Tools

- -

VisualText provides a number of debugging tools to make building text - analyzers easier. Screen shots from this section are from the VisualText - Sample "Corporate" analyzer.

- -

Debug Structure: - Trees

- -

Almost all debugging tools in VisualText are based on parse trees. VisualText - analyzers build parse trees, which organize information into meaningful - chunks. You can debug an analyzer pass-by-pass, by setting the Generate - Logs button and running the analyzer:

- -

  followed by

- -

This will generate analyzer log files that contain trees for each text - processed. Each text will have its own set of trees. The log files are - displayed in the Text Tab under (indented and below) the corresponding - text file. (A red checkmark in a file indicates the file has been analyzed.) - You can see each of the log files generated for a file by clicking on - the plus sign in front of the text file. Log files for each pass in the - analyzer sequence have the name 'ana###.log' where ### is the number of - the pass in the analyzer sequence:

- -

- -

Debug Context: - Ana Tab

- -

The Ana Tab provides the context for what tree is displayed during debugging. - There is a tree built for every pass and when a pass is chosen in the - Ana Tab, all subsequent trees will be built from the selected pass.

- -

Highlighting

- -

Highlighting is the quickest way to see which rules in a pass have fired. - It operates on the text in the text file you are analyzing. Highlighting - is activated by clicking the Toggle - Highlight Mode button on the toolbar. It must be turned on before - you run the analyzer.

- -

- -

Within the trees generated for each pass in the analyzer, VisualText - records which nodes were matched and which nodes were built in the pass. - Built nodes are highlighted in blue. Green highlighting indicates other - nodes that matched rules in the pass.

- -

Blue Highlighting

- -

If a pass has rules that match a text and a tree is constructed or modified, - the highlighting appears in blue. Below, the rule for "dollars" - builds a new concept "_money" in the tree and therefore highlights - the text in blue.

- -

- -

Green Highlighting

- -

For rules that do not construct or modify a tree, highlighting is green. - Below, the rule containing the concept "_company" matches and - constructs a concept in the Knowledge Base but does not change the parse - tree, therefore the text is highlighted in green:

- -

- -

Dump Files

- -

Dump files are a convenient - aid to debugging. They record information useful in debugging the analyzer. - Dump files can be displayed by using the "Wheelbarrow" icon - in the Debug Toolbar with a pulldown menu:

- -

- -

Dump files are created by printing output to any file with a "txt" - extension using the piping printout operator "<<". Below - is the NLP++ code for printing out to the above "lookup.txt" - dump file:

- -

- -

One convenient Right-Click menu function will get you started in printing - to a dump file. It printouts out a line to a "txt" file with - the same name as the pass file. In this case, we clicked on the "lookup" - pass in our corporate analyzer and the Right-Click menu adds the line - above the menu as shown below:

- -

- -

Page Mode

- -

Page Mode is used to see changes in parse trees between passes in the - Ana Tab, or between text files themselves.  Page - Mode is activated by selecting the Toggle Page Mode button on the Debug - Toolbar:

- -

- -

Page Mode: Between - Passes

- -

Page Mode provides a convenient way to shuffle among sets of views. -  For example, - selecting pass 16 below causes the parse tree for that pass to display - in the parse tree view.  Selecting - pass 17 causes that parse tree view to automatically - display the parse tree for pass 17.

- -

- -

- -

Page Mode: Moving - Among Texts

- -

A second use for Page Mode is to look at changes across text files. - With the Page Mode turned on, text files (including dump files), parse - tree windows, and browser windows will update their content with the current - text selected. The arrows next to the Page Mode will move the text selection - in the Text Tab either up or down:

- -

- -

Syntax Errors

- -

VisualText informs you of syntax errors via the Log Window. If a syntax - error is detected, you can often go directly to the line in the pass file - to locate the problem by double-clicking on the error in the Log Window. - Below, we forgot an opening parenthesis and by double-clicking on the - error, VisualText opens up the correct pass file and moves the cursor - to the line containing the error:

- -

- -

Searching - in Text

- -

VisualText's Find - in Files dialog is handy for finding words and concepts in an analyzer - project. One can search for a word or phrase in only the word files, text - files, or even in other analyzers in the user's analyzer area:

- -

- -

Search results are displayed in the Find - Window.  By - double-clicking on a search result line, you can go immediately to the - file and line containing an instance of the search string:

- -

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Basics/Library_Passes.htm b/Help/helps/VisualText_Basics/Library_Passes.htm deleted file mode 100644 index 7d8e053..0000000 --- a/Help/helps/VisualText_Basics/Library_Passes.htm +++ /dev/null @@ -1,401 +0,0 @@ - - - - - - - - - - -Library Passes - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Library Passes

- -

To help you get started building your text analyzer, VisualText comes - with several pre-built pass files.  These - pre-built pass files are called library - passes.  They - are available within Add menu - item of the Ana Tab popup menu.

- -

Library passes contain NLP++ rules and code to recognize common or generic - types of data. Generic data includes telephone numbers, email addresses, - different types of money formats, dates, etc.  VisualText - also provides library passes for importing existing databases into the - knowledge base.

- -

Types of Library Passes

- -

There are two types of library passes:  those - that are used by themselves and those that are part of a series of passes. -  

- -

Below is a summary of available library passes.  Where - library passes are part of a series of passes, the total number of passes - is indicated next to the library pass name.

- - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Library Pass Name

 -

Description

-

Database (10)

 -

Import a database into the KB hierarchy.  Passes - can import database with both text and numeric fields.    

-

Assumes first line of database file is a list - of field names.  To - add a large database, it is more efficient to split up the database file - into smaller files.  If - splitting large database into smaller files, the first line of each file - should include the list of field names.

-

Date Plus Numerics (9)

 -

Recognize simple and compound numeric words, - numeric sequences and lists, various character formatting such as punctuation - and contractions, telephone numbers, and dates.

-

Emails

 -

Recognize email addresses.

-

HTML (5)

 -

Recognize HTML formatting elements, text tagging, - special HTML characters, and tagging structures.

-

Lines

 -

 

-

Numerics (7)

 -

Recognize simple and compound numeric words, - numeric sequences and lists, various character formatting such as punctuation - and contractions, and telephone numbers.

-

Remove Blank Lines

 -

Remove blank lines.

-

Remove White Space

 -

Remove white spaces.

-

Trends

 -

Count the occurrence of words and phrases - in a text.

-

XML DTD (14)

 -

XML document type definition.  DTD - is used to define the XML document structure.  

-

DTDs can be contained in the XML document, - or can be referenced from an external document.

-

Xml (21)

 -

Process XML files.

- -

Using Library Passes

- -

Library passes can be added to any part of your analyzer sequence.  Note - however, that they may not be inserted before the system pass, tokenize. -  When a - library pass is composed of a series of passes, all of the passes in the - sequence are added.

- -

The insertion of library passes into an existing analyzer may not be - completely "seamless."  That - is, the developer may need to reorder passes and edit contexts within - pass files (e.g., the list within a @NODES - context selector) in order to get the analyzer working properly with the - added library passes.

- -

Location of Library Passes

- -

The set of library passes that accompanies VisualText is stored in the - spec folder of the installation directory.  If - you installed the VisualText software on your c drive for example, the - library passes are contained in the following directory:

- -

c:\Program Files\TextAI\VisualText\data\spec

- -

You can create your own library passes and add them to the spec directory. -  They will - automatically appear in the Library Passes submenu of the Ana Tab popup - menu.

- -

See Also

- -

Adding - passes to the analyzer sequence

- - - - diff --git a/Help/helps/VisualText_Basics/Standard_Rule_Format.htm b/Help/helps/VisualText_Basics/Standard_Rule_Format.htm deleted file mode 100644 index 9ce606c..0000000 --- a/Help/helps/VisualText_Basics/Standard_Rule_Format.htm +++ /dev/null @@ -1,190 +0,0 @@ - - - - - - - - - - -Standard Rule Format - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Standard Rule Format

- -

To facilitate the editing of rules in pass files, we've developed a - convention called the standard rule format. - Its main requirement is that each element of a rule occupy a separate - line of text.  If - is possible to write rules such that all rule elements are on the same - line.  When - referencing nodes however, it is often helpful to have elements written - on separate lines.  

- -

The rules below are not in standard rule format. Notice that the elements - (_VP, _verb, _xWHITE, etc.) in each rule are all on the same line.

- -

- -

Converting to Standard Form

- -

Rules written on a single line can be converted using the Pass File - Popup Menu item One - Line to Standard Rule Format:

- -

- -

 

- -

This results in a rule in standard rule format:

- -

- -

where each rule element (_VP, _verb, _xWHITE, - _NP, _xWHITE, _PP) appears on its own line.

- -

Standard rule format has three distinctive features. Rule elements are - indented by a tab space, numbered with comments and displayed one per - line.

- -

Indentation

- -

Indentation is for readability.

- -

Rule Element Numbering

- -

Numbering rule elements with comments makes for a convenient reference - when using NLP++ node functions or functions that reference individual - elements in a rule. To see how this works, some NLP++ functions have been - added to the rule given above:

- -

- -

The NLP++ node function N("$text",1) refers to the text for - rule element 1, N("$text",3) the text for rule element 3, and - N("$text",5) the text for rule element 5. In standard rule format, - if a person deletes a rule element with the right-click menu, the rule - elements will be renumbered automatically to facilitate the user in renumbering - of the NLP++ node function "N" by hand. In the future, node - function renumbering will be automatic with such edits.

- -

One-Per-Line and Element Editing

- -

Having one rule element per line allows for the use of right-click menu - items for quick editing. In standard rule format, users can quickly change, - convert, and delete rule elements and rule element attributes. In the - example below, a rule element is made optional by clicking anywhere on - the rule element of interest, and selecting the Add/Remove attribute "optional".

- -

- -

 

- -

Doing the above yields:

- -

- -

See Also

- -

Rule - Syntax

- -

 

- - - - diff --git a/Help/helps/VisualText_Basics/Structure_of_Analyzer_Projects.htm b/Help/helps/VisualText_Basics/Structure_of_Analyzer_Projects.htm deleted file mode 100644 index 1696ef0..0000000 --- a/Help/helps/VisualText_Basics/Structure_of_Analyzer_Projects.htm +++ /dev/null @@ -1,497 +0,0 @@ - - - - - - - - - - -Structure_of_Analyzer Projects - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Structure of Analyzer Projects

- -

When an analyzer is created and saved in VisualText, a directory structure - is created in the working directory (e.g. C:\apps). This structure is - referred to as the analyzer project. The analyzer project stores - the analyzer and its related files.

- -

An analyzer project is organized by a single analyzer - project file. The analyzer project file is automatically created - when you create an analyzer. The name of the analyzer project file is - the name you give the analyzer project appended with a .ana - extension. The analyzer project file, also called the .ana (pronounced - "dot ana") file, is used to load the analyzer.

- -

NOTE 2.0.1.x - and later: Run, KB, and user DLL files are now renamed and moved - to the \bin folder when they are compiled.  (And - these must be compiled manually.)

- -

The folders and files in the analyzer project are outlined below.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Folder or - File

 -

Description

-

bin

 -

Executables and libraries. Stores the kb.dll - compiled knowledge base library.
-NEW in 2.0.1.x and later: DLLs have been renamed and moved to bin: kb.dll, - kbu.dll, run.dll, runu.dll, user.dll, useru.dll

-

data

 -

Miscellaneous data files created or used by - the app.

-

logs

 -

Directory for the analyzer log file.

-

input

 -

Input file tree.

-

kb

 -

The interpreted and compiled KBs (kb.dll) - are stored here.

-

output

 -

Directory for output files, dump files, log - files, intermediate output files, and so on.

-

rfa

 -

Output for the RFA, which parses the RFB analyzer - definition.

-

rfb

 -

Output for the RFB, which parses pass files.

-

run

 -

Directory for the compiled analyzer library, - run.dll.

-

spec

 -

Directory for the analyzer sequence file and - pass files.

-

user

 -

Directory for the user - MS Visual C++ project. Can be used to write new NLP++ functions and passes.

-

 

 -

 

-

cgerr.log

 -

KB error log file.

-

gui.kb

 -

KB command file that stores analyzer sequence - and input folder hierarchy.

-

main.cpp

 -

Code file for a standalone driver program - to call the analyzer.

-

NAME.ana

 -

The project file for the current analyzer - app.

-

Uanalyze.cpp

 -

Code file for standalone driver program.

-

Uanalyze.h

 -

Code file for standalone driver program.

- -

This folder structure houses all the information - needed to run the analyzer, including folders for input and output, a - folder for the knowledge base, etc.

- - - - diff --git a/Help/helps/VisualText_Basics/Writing_Analyzers_with_VisualText.htm b/Help/helps/VisualText_Basics/Writing_Analyzers_with_VisualText.htm deleted file mode 100644 index 8516095..0000000 --- a/Help/helps/VisualText_Basics/Writing_Analyzers_with_VisualText.htm +++ /dev/null @@ -1,82 +0,0 @@ - - - - - - - - - - -Writing Analyzers with VisualText - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Writing Analyzers with VisualText

- -

The VisualText Basics section includes a set of topics designed to help - you build analyzers. As a prerequisite, we recommend you work through - the Tutorials to understand the workings of the VisualText GUI. Also, - working through the Corporate analyzer documentation and studying the - Corporate analyzer will provide a feel for the ordering and specification - of the steps, or passes, in a text analyzer.

- -

Building text analyzers is more art than science. VisualText favors - a bottom up approach to text processing, where small constructs are processed - and grouped (or reduced) first, followed by identification of successively - larger components of a text.

- -

One approach that often works well for restricted domains is to create - a prototype that processes a single input text end-to-end, to produce - the desired analyzer output. In building such a prototype, you should - strive to build each pass in as general a way as possible. Once such a - prototype is built, a second input text is tested, and the analyzer is - reworked so that both inputs are processed in a general way. As more inputs - are handled in this fashion, the analyzer will perform better and do more - with unseen texts. In a restricted domain, even a small number of representative - input texts (e.g., 25-50) can yield dramatic improvements in the processing - of novel texts.

- -

Another aspect of this approach is that it is "knowledge free." - For a restricted domain, you don't usually need an ontology or knowledge - base for all human knowledge. You build or import knowledge as needed. - The broader the domain, of course, the more the analyzer will benefit - from large-coverage knowledge bases, such as WordNet.

- -

If you are interested in pre-built generic - text analysis capabilities for VisualText, you may wish to download one - of the versions of TAIParse available in the Apps section of our website - at

- -

http://www.textanalysis.com/Apps/apps.html

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/AnaArchViewLocal.gif b/Help/helps/VisualText_Interface/AnaArchViewLocal.gif deleted file mode 100644 index 914a682..0000000 Binary files a/Help/helps/VisualText_Interface/AnaArchViewLocal.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/AnaArchViewRemote.gif b/Help/helps/VisualText_Interface/AnaArchViewRemote.gif deleted file mode 100644 index 2d1ef43..0000000 Binary files a/Help/helps/VisualText_Interface/AnaArchViewRemote.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Interface_Intro.htm b/Help/helps/VisualText_Interface/Interface_Intro.htm deleted file mode 100644 index f73baef..0000000 --- a/Help/helps/VisualText_Interface/Interface_Intro.htm +++ /dev/null @@ -1,302 +0,0 @@ - - - - - - - - - - -Interface Intro - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Overview

- -

The major components of the VisualText user interface (or Main Window) - are called out below.

- -

- -

Apart from the Main Menu, you can resize and "undock" (i.e., - move about) the components shown.

- -

Interface Components

- -

A short description of each of the interface components is given below.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Interface

 -

Description

-

Main - Menu

 -

Functions accessed via pulldown menus.  The - Main Menu includes the following menus:  File, Edit, - Analyzer, KB, - View, Browser, - Parse Tree, Tools, - Window, Help.

-

Tab Window

 -

Controls access to the sample hierarchy, analyzer - sequence and text data files.

-

Log Window

 -

Displays processing information and error - messages.

-

Find Window

 -

Displays search information.

-

Workspace

 -

Area where input and output files, the knowledge - base and tree structures are displayed and manipulated.

-

Main Toolbar

 -

Commonly used functions accessed by buttons.

-

Tab Toolbar

 -

Iconic functions for the Tab Window.

-

Browser - Toolbar

 -

World Wide Web access and browser control.

-

Workspace - Toolbar

 -

Iconic access to the Analyzer, Knowledge Base, - and more.

-

Debug - Toolbar

 -

Functions to examine parse trees, intermediate - files, and output files.

- -

In addition to these objects, popup menus appear when you select a window - and right-click the mouse. See Popup - Menus for detailed description of available functions from popup menus.

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Knowledge_Base.htm b/Help/helps/VisualText_Interface/Knowledge_Base.htm deleted file mode 100644 index c8d903c..0000000 --- a/Help/helps/VisualText_Interface/Knowledge_Base.htm +++ /dev/null @@ -1,357 +0,0 @@ - - - - - - - - - - -Knowledge Base - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Introduction to the Knowledge Base

- -

The Conceptual Grammar (CG) Knowledge Base Management System (KBMS) - provides a permanent store for linguistic, conceptual, and domain knowledge. - The KBMS serves as a flexible "vanilla" framework in which users - may implement arbitrary representation schemes. Users can manage the KB - manually, while text analyzers can automatically access and update information - in the KB. The KB can be used to collect information across multiple texts, - for example, in training a categorizer or in managing an electronic chat.

- -

Concepts and Attributes in the Knowledge - Base

- -

The basic unit of knowledge in the KB is called a concept, - which is analogous to an object - in an object-oriented representation. A concept has attributes, - which consist of a key (or slot) and zero or more values. - The major organizing principle in the KB is a hierarchy - of concepts. Concepts can also refer to other concepts via attributes, - yielding a rich graph of knowledge.

- -

 A - unique feature of our Conceptual Grammar KBMS is that each concept may - optionally "own" a phrase, - which is a data structure that represents sequential information in a - convenient way. A phrase is composed of nodes, - which differ from concepts in that they are not attached to the hierarchy. - A node is a proxy or instance - of a concept in the hierarchy. Most NLP++ functions that operate on concepts - also operate on nodes. The main data structures in the KB are:

- - - --- - - - - - - - - - - - - - - - - - - - - - - - -
-

Data Structure

 -

Description

-

CONCEPT

 -

Organizing unit of knowledge in the KB.

-

ATTR

 -

Attribute key or slot.

-

VAL

 -

Value of an attribute.

-

PHRASE

 -

Sequence of nodes attached to a concept.

-

NODE

 -

A proxy or instance of a concept.

- -

Attributes may take on three types of values: integer, string, or concept. - A single attribute may mix values with any or all of these types, as desired.

- -

Phrases can be used to implement idioms, patterns, samples, rules, and - other sequential information. In the Gram Tab, the phrase associated with - a rule concept is used to represent user-highlighted samples. Each node - represents a single sample, with the name of the node doubling as the - sample's text.

- -

NLP++ Functions for the Knowledge Base

- -

See Table of KB Functions - for a listing of the available NLP++ functions for managing the KB.  These - functions are also treated individually in the Knowledge Base Functions - section.

- -

Concept Hierarchy

- -

The backbone of the KB is a knowledge hierarchy. Each concept therefore - has hard-coded parent and child links for efficiency.

- -

VisualText utilizes part of the concept hierarchy to implement the Gram Tab. The Gram - hierarchy (also called the sample - hierarchy) displayed in the Gram Tab is embedded directly in the - concept hierarchy. The KB as a whole is accessible in the KB - Editor:

- -

- -

When users highlight samples and add them to the - Gram Tab, they are modifying the KB. The left panel above shows the sample - hierarchy within the Gram Tab, while the right panel shows the same hierarchy - as it appears within the KB. In the KB Editor, a concept that contains - a phrase is colored yellow. For example, the adj concept contains a phrase, - where each node of the phrase represents one of the samples "quick", - "brown", etc. Therefore det, adj, noun, verb, and prep are all - colored yellow. To view a concept's phrase from the KB Editor, right click - on the concept and select Phrase Tool from the menu.

- -

Phrases

- -

A phrase is a sequence of nodes "owned" by a concept, and - is not visible from the concept hierarchy. For example, the concept "noun" - has the phrase shown below, consisting of the nodes "fox", "dog", - "John", and so on. (In the present example, the ordering of - nodes in the sequence is not important. But, in other uses of phrases, - for example to represent grammar rules, ordering is important.) A concept - can only have one phrase, with no restriction on the number of nodes. - The Phrase Tool, shown below, enables you to view a concept's phrase. - Each line in the Phrase Tool displays the name of one node, which can - be an arbitrary string.

- -

Phrase Tool

- -

The phrase belonging to a concept can be viewed in - the KB Editor by highlighting a yellow - concept in the hierarchy, right clicking and selecting Phrase Tool from - the Popup Menu. This displays the nodes of the phrase, one per line. In - the example below, we selected the concept 'noun' and then selected Phrase - Tool in the menu to display the phrase associated with 'noun'. In this - example, each node in the phrase represents a noun sample in the Gram - Tab.

- -

- -

Nodes

- -

A node is an instance of a concept, and does not - appear in the concept hierarchy. Nodes occur only as components of phrases - in the KB. The nodes belonging to the noun concept happen to be instances - of the KB dictionary concepts for "fox", "dog", etc.

- -

You can access the attributes for a particular node - by highlighting the node in the Phrase Tool and selecting Attributes, - or by double clicking on a line. This launches the Attribute - Editor, where the attributes and values for the node can be viewed - and edited. In the graphic below, the attributes for the selected node, - 'fox', are displayed.

- -

- - - - - - -
-

WARNING: - The Gram Tab is the only safe way to edit the Gram hierarchy. Editing - the gram or sys - hierarchies manually can easily corrupt the KB, so please make a backup - copy of your analyzer project before experimenting.

- -

 

- -

Accessing the Knowledge Base

- -

The KB may be managed in many ways. The KB Editor, Gram Tab, Attribute Editor, - and Phrase Tool provide interactive means for accessing the KB. NLP++ - function calls and C++ API function calls provide more direct access to - the internals of the knowledge base. (For more information on accessing - the KB, see the Table - of KB Functions and the page-per-functions in the Knowledge Base Functions - section .)

- - - - diff --git a/Help/helps/VisualText_Interface/Log_Window_Messages.htm b/Help/helps/VisualText_Interface/Log_Window_Messages.htm deleted file mode 100644 index 78a9d13..0000000 --- a/Help/helps/VisualText_Interface/Log_Window_Messages.htm +++ /dev/null @@ -1,544 +0,0 @@ - - - - - - - - - - -Log Window Messages - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Log Window Messages

- -

Two types of messages are given in the Log Window: informational - messages and error messages.

- -

The Log Window is divided into two parts: Events and Details.

- -

Informational Messages

- -

Here are some of the most common informational messages displayed in - the Log Window.

- - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Message

 -

Description

-

Loading stat: Initializing time

 -

Time basis for start of boot sequence timing.

-

Loading stat: Auto-load analyzer turned - off

 -

Time required to load IDE without an analyzer. - (See the General Preferences - for auto-loading an analyzer at startup.)

-

Loading stat:  Total - time

 -

Time required to load IDE.  

-

Loading stat:  Loading - "analyzerName" KB - NUM

 -

Time required to load the KB for analyzer - called analyzerName.

-

Loading stat:  Loading - "analyzerName" Analyzer - NUM

 -

Time required to load the analyzer.

-

Loading stat:  Finalizing - - NUM

 -

Time required to finish load sequence.

-

Loading stat:  Total - time: NUM

 -

Total time required to load knowledge base - and analyzer.

-

Analyzer Load: 0 0 [Date: Day, Month, Time, - Year]

 -

Day, date and time analyzer is loaded.

-

Analyzer Load:  0 - 0 [Read analyzer sequence from kb.]

 -

Indicates analyzer is being read from the - knowledge base.

-

Analyzer Load: 0 0 [Build analyzer time=NUM - sec]

 -

Total time to build analyzer.

-

Analyzing (NUM)  TextFile

 -

Name of text currently being analyzed.

-

Time = NUM

-

Time required to process text.

-

Analysis done (NUM) FilePath

 -

Name and directory path of analyzed file.

- -

Error Messages

- -

Error and warning messages are displayed in the Log Window. These messages - provide you with a way of seeing what is happening when the analyzer is - run. They can be a valuable tool when debugging your analyzer. Some of - the most common error messages are given below.

- - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Message

 -

Description

-

RightMatchSpecial unimplemented

 -

Indicates _xWILD may have been used at the - beginning and/or end of a rule's phrase. This is currently not supported. - Try _xANY instead. Or try _xWILD in combination with _xSTART and/or _xEND - such that _xWILD does not occur at the beginning or end of the phrase.

-

No Rule Reduce: Cant attach semantic data

 -

Indicates that an S variable will be discarded, - since the rule does not create a suggested node.

-

Check: Error in NLP++.
-Code actions: Error in NLP++.
-Post actions: Error in NLP++.

 -

Indicates a runtime error in a CODE, CHECK, - or POST region of a pass file. Possible errors are an NLP++ syntax error - or a mismatch between the code and the rules that it applies to. For example, - if an optional rule element matched no nodes, but the code attempts to - assign a variable to the element's node, this error message will appear.

-

Error. Analyzer pass '<NAME>' is uninterned

 -

Indicates that a pass in the analyzer sequence - contains syntax errors and is being skipped over.

-

Pat pass given no rules

 -

Indicates possible syntax error in a rule - of the pass file. Check for missing CODE Region labels.

- - - - diff --git a/Help/helps/VisualText_Interface/Main_Analyzer_Menu.htm b/Help/helps/VisualText_Interface/Main_Analyzer_Menu.htm deleted file mode 100644 index 084711e..0000000 --- a/Help/helps/VisualText_Interface/Main_Analyzer_Menu.htm +++ /dev/null @@ -1,939 +0,0 @@ - - - - - - - - - - -Main Analyzer Menu - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Analyzer Menu

- -

The Analyzer Menu controls the actions of an analyzer.  It - is also used in generating rules.

- -

- -

The Analyzer Menu corresponds to some elements in - the Workspace Toolbar.

- -

- -

When there is a corresponding element, the toolbar button is shown in - the following table:

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Menu Item

 -

Description

-

 -

Compile Analyzer

 -

Turning on this toggle, followed by loading - an analyzer, results in compiling the analyzer while it is being loaded. - This results in the creation of a RUN.DLL library, which can be called - from an external program. Currently, VisualText does not load the compiled - analyzer library. The Compile Analyzer toggle is turned off after an analyzer - is loaded.

-

 

 -

Generate Rules

 -

Submenu for generating rules. (See below.)

-

Run

 -

Run

 -

Runs the analyzer on the currently selected - input file. (Analyzers cannot be run on files with a .log extension.)

-

Run Next

 -

Run - Next

 -

Runs the analyzer on the next file in the - directory after the currently selected file.

-

Slog

 -

Slog

 -

Does a Generate Rules All, followed by a Run - on the currently selected input.

-

Stop Analyzer

 -

Stop

 -

Stops the current analyzer.

-

Toggle Generate Logs

 -

Generate - Logs

 -

When selected, enables generation of intermediate - logs for each pass in the analyzer sequence. Must be enabled before an - analyzer is run over text. This function gives you the ability to display - and inspect the intermediate parse trees, one for each pass, and the final - tree. It may slow the parse substantially in many-pass analyzers, since - all intermediate trees are written to disk. If the function is not enabled, - only the final tree is available for display.

-

 

 -

View

 -

Submenu for viewing text analyzer information - such as input files, output files, pass files, and parse trees. (See below.)

- -

Generate Rules Submenu

- -

- - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Menu Item

 -

Description

-

 

 -

All

 -

Exhaustively generates all rules as specified - by samples in the Gram Tab. - Depending on the number of samples in the Gram - Tab, this can be slow. However, to assure that rule generation - is correct and complete, this option should be used whenever feasible.

-

Generate Dirty

 -

Dirty

 -

Generates rules only for those rule concepts - in the Gram Tab that have the - Dirty attribute set. This attribute - is set by the system, for example, when new sample data is added to the - concept. This is a "quick-and-dirty" rule generation that most - often works well, but may not always agree with the Generate Rules > - All option. It is used for rapid prototyping. See Mark - for Generation in the Gram Tab Popup Menu, where the Dirty attribute - can be set.

-

 

 -

Unlocked

 -

Generates rules only for those concepts in - the Gram Tab that have the unlocked - attribute set. Locking concepts should only be done by the expert user. - A concept's attributes can be viewed and hand edited via the Attribute - Editor.

- -

Analyzer View Submenu

- -

- -

 

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Menu Item

 -

Description

-

View Concept Tree

 -

Concept Tree

 -

Displays the concept - tree of a parsed document in the Workspace.

-

View Current Text

 -

Current Text

 -

Displays the currently selected input data - file in the Workspace.

-

View Dump Files

 -

Dump Files

 -

Displays the output - files associated with a parsed document in the Workspace. When there are - multiple files, a pulldown tab to the right of the button lists available - files.

-

View Output File

 -

Output File

 -

Displays output.txt file associated with parsed - input document in the Workspace.  This - icon is reserved for files with the name output.txt only.

-

View Parse Tree

 -

Parse Tree

 -

Displays the full parse - tree of a parsed document in the Workspace.

-

View Pass File

 -

Pass File

 -

Displays selected pass - file in the Workspace.

-

View Web Page

 -

Web Page

 -

Displays selected html page in a browser window - in the Workspace.

- - - - diff --git a/Help/helps/VisualText_Interface/Main_Browser_Menu.htm b/Help/helps/VisualText_Interface/Main_Browser_Menu.htm deleted file mode 100644 index 5cf8537..0000000 --- a/Help/helps/VisualText_Interface/Main_Browser_Menu.htm +++ /dev/null @@ -1,349 +0,0 @@ - - - - - - - - - - -Main Browser Menu - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Browser Menu

- -

The Browser Menu provides access to the World Wide Web and various controls - for browsers.

- -

- -

The Browser Menu corresponds to some functions on the - Browser Toolbar.

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Menu Item

 -

Description

-

Back

 -

Back

 -

Loads previous Web page (if there is one) - in the history list. Function is disabled if no Urls are in the history - list.

-

Forward

 -

Forward

 -

Loads next Web page (if there is one) in the - history list. Function is disabled if no Urls are in the history list.

-

 

 -

Open URL

 -

Launches dialog box to enter URL. Web page - is opened in the Workspace.

-

Stop

 -

Stop

 -

Stops loading a Web page.

-

New Browser

 -

New Browser

 -

Opens a new browser window.  Browser - opens to location of existing open browser.  the - desired Web address can be typed directly into the Location Input Panel.

-

Save URL Text

 -

Save URL - Text

 -

Copies the URL name and html file to the Text - Tab. Function is enabled when a browser is selected in the Workspace.

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Main_Edit_Menu.htm b/Help/helps/VisualText_Interface/Main_Edit_Menu.htm deleted file mode 100644 index d8349bc..0000000 --- a/Help/helps/VisualText_Interface/Main_Edit_Menu.htm +++ /dev/null @@ -1,514 +0,0 @@ - - - - - - - - - - -Main Edit Menu - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Edit Menu

- -

The Edit Menu controls actions done to files in the Workspace.

- -

- -

The Edit Menu corresponds to some buttons on the Main - Toolbar.

- -

- -

When there is a corresponding element, the toolbar button is shown in - the following table:

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Menu Item

 -

Description

-

 

 -

Undo

 -

Reverses the last action performed in a file - in the Workspace.

-

 

 -

Redo

 -

Reverses an Undo action performed in a file - in the Workspace.

-

Cut

 -

Cut

 -

Deletes selected object in a Workspace file.

-

Copy

 -

Copy

 -

Copies selected object to the clipboard.

-

Paste

 -

Paste

 -

Pastes object from the clipboard at insertion - point in currently selected file.

-

Save

 -

Save

 -

Saves selected file in the Workspace.

-

Save All

 -

Save All

 -

Saves all unsaved files in the Workspace.

-

 

 -

Find

 -

Launches Find dialog box to search for item - in open file.

-

 

 -

Replace

 -

Launches Replace dialog box to find and replace - item in open file.

-

Find in Files

 -

Find in Files

 -

Launches Find - in Files dialog box to search for items across multiple files. File - type can be specified. Search results are displayed in the Find - Window.

-

 

 -

Replace in Files

 -

Launches Replace - in Files dialog box to find and replace items across multiple files. -  Search - and replace results are displayed in the Find - Window.

- - - - diff --git a/Help/helps/VisualText_Interface/Main_File_Menu.htm b/Help/helps/VisualText_Interface/Main_File_Menu.htm deleted file mode 100644 index 8d879a4..0000000 --- a/Help/helps/VisualText_Interface/Main_File_Menu.htm +++ /dev/null @@ -1,1920 +0,0 @@ - - - - - - - - - - -Main File Menu - - - - - - - - - - - - - - - - - - - - - - - - - - - -

File Menu

- -

The File Menu is context sensitive. Available menu items change when - a text file is open in the Workspace.  Here - is the File Menu when all options are available.  

- -

- -

The File Menu corresponds to some elements in the Main Toolbar.

- -

- -

When there is a corresponding element, the toolbar button is shown in - the following table:

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Menu Item

 -

Description

-

New

 -

New

 -

Creates a new file and displays it in the - Workspace window.

-

Open

 -

Open

 -

Opens a file and displays it in the Workspace.

-

 

 -

Close

 -

Closes currently selected file in the Workspace.

-

Save

 -

Save

 -

Saves currently selected file in the Workspace.

-

 

 -

Save As

 -

Saves currently selected file in the Workspace - with a new name.

-

 

 -

New Analyzer

 -

Creates a new analyzer. Only one analyzer - can be open at a time in an instance of VisualText. If an existing analyzer - is open, it will be closed and the new one loaded. Two instances of VisualText - can be launched however, thereby having two different analyzers running - at the same time.

-

 

 -

Open Analyzer

 -

Launches Open dialog box to navigate to location - of .ana - file to open the analyzer.

-

 

 -

Save Analyzer

 -

Saves currently loaded analyzer including - the Knowledge Base.

-

 

 -

Save Analyzer As

 -

Saves currently loaded analyzer with a new - name.

-

 

 -

Close Analyzer

 -

Closes currently loaded analyzer.

-

Print

 -

Print

 -

Prints currently selected file in the Workspace.

-

 

 -

Print Preview

 -

Prints preview of currently selected file - to the screen.

-

 

 -

Page / Print Setup

 -

Launches Page Setup dialog box with printing - options.

-

 

 -

Analyzer - Archive

 -

Submenu for archiving the analyzer. (See below.)

-

 

 -

Recent Files

 -

Lists recently opened files in a submenu. - Selected file is opened in the Workspace.

-

 

 -

Recent - Analyzers

 -

Submenu for selecting recently opened analyzers. - (See below.)

-

 

 -

Preferences

 -

Launches VisualText Preferences dialog box - to set various preferences.  (See - below.)

-

 

 -

Exit

 -

Exits the VisualText program.

- -

Analyzer Archive Submenu

- -

The archive functions allow you to load and store analyzers compactly - in zip files either on a local machine or on a server. Archiving is used - to create quick backups of your work and to facilitate communication between - developers working on the same analyzers. Files are transmitted via FTP. - The FTP parameters can be set by selecting the Archiving - tab in Preferences. When you archive directly to the server, a file is - created on your local disk. Information on uploading to the server is - displayed in the Log Window.

- -

- - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Archive Local

 -

Launches an Archive dialog to create a local - archive of the analyzer project file in a zip file. By default, the project - is archived in the project directory. The name of archive file defaults - to the name of the current analyzer, suffixed with the current date and - time. User has option to save C files and log files for analyzer session.

-

Archive - Remote

 -

Launches an Archive dialog to create a remote - archive of the analyzer project files in a zip file. An archive file is - also created locally. Local archive location defaults to the project directory. - The name of archive file defaults to the name of the current analyzer, - suffixed with the current date and time. User has option to save C files - and log files for analyzer session.

-

View Local

 -

Displays analyzers in the local archive, and - presents options to delete, rename, upload (send to server) or load into - current analyzer Workspace. Listings in archive can be sorted by clicking - on column headers.

-

View Remote

 -

Displays analyzers in the server archive, - and presents options to delete, rename, download (send to local archive) - or load into current analyzer Workspace.  Listings - in archive can be sorted by clicking on column headers.

- -

Archive Dialog

- -

- -

View Local

- -

- -

 

- -

View - Remote

- -

- -

Recent Analyzers Submenu

- -

Selecting Recent Analyzers will bring up the list of recently loaded - analyzers. The selected analyzer is opened in the Workspace. Selecting - an analyzer closes an already open analyzer. The analyzer listed at the - top of the list is the most current.

- -

- -

Preferences

- -

The Preferences menu item launches the VisualText Preferences dialog. -  The VisualText - Preferences dialog has tabs to set the following preferences:

- - - -

General Preferences

- -

The General Preferences tab allows you to set initialization, analyzer - and application directory information.

- -

- -

 

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Dialog Item

 -

Description

-

Auto Load - Last Analyzer

 -

If checked, the last analyzer loaded will - be loaded on program launch. The last analyzer loaded is displayed under - the check box. If the check box is cleared, no analyzer is loaded on program - launch.

-

Full Screen

 -

If checked, VisualText launches into full - screen. If the box is cleared, it launches in reduced screen.

-

Auto Save - KB on Logout

 -

If checked, automatically saves the knowledge - base when exiting an analyzer. If not checked, user is prompted to save - the knowledge base when exiting.

-

KB - Safe Edit Mode

 -

By default, checked when an analyzer is launched. -  Enabling - this mode prevents manual editing of the gram and sys hierarchy concepts - in the knowledge base, by greying out portions of KB - Editor Popup Menu when the gram or sys hierarchy concepts are selected. -  Manual - editing of the gram and sys hierarchies can easily corrupt the knowledge - base.  

-

Auto Comment - Name

 -

Name for current author or developer. Specified - name will be added to the header comment section when pass files are generated.

-

Apps Directory

 -

Path for the default directory for analyzer - projects.

-

Done

 -

Closes the VisualText Preferences dialog.

- -

 

- -

Display Preferences

- -

The Display Preferences tab allows you to set preferences for the VisualText - interface.

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Dialog Item

 -

Description

-

Size to Content

 -

If checked, window size for file when opened - in the Workspace is determined by the content of the file.

-

Size Dynamically

 -

If checked, then every window opened in the - Workspace grows as the user expands items in that window. (Note: Window - grows only to the default size specified in Preferences. Size to Content - must also be checked. See Programming - Notes and Issues.)

-

Rule Tab - Size

 -

Amount of indenting space in rule files.

-

Fixed Font - Size

 -

Default font size settings.

-

Width

 -

Default width of files opened in VisualText - Workspace.

-

Height

 -

Default height of files opened in VisualText - Workspace.

-

Done

 -

Closes the VisualText Preferences dialog.

- -

Email Preferences

- -

The Email Preferences tab is where you set email related information. -  Before - you can use the Email Tool, you - must these preferences.

- -

- - - - - - - - - - - - - - - - - - - -
-

Dialog Item

 -

Description

-

SMTP Server

 -

SMTP (Simple Mail Transport Protocol) address - of the server where email is received.

-

From Email

 -

Email address of author or developer.

-

Done

 -

Closes the VisualText Preferences dialog.

- -

Archiving Preferences

- -

The Archiving Preferences tab is where you set server and user information.

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Dialog Item

 -

Description

-

Author Name

 -

Name for archive author.

-

FTP Location

 -

FTP site of archive.  For - example, ftp.companyName.com.

-

Directory - Path

 -

Location of archive at FTP site. For example, - archiveDirectory/myArchives.

-

Username

 -

Name used to log into FTP site.

-

Password

 -

Password used to access FTP site.

-

Done

 -

Closes the VisualText Preferences dialog.

- -

Dictionary Lookup

- -

The Dictionary Lookup Preferences tab is where you set up preferences - for loading online dictionaries.  Dictionaries - specified here are used by the Dictionary - Lookup Tool.  The - URL path for the WordNet dictionary is set by default.  Specified - dictionaries are listed in the table once they are added.

- -

- - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Dialog Item

 -

Description

-

Name

 -

Input panel to specify the name of the online - dictionary.

-

URL

 -

Input panel to specify URL for online dictionary.  For - common online sites and more information on specifying URLs, see URLs - for Dictionary Lookup.

-

Add

 -

Adds specified online dictionary and URL to - the list of dictionaries.

-

Update

 -

Saves changes made to an existing dictionary entry.

-

 -

Moves table selection up one position.

-

 -

Moves table selection down one position.

-

Delete

 -

Deletes selected table item.

-

Done

 -

Closes the VisualText Preferences dialog.

- -

Analyzer Dependent

- -

The Analyzer Dependent Preferences tab is where you set analyzer dependent - preferences.  When - the Load Compiled KB box is checked, the compiled knowledge base is loaded - when the analyzer is launched.  

- -

If the knowledge base has been edited during a session, upon exit the - knowledge base will be compiled again.  If - Auto Save KB on Logout is not checked, you will be prompted with a confirmation - prompt.

- -

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Main_Help_Menu.htm b/Help/helps/VisualText_Interface/Main_Help_Menu.htm deleted file mode 100644 index d6f5f5b..0000000 --- a/Help/helps/VisualText_Interface/Main_Help_Menu.htm +++ /dev/null @@ -1,154 +0,0 @@ - - - - - - - - - - -Main Help Menu - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Help Menu

- -

The Help Menu provides access to the VisualText Help documentation and - to general information about VisualText.

- -

- - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

VisualText - Help

 -

Launches the VisualText - Help documentation.

-

Context Help

 -

Enables Context-Sensitive - Help.  

-

About - VisualText

 -

Launches VisualText product and version information.

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Main_KB_Menu.htm b/Help/helps/VisualText_Interface/Main_KB_Menu.htm deleted file mode 100644 index 0ead3a5..0000000 --- a/Help/helps/VisualText_Interface/Main_KB_Menu.htm +++ /dev/null @@ -1,379 +0,0 @@ - - - - - - - - - - -Main KB Menu - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

KB Menu

- -

The KB (Knowledge Base) Menu is used to control actions in the Knowledge - Base.  

- -

The Knowledge Base can store knowledge - of any type. VisualText itself uses the KB to manage information, including - the Concept Hierarchy, - the sequence of passes in the analyzer, the dictionary, and much more.

- -

- -

The KB Menu corresponds to the first part of the - Workspace Toolbar:

- -

- -

The corresponding element on the toolbar button is shown in the following - table:

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Menu Item

 -

Description

-

KB Editor

 -

KB Editor

 -

Launches the KB - Editor.

-

 -

Attribute Editor

 -

Launches the Attribute - Editor.

-

 -

Compile KB

 -

Compiles the Knowledge Base. Action results - in the creation of a KB.DLL. Preferences to load the compiled Knowledge - Base can be set in Preferences from the File Menu. The KB.DLL library - can be used with either an interpreted or compiled VisualText-built analyzer.

-

Move Concept Up

 -

Move Concept Up

 -

Moves a selected concept in the concept hierarchy - 'up' one position.

-

Move Concept Down

 -

Move Concept Down

 -

Moves selected concept in the concept hierarchy - down one position.

-

Save KB

 -

Save KB

 -

Makes changes to the KB permanent. Changes - made during a VisualText session are stored in memory. If KB is not saved, - all changes made are lost upon exiting.

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Main_Parse_Tree_Menu.htm b/Help/helps/VisualText_Interface/Main_Parse_Tree_Menu.htm deleted file mode 100644 index 5d6fa56..0000000 --- a/Help/helps/VisualText_Interface/Main_Parse_Tree_Menu.htm +++ /dev/null @@ -1,300 +0,0 @@ - - - - - - - - - - -Main Parse Tree Menu - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Parse Tree Menu

- -

The Parse Tree Menu controls actions with parse tree displays.  

- -

There are several ways to enable this menu. One way is by running an - analyzer on an input file, then selecting View - > Parse Tree from the Analyzer Menu or by clicking on the Parse Tree button .

- -

Another way is by selecting Analyzer - > View > Concept Tree or selecting the Concept - Tree button .

- -

Yet another way to activate the Parse - Tree Menu, is by right clicking on a text that has been analyzed - and selecting View > Parse Tree.

- -

- -

A tree display is affected by the selection of the Generate Logs button. - If the analyzer has been run with Generate Logs on, the parse tree for - the currently selected pass in the Ana Tab is displayed. If however, Generate - Logs is off, the final parse tree is displayed.

- -

These are the choices in the Parse - Tree Menu:

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Close All

 -

Collapses the tree display to the root node only.

-

Open All

 -

Expands all nodes in the tree display.

-

Open Item - All

 -

Expands all nodes under the selected node - in the tree display.

-

Open One - Level

 -

Expands one level under the selected node - in the tree display.

-

Highlight - Text

 -

Highlights the text in the input file corresponding - to the selected node in the tree display.

-

Display as - Root

 -

Displays the selected node as the root of - the tree display. Only the selected node is displayed.

-

Display Entire - Tree

 -

Displays the full parse tree, starting at - the root, assuming that the current selection in the Text Tab has been - analyzed.

-

Find Selected - Node in Files

 -

Searches for the selected node's name in the - project files. Results are displayed in the Find Window.

- -

The View > Node Variable Mode - menu item and the View > Offsets Mode - menu item both affect the display of a parse tree as well.  The - former displays variables and values for a node in a yellow box, and the - latter displays the text start and end offset for each node in the parse - tree.

- - - - diff --git a/Help/helps/VisualText_Interface/Main_Tools_Menu.htm b/Help/helps/VisualText_Interface/Main_Tools_Menu.htm deleted file mode 100644 index 6659f10..0000000 --- a/Help/helps/VisualText_Interface/Main_Tools_Menu.htm +++ /dev/null @@ -1,361 +0,0 @@ - - - - - - - - - - -Main Tools Menu - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tools Menu

- -

The Tools Menu controls the display of some of VisualText tools.  Tools - listed in this menu are also available from the Text - Tab Popup and the Text File Popup - menus.

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

ASCII - Table

 -

Displays the ASCII - Table. ASCII Table displays the decimal and hexadecimal numbers and - their corresponding ASCII characters.

-

Character Viewer

 -

Launches an Open dialog to navigate to a file. -  Once file - is selected, Character Viewer - is displayed. Character Viewer shows the line number, cumulative count - of total characters seen per line, characters on a line, and ASCII characters - for each line of text in selected file. Under each line is the hexadecimal - value of each character above it.

-

Dictionary

 -

Submenu to perform dictionary related functions. -  (See below.)

-

Email Tool

 -

Launches the Email - Tool. Email Tool enables sending email directly from the VisualText - interface. Preferences must be set in the Email tab of the VisualText - Preference dialog which can be accessed by selecting File - > Preferences.

-

Hex Viewer

 -

Launches an Open dialog to navigate to a file. -  Once file - is selected, Hex Viewer is displayed.

-

Text File Info

 -

Launches an Open dialog to navigate to a file. -  Once file - is selected, the Text File Info - dialog is displayed. This tool provides statistics about the selected - text file, including creation date, character, word and line counts.

- -

 

- -

Dictionary Submenu

- -

The Dictionary menu option allows you to perform dictionary related - functions.  

- -

- - - - - - - - - - - - - - - - - - - -
-

Menu - Item

 -

Description

-

Add Word

 -

Launches a dialog to add a word to the dict - hierarchy of the Knowledge Base. Words can only be added to the dict hierarchy - one at a time.  Word - added to Knowledge Base is noted in Log Window.  

-

KB Attributes

 -

Launches a dialog to find a word in the dict - hierarchy of the Knowledge Base and view its attributes.  If - there is an entry for the word in the dict hierarchy, the Attribute - Editor is launched.  If - the word does not exist in the hierarchy, a message is written to the - Log Window.

-

Lookup

 -

Displays list of online dictionaries specified - in the Dictionary Lookup - Preferences tab.  Selecting - an option in the Lookup menu launches a dialog to search for a word in - the selected online source.  See - Dictionary Lookup for more - information on using this tool.

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Main_View_Menu.htm b/Help/helps/VisualText_Interface/Main_View_Menu.htm deleted file mode 100644 index 650187e..0000000 --- a/Help/helps/VisualText_Interface/Main_View_Menu.htm +++ /dev/null @@ -1,921 +0,0 @@ - - - - - - - - - - -Main View Menu - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

View Menu

- -

The View Menu controls the display of various windows and toolbars in - the Interface.  The - View Menu is also used to control certain VisualText modes.

- -

- -

Part of the View Menu corresponds roughly to the following sections - of the Workspace Toolbar and the Debug - Toolbar

- -

- -

Another part corresponds to a piece of the Main - Toolbar:

- -

- -

When there is a corresponding element, the toolbar button is shown in - the following table:

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Menu Item

 -

Description

-

Toggle Batch Mode

 -

Batch - Mode

 -

Instructs VisualText to analyze groups of - files in an efficient mode. VisualText does not save intermediate and - output files in batch mode. The text analyzer must explicitly create a - location for output that is to be retained in batch mode. The NLP++ function - called batchstart() enables a text analyzer to check when the first text - in a batch run is being analyzed.

-

Toggle Highlight Mode

 -

Highlight - Mode

 -

When selected, enables highlighting of words - or phrases in a text file matching selected pass in the Ana Tab or selected - concept in the Gram Tab. Must be selected before an analyzer is run over - text. This tool allows you to quickly see whether and where a rule is - applying.

-

 -

Node - Variable Mode

 -

When selected, enables the display of node - variables in the parse tree.

-

 

 -

Offsets - Mode

 -

Enables the display of offset information - for terminal nodes in a parse tree display. Offset information is displayed - as two numbers.  For - example [0,10], indicates the first character of the node starts at character - position 0 (i.e. the first character on a line) and ends at character - position 10.

-

 

 -

Tree Text Mode

 -

Parse tree displays text based on what is - in the parse tree.  If - nodes are removed from the tree, such as HTML tags, then the parse tree - view removes the tag text as well.

-

Toggle Page Mode

 -

Page Mode

 -

Maintains coordinated sets of documents. For - example, when the user runs the analyzer, its corresponding output and - tree windows are updated automatically. Similarly when selecting passes - in the Ana Tab, Page Mode activates the Next Text and Previous Text buttons - to its right.

-

Toggle Verbose Mode

 -

Verbose - Mode

 -

Enables the display of details from the loading - of the analyzer. Results are displayed in the Log Window.  When - a folder of files is analyzed, the processing time for each file is displayed.

-

Next Text

 -

Next Text

 -

Displays the next input text file according - to the ordering in the Text Tab. Works in Page Mode to automatically update - the open documents associated with the input text.

-

Previous Text

 -

Previous - Text

 -

Displays previous input text file according - to the ordering in the Text Tab. Works in Page Mode to automatically update - the open documents associated with the input text.

-

 -

Status Bar

 -

Displays status or description of GUI objects - and actions.

-

 

 -

Toolbars

 -

Submenu that controls the visibility of toolbars - in the VisualText Main Window. (See below.)

-

Toggle Find Window

 -

Find Window

 -

Toggles the visibility of the Find Window. - By default, the Find Window is located in the bottom right portion of - the interface.

-

Toggle Log Window

 -

Log Window

 -

Toggles the visibility of the Log Window. - By default, the Log Window is located in the bottom left portion of the - interface.

-

Toggle Tab Window

 -

Tab Window

 -

Toggles the visibility of the Tab Window. - By default, the Tab Window is located in the middle left portion of the - interface.

- -

View Toolbars Submenu

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Browser

 -

Controls the display of the Browser Toolbar.

-

Debug

 -

Controls the display of the Debug Toolbar.

-

Main

 -

Controls the display of the Main Toolbar.

-

Tab

 -

Controls the display of the Tab Toolbar.

-

Workspace

 -

Controls the display of the Workspace Toolbar.

- - - - diff --git a/Help/helps/VisualText_Interface/Main_Window_Menu.htm b/Help/helps/VisualText_Interface/Main_Window_Menu.htm deleted file mode 100644 index 52f1005..0000000 --- a/Help/helps/VisualText_Interface/Main_Window_Menu.htm +++ /dev/null @@ -1,221 +0,0 @@ - - - - - - - - - - -Main Window Menu - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Window Menu

- -

The Window Menu controls the arrangement and display of open windows - in the Workspace.  

- -

- -

These options arrange the windows in the Workspace in standard ways. - The bottom part of the menu consists of a list of windows and tools that - are currently open in the Workspace. Selecting one of these brings the - corresponding window to the foreground.

- - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Cascade

 -

Arranges the open windows in the Workspace - in a staggered or cascaded sequence.

-

Tile Horizontally

 -

Arranges the open windows in the Workspace - in a horizontal stack.

-

Tile Vertically

 -

Arranges the open windows in the Workspace - in a vertical stack.

-

Arrange Icons

 -

Arranges the minimized icons at the bottom - of the Workspace window.

-

Close All

 -

Closes all the open windows in the Workspace.

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Menus/Help_menu.gif b/Help/helps/VisualText_Interface/Menus/Help_menu.gif deleted file mode 100644 index 45baf92..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/Help_menu.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/analyzer_menu.gif b/Help/helps/VisualText_Interface/Menus/analyzer_menu.gif deleted file mode 100644 index 5a0738e..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/analyzer_menu.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/analyzer_menu_genrules.gif b/Help/helps/VisualText_Interface/Menus/analyzer_menu_genrules.gif deleted file mode 100644 index fb19b57..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/analyzer_menu_genrules.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/analyzer_view_menu.gif b/Help/helps/VisualText_Interface/Menus/analyzer_view_menu.gif deleted file mode 100644 index 6adf672..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/analyzer_view_menu.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/browser_menu.gif b/Help/helps/VisualText_Interface/Menus/browser_menu.gif deleted file mode 100644 index dc090ba..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/browser_menu.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/edit_menu0610.gif b/Help/helps/VisualText_Interface/Menus/edit_menu0610.gif deleted file mode 100644 index 0f29d62..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/edit_menu0610.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/file_menu_all.gif b/Help/helps/VisualText_Interface/Menus/file_menu_all.gif deleted file mode 100644 index 5b483ff..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/file_menu_all.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/file_menu_analyzer_archive.gif b/Help/helps/VisualText_Interface/Menus/file_menu_analyzer_archive.gif deleted file mode 100644 index 4bc4b02..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/file_menu_analyzer_archive.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/file_menu_recentanalyzers.gif b/Help/helps/VisualText_Interface/Menus/file_menu_recentanalyzers.gif deleted file mode 100644 index b120d1b..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/file_menu_recentanalyzers.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/kb_menu.gif b/Help/helps/VisualText_Interface/Menus/kb_menu.gif deleted file mode 100644 index dff5436..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/kb_menu.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/parse_tree-menu.gif b/Help/helps/VisualText_Interface/Menus/parse_tree-menu.gif deleted file mode 100644 index a798f96..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/parse_tree-menu.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/tools_dictionary.gif b/Help/helps/VisualText_Interface/Menus/tools_dictionary.gif deleted file mode 100644 index c3e51f2..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/tools_dictionary.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/tools_menu.gif b/Help/helps/VisualText_Interface/Menus/tools_menu.gif deleted file mode 100644 index f163c6e..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/tools_menu.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/viewMenu.gif b/Help/helps/VisualText_Interface/Menus/viewMenu.gif deleted file mode 100644 index 4da2baf..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/viewMenu.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Menus/viewMenuToolbars.gif b/Help/helps/VisualText_Interface/Menus/viewMenuToolbars.gif deleted file mode 100644 index ac8ad78..0000000 Binary files a/Help/helps/VisualText_Interface/Menus/viewMenuToolbars.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/Parse_Tree_Popup.htm b/Help/helps/VisualText_Interface/Popups/Parse_Tree_Popup.htm deleted file mode 100644 index c5bb6cf..0000000 --- a/Help/helps/VisualText_Interface/Popups/Parse_Tree_Popup.htm +++ /dev/null @@ -1,103 +0,0 @@ - - - - - - - - - - -Parse Tree Popup - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Parse Tree Popup

- -

The Parse Tree Popup menu is launched by selecting an item in a parse - tree file in the Workspace and right mouse clicking.

- -

- -

The Parse - Tree Menu on the Main Menu describes all of the items in this submenu.

- -

Toggle Node Variables  Toggle - Node Variables button

- -

With View > Node Variable Mode selected - (or the Toggle Node Variables button depressed on the toolbar), placing - the cursor near a parse tree node displays a yellow box with the node's - variables and values, as below.

- -

- - - - diff --git a/Help/helps/VisualText_Interface/Popups/PassFilePopup_ChangeTo.gif b/Help/helps/VisualText_Interface/Popups/PassFilePopup_ChangeTo.gif deleted file mode 100644 index af9d75e..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/PassFilePopup_ChangeTo.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/Pass_File_Popup.htm b/Help/helps/VisualText_Interface/Popups/Pass_File_Popup.htm deleted file mode 100644 index 6a00215..0000000 --- a/Help/helps/VisualText_Interface/Popups/Pass_File_Popup.htm +++ /dev/null @@ -1,926 +0,0 @@ - - - - - - - - - - -Pass File Popup - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Pass File Popup

- -

The Pass File Popup menu is launched by selecting a pass file in the - Workspace and right mouse clicking. Many of the rule-editing actions assume - that rules are in written in standard - rule format.

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Add/Remove

 -

Submenu for adding or removing phrase - element modifiers at the selected line of a pass file. (See below.)

-

Change - to

 -

Submenu for editing special - elements in the current rule. (See below.)

-

Comment/Uncomment

 -

Comments or uncomments the currently selected - line of a pass file.

-

Convert

 -

Submenu for converting text to rules and for - converting rules to standard format. (See below.)

-

Delete Line

 -

Deletes selected line in a pass file.

-

Duplicate - Line

 -

Duplicates selected line in a file and inserts - it below the current line.

-

Insert

 -

Submenu for inserting a commonly used rule - components at the current line. (See below.)

-

Insert - Line

 -

Submenu for inserting a special rule element - below the current line. (See below.)

-

Reformat - Standard Rule

 -

Reformats a rule written in standard - rule format. Cursor must be somewhere within the rule. This accelerator - normalizes rules whose indentation and numbering have been altered by - previous editing operations or by hand-editing.

-

Find

 -

Submenu for finding selected item in a pass - file. (See below.)

-

Highlight - Matches

 -

Highlights text in the current input text - file based on the selection in the current pass file. If a rule element - is selected in the pass file, then matches in the parse tree are used - to highlight corresponding portions of the input text. If no item is selected - in the pass file, the user is prompted to supply a rule element.

-

Select Analyzer - Pass

 -

Selects the pass in the analyzer sequence - corresponding to the open pass file.  Selected - pass is highlighted in the Ana Tab if Ana Tab is currently open.  If - either the Gram Tab or the Text Tab are open, changes the display to the - Ana Tab.

-

Run

 -

Runs the analyzer on the currently selected - input file. (Analyzers can not be run on files with a .log extension.) - Same as Run Button on Workspace - Toolbar and Run in the  Analyzer - Menu.

-

Save

 -

Saves selected pass file. Same as Save - in the Edit Menu and the Save - button on the Main Toolbar.

-

View Log

 -

If an intermediate log file is available, - displays log file corresponding to the currently selected pass file. If - intermediate log file is not available, displays the final.log file.

-

Close Rest

 -

Closes all windows in the Workspace except - currently selected one.

-

Help on Selection

 -

Displays Help documentation on selected item.

-

NLP++ Help

 -

Launches VisualText Help documentation for - NLP++.

-

Pass File - Help

 -

Displays Help for writing rules in pass files.

- -

Add/Remove Submenu

- -

The Add/Remove submenu adds - or removes element modifiers for the rule element on the selected line. - See Phrase Element - Modifiers for detailed explanations of the various element actions.

- -

- -

 

- -

Change To Submenu

- -

The Change To submenu changes - the name of the selected special rule element to the selected special - rule element name (e.g., _xWILD).  See - Special Rule Elements - for detailed explanations of the various element actions.

- -

- -

Convert Submenu

- -

Accelerators in the Pass File Popup - menu operate on rules laid out with one element per line. The Convert - submenu can format rules to the desired standard form the accelerators - expect.

- - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

One - Line to Standard Rule Format

 -

Rewrites single line rule into the Standard - Rule Format. Item is activated if cursor is on a rule whose elements - are written on the same line. (See example below.)

-

String - to Literal Rule

 -

Converts a string of text into a literal rule. - (See example below.)

-

String - to General Rule

 -

Converts a string of text into a general rule. - (See example below.)

- -

 

- -

Convert One Line to - Standard Rule Format

- -

This editing tool changes a single line rule into a multi-line rule, - with one rule element per line. Breaking up the rule elements like this - makes writing NLP++ rules easier, since statements that refer to rule - elements refer to them by number, and counting rule elements across a - single line seems more difficult than counting lines.

- -

Using One Line to Standard Rule Format - to format a one line rule like the one here:

- -

- -

results in the following rule format:

- -

- -

 

- -

Convert String to Literal Rule

- -

This menu item converts a string into a literal rule, inserting default - white spaces and rule element modifiers as well as comment symbols and - rule element numbers.

- -

Selecting String to Literal Rule for - the text string, "This is a test"

- -

- -

results in the following rule:

- -

- -

Convert String to General Rule

- -

Convert String to General Rule - is similar to String to Literal Rule except that rule elements are abstracted - to their variable types. Here's the output on the same example input, - This is a test:

- -

- -

Insert Submenu

- -

The Insert submenu inserts a - commonly used rule component at the current line. The submenu includes - the most commonly used elements and also a default for directing the output - to a file (Dump Phrase to File). The Header - menu item inserts a comment line into the file.

- -

- -

Insert Line Submenu

- -

The Insert Line submenu allows - you to insert rule elements into a rule. The selected rule element is - inserted on the line below the line the cursor is on. The selected rule - elements are _xALPHA, - _xNUM, - _xSTART, - _xWHITE, - and _xWILD.

- -

- -

Find Submenu

- -

- - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Find Selected - Text

 -

If item in pass file is highlighted, finds - highlighted item in the pass file. If no item is selected, launches a - dialog box to enter words or phrases to be searched in currently selected - file. Target is highlighted in text.

-

Find Again

 -

Searches for next occurrence of find target - in currently selected file. Next found target is highlighted in text.

-

Find in Files

 -

Launches a dialog box to enter words or phrases - to be searched in multiple files in current analyzer. Search results are - printed in the Find Window.

-

Find in Log

 -

If item in pass file is highlighted, searches - for highlighted item in the log file corresponding to currently selected - pass in the analyzer sequence. If no item is highlighted, launches a dialog - box to enter words or phrases to be searched in the corresponding log - file for currently selected analyzer pass. The log file contains the file - version of the pass output. Results are displayed in the log file in the - Workspace.

- - - - diff --git a/Help/helps/VisualText_Interface/Popups/Popup_Menus.htm b/Help/helps/VisualText_Interface/Popups/Popup_Menus.htm deleted file mode 100644 index 6f4c20d..0000000 --- a/Help/helps/VisualText_Interface/Popups/Popup_Menus.htm +++ /dev/null @@ -1,218 +0,0 @@ - - - - - - - - - - -Popup Menus - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Popup Menus

- -

There are a number of popup menus in VisualText. The main popup menus - are provided for the following: Ana Tab, Gram - Tab, Text Tab, Text Files, - Pass Files, Parse Trees - and the KB Editor. To see a description of each - of these menus, click on the titles below.

- -

Ana Tab Popup

- -

The Ana Tab Popup controls activity in the analyzer sequence. To bring - up this menu, select a pass in the Ana Tab and right mouse click.

- -

- -

 

- -

Gram Tab Popup

- -

The Gram Tab Popup controls actions on the Gram Hierarchy. To bring - up this menu, select a concept in the - Gram Hierarchy and right mouse click.

- -

- -

 

- -

Text Tab Popup

- -

The Text Tab Popup organizes functions for operation in the Text Tab. - To bring up this menu, select any folder or text file in the Text Tab - and right mouse click.

- -

- -

Text File Popup

- -

The Text File Pop controls actions to a selected text file in the Workspace. - To bring up this menu, select the text file in the Workspace and right - mouse click.

- -

- -

Pass File Popup

- -

The Pass File Popup controls functions for editing the rules contained - within the pass file. To open this menu, open a pass file in the Workspace - and right mouse click.

- -

- -

 

- -

Parse Tree Popup

- -

The Parse Tree Popup controls actions to an item in a parse tree display. - To bring up this menu, select an item in a parse tree display in the Workspace - and right mouse click.

- -

- -

KB - Editor Popup

- -

The KB Editor Popup controls actions to items in the Knowledge Base. - To bring up this menu, right click in the body of the KB Editor.

- -

- -

.

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Popups/TextTabPopup_Add.gif b/Help/helps/VisualText_Interface/Popups/TextTabPopup_Add.gif deleted file mode 100644 index 42ef061..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/TextTabPopup_Add.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/TextTabPopup_Add_TopLevel.gif b/Help/helps/VisualText_Interface/Popups/TextTabPopup_Add_TopLevel.gif deleted file mode 100644 index 9b5b4ef..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/TextTabPopup_Add_TopLevel.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/Text_File_Popup.htm b/Help/helps/VisualText_Interface/Popups/Text_File_Popup.htm deleted file mode 100644 index 4e91017..0000000 --- a/Help/helps/VisualText_Interface/Popups/Text_File_Popup.htm +++ /dev/null @@ -1,971 +0,0 @@ - - - - - - - - - - -Text File Popup - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Text File Popup

- -

The Text File Popup menu is launched by selecting a text file in the - Workspace and right mouse clicking.

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

Find Selected - Text

 -

Searches for highlighted text in file.  If - no text is highlighted, launches a dialog box to enter search term. Found - search term is highlighted in text.

-

Find Again

 -

Searches for next occurrence of find target - in currently selected file. Next found target is highlighted in text.

-

Find in Files

 -

Searches for highlighted text across multiple - files in the current analyzer.  If - no word or phrase is highlighted in file, launches a dialog box to enter - search terms. Search results are printed in the Find Window.

-

Find in Log

 -

Searches for highlighted text in log files. -  If no word - or phrase is highlighted in file, launches a dialog box to enter search - terms. If an intermediate log file is available, search is performed in - the log file corresponding to the currently selected pass in the analyzer - sequence. If intermediate log file is not available, the final.log file - is searched.

-

Close Rest

 -

Closes all windows in the Workspace except - currently selected one.

-

Dictionary

 -

Submenu to perform dictionary related functions. -  (See below.)

-

Generate - Rule

 -

Automatically creates a rule from the currently - selected text. Appends rule to the currently selected pass file in the - Ana Tab. Generate Rule only works if you have previously run the analyzer - on the text with the "Generate Logs" toggle turned on

-

Run

 -

Runs the analyzer on the currently selected - input file. (Note: analyzers can not be run on files with a .log extension.) - Same as the Run Button on Workspace - Toolbar and selecting Run from - the Analyzer Menu.

-

Sort Unique - Text

 -

Sorts words in the currently selected text - file in ascending alphabetic order. Multiple occurrences of words are - discarded. Results are displayed in the currently selected text file. - Useful in building sample files.

-

Tools

 -

Submenu for accessing VisualText tools. (See - below.)

-

View

 -

Submenu for viewing parse tree, output - files and parse tree of selected text.

-

Wrap Text

 -

Launches a dialog to specify the maximum number - of characters per line, then "wraps" the text lines to that - limit.

-

Save

 -

Saves currently selected file in the Workspace.

-

Save As

 -

Saves currently selected file in the Workspace - with a new name.  Refresh - the Text Tab to see the new file.

- -

Dictionary Submenu

- -

The Dictionary menu option allows you to perform dictionary related - functions.

- -

- - - - - - - - - - - - - - - - - - - -
-

Menu - Item

 -

Description

-

Add Word

 -

Adds the highlighted word in a file to the - dict hierarchy of the Knowledge Base. Words can only be added to the dict - hierarchy one at a time.  Only - single words can be added. If the word already exists in the hierarchy - a message is written to the Log Window.  Log - message is also written indicating which word is added.

-

KB Attributes

 -

Highlighting a word in the text file and selecting - KB Attributes launches the Attribute - Editor.  Attribute - Editor is only launched if highlighted word exists in the dict hierarchy - of the Knowledge Base.

-

Lookup

 -

Displays list of online dictionaries specified - in the Dictionary - Lookup Preferences tab.  Selecting - an option in the Lookup menu starts a search in the online dictionary - source.  See - Dictionary Lookup for - more information on using this tool.

- -

  

- -

Tools Submenu

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

ASCII - Table

 -

Displays the ASCII - Table. ASCII Table displays the decimal and hexadecimal numbers and - their corresponding ASCII characters.

-

Character Viewer

 -

Displays the Character - Viewer for the selected file. Character Viewer shows the line number, - cumulative count of total characters seen per line, characters on a line, - and ASCII characters for each line of text in selected file. Under each - line is the hexadecimal value of each character above it.

-

Dictionize

 -

Activates the Dictionize - Tool.  

-

Email

 -

Launches the Email - Tool. Email Tool enables sending email directly from the VisualText - interface.

-

Hex Viewer

 -

Launches the Hex - Viewer.

-

Jumble

 -

Activates the Jumble - Tool.  This - tool jumbles the letters in words in a text file.  This - tool is useful when looking for structural clues in a text, since it removes - distracting content.  Paragraph - and sentence structure are preserved.  

-

Mask

 -

Activates the Mask - Tool.  This - tool takes alphabetic content of a text with strings of the single letter - 'a', but leaves the form, the non-alphabetic structure, including punctuation - and white spaces. This tool is useful when looking for structural clues - in a text, since it removes distracting content.

-

Text File Info

 -

Displays the Text - File Info dialog. This tool provides statistics about the selected - text file, including creation date, character, word and line counts.

- -

 

- -

View Submenu

- -

- - - - - - - - - - - - - - - - - - - -
-

Menu - Item

 -

Description

-

Parse Tree

 -

Displays the full parse tree of a parsed document - in the Workspace. This includes character position and white spaces in - the document.

-

Output File

 -

Displays output.txt file associated with parsed - input document in the Workspace.

-

Tree of Selected

 -

Displays parse tree for selected portion in - a text file in the Workspace.

- -

 

- -

Example of Tree of Selected

- -

- - - - diff --git a/Help/helps/VisualText_Interface/Popups/ana_tab_popup.gif b/Help/helps/VisualText_Interface/Popups/ana_tab_popup.gif deleted file mode 100644 index 0d018a9..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/ana_tab_popup.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add.gif b/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add.gif deleted file mode 100644 index bd230db..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add_librarypass.gif b/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add_librarypass.gif deleted file mode 100644 index 223cc85..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add_librarypass.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add_toplevel.gif b/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add_toplevel.gif deleted file mode 100644 index 24ce7d5..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add_toplevel.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add_toplevel_librarypass.gif b/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add_toplevel_librarypass.gif deleted file mode 100644 index 0b6eb2a..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_add_toplevel_librarypass.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_archive.gif b/Help/helps/VisualText_Interface/Popups/ana_tab_popup_archive.gif deleted file mode 100644 index addeaec..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_archive.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_properties.gif b/Help/helps/VisualText_Interface/Popups/ana_tab_popup_properties.gif deleted file mode 100644 index 44b6e43..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/ana_tab_popup_properties.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/ana_tap_popup_view.gif b/Help/helps/VisualText_Interface/Popups/ana_tap_popup_view.gif deleted file mode 100644 index d26e54f..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/ana_tap_popup_view.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/gram_tab_popup0725.gif b/Help/helps/VisualText_Interface/Popups/gram_tab_popup0725.gif deleted file mode 100644 index 94498eb..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/gram_tab_popup0725.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/gram_tab_popup_add.gif b/Help/helps/VisualText_Interface/Popups/gram_tab_popup_add.gif deleted file mode 100644 index ca460b3..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/gram_tab_popup_add.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/gram_tab_popup_archive.gif b/Help/helps/VisualText_Interface/Popups/gram_tab_popup_archive.gif deleted file mode 100644 index fd4d1eb..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/gram_tab_popup_archive.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/gram_tab_popup_samplefile.gif b/Help/helps/VisualText_Interface/Popups/gram_tab_popup_samplefile.gif deleted file mode 100644 index 6781d3b..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/gram_tab_popup_samplefile.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/kb_editor_pop0814.gif b/Help/helps/VisualText_Interface/Popups/kb_editor_pop0814.gif deleted file mode 100644 index 8bcf068..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/kb_editor_pop0814.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/parse_tree_popup.gif b/Help/helps/VisualText_Interface/Popups/parse_tree_popup.gif deleted file mode 100644 index d825539..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/parse_tree_popup.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/pass_file_popup.gif b/Help/helps/VisualText_Interface/Popups/pass_file_popup.gif deleted file mode 100644 index 65ab703..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/pass_file_popup.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/pass_file_popup_add_remove.gif b/Help/helps/VisualText_Interface/Popups/pass_file_popup_add_remove.gif deleted file mode 100644 index 8f7868d..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/pass_file_popup_add_remove.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/pass_file_popup_example_rule2.gif b/Help/helps/VisualText_Interface/Popups/pass_file_popup_example_rule2.gif deleted file mode 100644 index 73746ff..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/pass_file_popup_example_rule2.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/pass_file_popup_find_sub.gif b/Help/helps/VisualText_Interface/Popups/pass_file_popup_find_sub.gif deleted file mode 100644 index 9bf4236..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/pass_file_popup_find_sub.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/pass_file_popup_insert_submenu.gif b/Help/helps/VisualText_Interface/Popups/pass_file_popup_insert_submenu.gif deleted file mode 100644 index 248ea95..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/pass_file_popup_insert_submenu.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/pass_file_popup_insertlines.gif b/Help/helps/VisualText_Interface/Popups/pass_file_popup_insertlines.gif deleted file mode 100644 index a1e38ad..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/pass_file_popup_insertlines.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/pass_file_popup_rule_example.gif b/Help/helps/VisualText_Interface/Popups/pass_file_popup_rule_example.gif deleted file mode 100644 index d8638fe..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/pass_file_popup_rule_example.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/pass_file_popup_string_to_lit1.gif b/Help/helps/VisualText_Interface/Popups/pass_file_popup_string_to_lit1.gif deleted file mode 100644 index 596b989..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/pass_file_popup_string_to_lit1.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/pass_file_popup_string_to_lit2.gif b/Help/helps/VisualText_Interface/Popups/pass_file_popup_string_to_lit2.gif deleted file mode 100644 index d2610b8..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/pass_file_popup_string_to_lit2.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/pass_file_popups_string_to_general.gif b/Help/helps/VisualText_Interface/Popups/pass_file_popups_string_to_general.gif deleted file mode 100644 index 4b0d466..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/pass_file_popups_string_to_general.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/text_file_popup.gif b/Help/helps/VisualText_Interface/Popups/text_file_popup.gif deleted file mode 100644 index ade5d22..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/text_file_popup.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/text_file_popup_dictionary.gif b/Help/helps/VisualText_Interface/Popups/text_file_popup_dictionary.gif deleted file mode 100644 index b31b3cd..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/text_file_popup_dictionary.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/text_file_popup_tools.gif b/Help/helps/VisualText_Interface/Popups/text_file_popup_tools.gif deleted file mode 100644 index 1ce1ae0..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/text_file_popup_tools.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/text_file_popup_treeselected_eg.gif b/Help/helps/VisualText_Interface/Popups/text_file_popup_treeselected_eg.gif deleted file mode 100644 index 6e9614a..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/text_file_popup_treeselected_eg.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/text_file_popup_view.gif b/Help/helps/VisualText_Interface/Popups/text_file_popup_view.gif deleted file mode 100644 index 1b878a2..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/text_file_popup_view.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/text_tab_dictionary.gif b/Help/helps/VisualText_Interface/Popups/text_tab_dictionary.gif deleted file mode 100644 index fcd4684..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/text_tab_dictionary.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/text_tab_popup.gif b/Help/helps/VisualText_Interface/Popups/text_tab_popup.gif deleted file mode 100644 index b71dde8..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/text_tab_popup.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/text_tab_popup_archive.gif b/Help/helps/VisualText_Interface/Popups/text_tab_popup_archive.gif deleted file mode 100644 index c944f61..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/text_tab_popup_archive.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Popups/text_tab_popup_tool.gif b/Help/helps/VisualText_Interface/Popups/text_tab_popup_tool.gif deleted file mode 100644 index 872ff6d..0000000 Binary files a/Help/helps/VisualText_Interface/Popups/text_tab_popup_tool.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tab.htm b/Help/helps/VisualText_Interface/Tab.htm deleted file mode 100644 index 787fe2b..0000000 --- a/Help/helps/VisualText_Interface/Tab.htm +++ /dev/null @@ -1,255 +0,0 @@ - - - - - - - - - - -Tab - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tab Window

- -

There are three tabs in the Tab Window, - Ana, Gram, and Text

- -

- -

Ana Tab

- -

The Ana Tab shows the analyzer sequence. This is where you control the - text analyzer sequence, passes, and pass files. You can operate on the - objects in the Ana Tab with buttons on the Tab - Toolbar and with the Ana Tab Popup - Menu.

- -

- -

There are two types of passes:

- -

system - pass
-user created pass

- -

System Pass

- -

Currently, the only system pass available is the tokenize - pass. System passes are identified by the rectangle icon.  System - passes cannot be modified. The tokenize executes first, traversing - the input text to convert it to a structure called a parse - tree. Unlike user created passes, the tokenize and pass does not - have an associated pass file.

- -

User - Created Pass

- -

User created passes are added after the default system pass. User created - passes can be identified by the DNA chain icon. The two types of user - created passes are Pattern and - Recursive. A pass in the analyzer - sequence that is recursive is indicated by an R - in the analyzer sequence.

- -

A region of passes that have been generated automatically is called - a stub region. The stub region - is delimited by two red markers (labeled stub_NAME and end_NAME) in the - analyzer sequence.

- -

 

- -

Gram Tab

- -

The Gram Tab is used to construct a concept (or sample) hierarchy and - to sample data from input files to the concept hierarchy. You can operate - on the objects in the Gram Tab by selecting them and using toolbar buttons - and the Gram Tab Popup Menu.

- -

- -

The Gram Tab can have five types of concepts:

- - - -

Stub - Concept

- -

The top level concept is the stub concept. It is identified by the DNA - chain icon. The stub concept is associated with a region of automatically - generated passes in the analyzer sequence. Folder and rule concepts can - be direct descendants of stub concepts.

- -

Folder - Concept

- -

The second level concept is the folder concept. This - concept uses the folder icon. Folder concepts help organize related - sets of samples. Folder and rule concepts can be direct - descendants of folder concepts.

- -

Rule - Concept

- -

The third level concept is the rule concept. Rule - concepts are identified by the right-pointing arrow. Rule concepts - own a set of related samples, from which the rule generator will generalize, - merge, and generate rules. Label concepts and sample - concepts can be direct descendants of rule concepts.

- -

Label - Concept

- -

The fourth concept type is the label concept. Label - concepts are identified by the triangle icon. Label concepts hold - subsamples of larger samples. Label concepts can only be direct descendants - of rule concepts.

- -

Sample - Concept

- -

The final concept type is the sample concept. Samples - are identified by a box icon. A sample concept stores a single - sample or subsample. Sample concepts are placed under rule concepts and - label concepts. (Note: in the KB, sample concepts are not directly in - the hierarchy, but are implemented as phrase nodes owned by rule concepts - and label concepts.)

- -

 

- -

Text Tab

- -

The Text Tab is where you organize and view input and output text data - files. You can operate on the objects in the Text Tab by selecting them - and using the Text Tab Popup Menu, and with various toolbar buttons.

- -

- -

The Text Tab contains folders and files. When a text file has been analyzed, - it will have a checkmark on it. Html files can be distinguished from the - other text files by the globe icon.

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Toolbars/Browser_Toolbar.htm b/Help/helps/VisualText_Interface/Toolbars/Browser_Toolbar.htm deleted file mode 100644 index 8f8c174..0000000 --- a/Help/helps/VisualText_Interface/Toolbars/Browser_Toolbar.htm +++ /dev/null @@ -1,383 +0,0 @@ - - - - - - - - - - -Browser Toolbar - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Browser Toolbar

- -

- -

The Browser Toolbar provides access to the World - Wide Web and various controls for browsers.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Name

 -

Description

-

Save URL Text

 -

Save URL - Text

 -

Copies the URL name and html file to the Text - Tab. Function is enabled when a browser is selected in the Workspace.

-

New Browser

 -

New - Browser

 -

Opens a new browser window.  If - a browser is already open, new browser opens to same location as open - browser.  The - desired Web address can be typed directly into the Location Input Panel.

-

Stop

 -

Stop

 -

Stops loading a Web page.

-

Back

 -

Back

 -

Loads previous Web page (if there is one) - in the history list. Function is disabled if no Urls are in the history - list.

-

Forward

 -

Forward

 -

Loads next Web page (if there is one) in the - history list. Function is disabled if no Urls are in the history list.

- - - - - - - - -
-

 -

Location - Input Panel

 -

Input panel to open URL.

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Toolbars/Debug_Toolbar.htm b/Help/helps/VisualText_Interface/Toolbars/Debug_Toolbar.htm deleted file mode 100644 index ffc2ad6..0000000 --- a/Help/helps/VisualText_Interface/Toolbars/Debug_Toolbar.htm +++ /dev/null @@ -1,588 +0,0 @@ - - - - - - - - - - -Debug Toolbar - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Debug Toolbar

- -

- -

The Debug Toolbar provides controls to display parse - trees, intermediate files and output files.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Name

 -

Description

-

Toggle Page Mode

 -

Toggle - Page Mode

 -

Maintains coordinated sets of documents. For - example, when the user runs the analyzer, its corresponding output and - tree windows are updated automatically. Similarly when selecting passes - in the Ana Tab, Page Mode activates the Next Text and Previous Text buttons.

-

Previous Text

 -

Previous - Text

 -

Displays previous input text file according - to the ordering in the Text Tab. Works in Page Mode to automatically update - the open documents associated with the input text.

-

Next Text

 -

Next - Text

 -

Displays the next input text file according - to the ordering in the Text Tab. Works in Page Mode to automatically update - the open documents associated with the input text.

-

View Current Text

 -

View - Current Text

 -

Displays the currently selected input data - file in the Workspace.

-

View Web Page

 -

View - Web Page

 -

Displays Web page for the HTML file selected - in the Text Tab.

-

View Pass File

 -

View - Pass File

 -

Displays the currently selected pass file - in the analyzer sequence in the Workspace.

-

View Parse Tree

 -

View - Parse Tree

 -

Displays the full parse tree of a parsed document - in the Workspace. This includes character position and white spaces in - the document.

-

View Concept Tree

 -

View - Concept Tree

 -

Displays the concept tree of a parsed document - in the Workspace.

-

Toggle Node Variables

 -

Toggle - Node Variables

 -

When selected, enables the display of node - variables in the parse tree.

-

View Dump Files

 -

View - Dump Files

 -

Displays the output files associated with - a parsed document in the Workspace. When there are multiple files, the - pulldown tab to the right of the button lists available files.

-

View Output File

 -

View - Output File

 -

Displays the output.txt file associated with - parsed input document in the Workspace.

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Toolbars/Main_Toolbar.htm b/Help/helps/VisualText_Interface/Toolbars/Main_Toolbar.htm deleted file mode 100644 index c873c72..0000000 --- a/Help/helps/VisualText_Interface/Toolbars/Main_Toolbar.htm +++ /dev/null @@ -1,694 +0,0 @@ - - - - - - - - - - -Main Toolbar - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Main Toolbar

- -

- -

The Main Toolbar provides access to commonly used - functions such as opening, closing and saving files. It also contains - the controls for toggling the Main Windows.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Name

 -

Description

-

New

 -

New

 -

Creates a new file and displays it in the - Workspace window.

-

Open

 -

Open

 -

Opens a file and displays it in the Workspace.

-

Save

 -

Save

 -

Saves currently selected file in the Workspace.

-

Save All

 -

Save - All

 -

Saves all unsaved files in the Workspace.

-

Cut

 -

Cut

 -

Deletes selected object in a Workspace file.

-

Copy

 -

Copy

 -

Copies selected object to the clipboard.

-

Paste

 -

Paste

 -

Pastes object from the clipboard at insertion - point in currently selected file.

-

Print

 -

Print

 -

Prints currently selected file in the Workspace.

-

Find in Files

 -

Find - in Files

 -

Launches Find - in Files dialog box to search for items across multiple files. File - type can be specified. Search results displayed in the Find - Window.

-

Find Previous

 -

Find - Previous

 -

Searches for the last occurrence of an item - in the find dialog.

-

Find Next

 -

Find - Next

 -

Searches for the next occurrence of an item - in the find dialog.

-

Toggle Tab Window

 -

Tab - Window

 -

Toggles the visibility of the Tab Window. - By default, the Tab Window is located in the middle left portion of the - interface.

-

Toggle Log Window

 -

Log - Window

 -

Toggles the visibility of the Log Window. - By default, the Log Window is located in the bottom left portion of the - interface.

-

Toggle Find Window

 -

Find - Window

 -

Toggles the visibility of the Find Window. - By default, the Find Window is located in the bottom right portion of - the interface.

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Toolbars/Tab_Toolbar.htm b/Help/helps/VisualText_Interface/Toolbars/Tab_Toolbar.htm deleted file mode 100644 index ddd1ebb..0000000 --- a/Help/helps/VisualText_Interface/Toolbars/Tab_Toolbar.htm +++ /dev/null @@ -1,314 +0,0 @@ - - - - - - - - - - -Tab Toolbar - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Tab Toolbar

- -

- -

The Tab Toolbar presents controls for some objects in the Tab Window, - and is active when the Ana Tab, Gram Tab or Text Tab are displayed.

- - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Name

 -

Description

-

Move Tab Selection Up

 -

Move Tab - Selection Up

 -

Moves selected pass in the Ana Tab or concept - in the Gram Tab up one position.

-

Move Tab Selection Down

 -

Move Tab - Selection Down

 -

Moves selected pass in the Ana Tab or concept - in the Gram Tab down one position.

-

Delete in Tab Window

 -

Delete - in Tab Window

 -

Deletes selected pass in the Ana Tab, concept - in the Gram Tab or text file in Text Tab.

- - - - - - - - -
-

 -

File - Location Panel

 -

Displays the name and path of currently selected - file in the Text Tab.

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Toolbars/Toolbar.htm b/Help/helps/VisualText_Interface/Toolbars/Toolbar.htm deleted file mode 100644 index c11f25d..0000000 --- a/Help/helps/VisualText_Interface/Toolbars/Toolbar.htm +++ /dev/null @@ -1,166 +0,0 @@ - - - - - - - - - - -Toolbar - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Toolbars

- -

The toolbars on the VisualText interface provide quick and easy access - to the most commonly used commands. You can move the toolbars to any location - on the interface. You can also open and close any combination of them - to suit your needs.

- -

To move a toolbar, left click on the two bars on the left edge of the - toolbar and drag it to the desired location. You can also move the toolbar - by putting the pointer on the edge of the toolbar in an area that doesn't - display a button or pulldown menu.

- -

When a toolbar is selected it will appear indented.

- -

To see a description of each of these toolbars, click - on the toolbar titles below.

- -

Main Toolbar

- -

- -

 

- -

Tab Toolbar

- -

- -

 

- -

Browser Toolbar

- -

- -

 

- -

Workspace Toolbar

- -

- -

 

- -

Debug Toolbar

- -

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Toolbars/Workspace_Toolbar.htm b/Help/helps/VisualText_Interface/Toolbars/Workspace_Toolbar.htm deleted file mode 100644 index d8a9aeb..0000000 --- a/Help/helps/VisualText_Interface/Toolbars/Workspace_Toolbar.htm +++ /dev/null @@ -1,817 +0,0 @@ - - - - - - - - - - -Workspace Toolbar - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Workspace Toolbar

- -

- -

The Workspace Toolbar provides access to the analyzer, - knowledge base and controls various functions with the display of parse - tree information.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Button

 -

Name

 -

Description

-

Attribute Editor

 -

Attribute - Editor

 -

Launches the Attribute - Editor

-

KB Editor

 -

KB - Editor

 -

Launches the KB - Editor.

-

Move Concept Up

 -

Move - Concept Up

 -

Moves a selected concept in the concept hierarchy - 'up' one position.

-

Move Concept Down

 -

Move - Concept Down

 -

Moves selected concept in the concept hierarchy - down one position.

-

Save KB

 -

Save KB

-

 

 -

Makes changes to the KB permanent. Changes - made during a VisualText session are stored in memory. If KB is not saved, - all changes made are lost upon exiting

-

Compile KB

 -

Compile - KB

 -

Compiles the Knowledge Base. Action results - in the creation of a KB.DLL. Preferences to load the compiled Knowledge - Base can be set in Preferences from the File Menu. The KB.DLL library - can be used with either an interpreted or compiled VisualText-built analyzer.

-

Compile Analyzer

 -

Compile - Analyzer

 -

Turning on this toggle, followed by loading - an analyzer, results in compiling the analyzer while it is being loaded. - This results in the creation of a RUN.DLL library, which can be called - from an external program. Currently, VisualText does not load the compiled - analyzer library. The Compile Analyzer toggle is turned off after an analyzer - is loaded.

-

Generate Dirty

 -

Generate - Dirty

 -

Generates rules only for those rule concepts - in the Gram Tab that have the Dirty - attribute set. This attribute is set by the system, for example, when - new sample data is added to the concept. This is a "quick-and-dirty" - rule generation that most often works well, but may not always agree with - the Generate Rules > All option. It is used for rapid prototyping. - See Mark for - Generation in the Gram Tab Popup Menu, where the Dirty attribute can - be set.

-

Slog

 -

Slog

 -

Does a Generate Rules All, followed by a Run - on the currently selected input.

-

Run

 -

Run

 -

Runs the analyzer on the currently selected - input file. (Analyzers can not be run on files with a .log extension.)

-

Run Next

 -

Run - Next

 -

Runs the analyzer file on the next file in - the directory after the currently selected file.

-

Stop Analyzer

 -

Stop - Analyzer

 -

Stops the current analyzer.

-

Toggle Highlight Mode

 -

Toggle - Highlight Mode

 -

When selected, enables highlighting of words - or phrases in a text file matching selected pass in the Ana Tab or selected - concept in the Gram Tab. Must be enabled before an analyzer is run over - text. This tool allows you to quickly see whether and where a rule is - applying.

-

Toggle Generate Logs

 -

Toggle - Generate Logs

 -

When selected, enables generation of intermediate - logs for each pass in the analyzer sequence. Must be enabled before an - analyzer is run over text. This function gives you the ability to display - and inspect the intermediate parse trees, one for each pass, and the final - tree. It may slow the parse substantially in many-pass analyzers, since - all intermediate trees are written to disk. If the function is not enabled, - only the final tree is available for display.

-

Toggle Verbose Mode

 -

Toggle - Verbose Mode

 -

Enables the display of details from the loading - of the analyzer. Results are displayed in the Log Window.

-

Toggle Batch Mode

 -

Toggle - Batch Mode

 -

Instructs VisualText to analyze groups of - files in an efficient mode. VisualText does not save intermediate and - output files in batch mode. The text analyzer must explicitly create a - location for output that is to be retained in batch mode. The NLP++ function - called batchstart() enables a text analyzer to check when the first text - in a batch run is being analyzed.

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Toolbars/browser_toolbar.gif b/Help/helps/VisualText_Interface/Toolbars/browser_toolbar.gif deleted file mode 100644 index 267e209..0000000 Binary files a/Help/helps/VisualText_Interface/Toolbars/browser_toolbar.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Toolbars/browser_toolbar_locpanel.gif b/Help/helps/VisualText_Interface/Toolbars/browser_toolbar_locpanel.gif deleted file mode 100644 index 97e2702..0000000 Binary files a/Help/helps/VisualText_Interface/Toolbars/browser_toolbar_locpanel.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Toolbars/debug_toolbar0614.gif b/Help/helps/VisualText_Interface/Toolbars/debug_toolbar0614.gif deleted file mode 100644 index 024a5ad..0000000 Binary files a/Help/helps/VisualText_Interface/Toolbars/debug_toolbar0614.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Toolbars/mainToolbar.gif b/Help/helps/VisualText_Interface/Toolbars/mainToolbar.gif deleted file mode 100644 index b42b5ea..0000000 Binary files a/Help/helps/VisualText_Interface/Toolbars/mainToolbar.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Toolbars/tab_toolbar.gif b/Help/helps/VisualText_Interface/Toolbars/tab_toolbar.gif deleted file mode 100644 index 12acf69..0000000 Binary files a/Help/helps/VisualText_Interface/Toolbars/tab_toolbar.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Toolbars/tab_toolbar_locpanel.gif b/Help/helps/VisualText_Interface/Toolbars/tab_toolbar_locpanel.gif deleted file mode 100644 index adae7d2..0000000 Binary files a/Help/helps/VisualText_Interface/Toolbars/tab_toolbar_locpanel.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Toolbars/workspaceToolbar.gif b/Help/helps/VisualText_Interface/Toolbars/workspaceToolbar.gif deleted file mode 100644 index 89bf674..0000000 Binary files a/Help/helps/VisualText_Interface/Toolbars/workspaceToolbar.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/Attribute_Editor.htm b/Help/helps/VisualText_Interface/Tools/Attribute_Editor.htm deleted file mode 100644 index a38e0db..0000000 --- a/Help/helps/VisualText_Interface/Tools/Attribute_Editor.htm +++ /dev/null @@ -1,914 +0,0 @@ - - - - - - - - - - -Attribute Editor - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Attribute Editor

- -

Function

- -

The Attribute Editor allows the user to edit the attributes of a concept - or node in the Knowledge Base.  An - attribute consists of a key (or - slot, or less accurately, a name) and zero or more values. - A value can be a string, an integer, or another concept. Each attribute - can have multiple values, each with a distinct type.

- -

Version 1.4 of VisualText introduces enhancements to the Attribute Editor. -  Concepts - can now be viewed and edited, as values of attributes.  Navigation - of the knowledge base from within the Attribute Editor has been made more - convenient.

- -

Accessing

- -

The Attribute Editor can be launched from several places within VisualText. -   The - easiest is to use the Attribute Editor button on the Workspace - Toolbar.

- -

Selecting Attributes from the Gram - Tab Popup Menu, the Ana Tab Popup - Menu, the Phrase Tool, - and the Text Tab Popup Menu will - launch the Attribute Editor.  You - can also access the Attribute Editor from the KB - Menu.

- -

- -

Attribute Editor Description

- -

The Attribute Editor is used to select attributes and/or values for - the selected concept or node. Here is a list of the options in the Attribute - Editor.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Item

 -

Description

-

(combo box)

 -

The combo box determines the data type of - the attribute value to be added or edited.  Selecting - an existing attribute and value will automatically update the combo box - to the type of the value already present.  Three - knowledge base data types are currently supported: string (A), concept - (C), and numeric (N).

-

Add Attribute

 -

Create a new line in the display in which - you can enter the name for the attribute. You must enter the first value - for the attribute or it defaults to "New Value".

-

Add Value

 -

Create a new line which represents an additional - value for the selected attribute.

-

Delete

 -

If there is only one value, delete the value - and attribute pair. If there are multiple values, delete the value on - the selected line.

-

no attrs

 -

Move to a concept having no attributes.

-

Browse

 -

(1) If a concept-valued attribute line is - selected, checking the Browse box highlights the concept within the KB - Editor window. (2)  If - the Browse button is checked, then clicking on a value field and then - on a concept in the KB Editor selects that concept as the new value of - the attribute.  This - enables convenient selection of a concept as the value of an attribute.

-

Quick - Edit >>

 -

Expand the Attribute Editor to include the - Quick Edit Table.  Commonly - used attribute-value pairs can be listed in the Quick Edit Table to make - adding and/or deleting attributes and attribute-value pairs easier.  When - Quick Edit Table is open, selecting button closes the Quick Edit Table - display.  (See - below.)

-

 -

Select previous or next concept.  

-

(up down arrows)

 -

Select the parent or child concept.

-

 -

Launch the KB Editor if it is not already - open.

-

 -

Search an online dictionary for the concept - currently being viewed in the Attribute Editor.  The - panel next to the dictionary icon can be used to specify which online - dictionary to access.  (For - more information on setting dictionary access preferences, see the Dictionary Lookup - preferences tab which can be accessed from the File Menu under Preferences.)

-

 -

Search input panel.  Searches - for word in the dict hierarchy.  If - item exists in the dict hierarchy, the attributes for word are displayed. -  The search - input panel locates  words - in the concept:sys:dict hierarchy only.  

-

Add

 -

Add a word concept to the dict hierarchy.

-

 -

Delete a word concept from the dict hierarchy.

-

 -

Search button.  Start - search for item entered in input panel.

-

Done

 -

Close the Attribute Editor.

- -

An attribute may have multiple values. For example, the attribute named - "algo" may have values "pat", "hello there", - and "170224929." By selecting a line and clicking Add Value, - successive values can be added. One can even mix the types of values (numeric, - "N", and string, "A") arbitrarily.

- -

The Attribute Editor is now smart enough to be able to treat "123" - as both a string or numeric type, depending on the type chosen in the - combo box.  Similarly, - "concept:gram" may represent a concept in the KB or a string, - depending on the type chosen in the combo box.

- -

Concept naming in the attribute editor looks like concept:child:grandchild:greatgrandchild - and so on.  The - concept named "concept" specifies the root of the KB hierarchy, - and colons (':') separate each successive child.

- - - - - - -
-

Notes: -  (1) The - "name" of an attribute is actually a concept in the KB, not - merely a string.  (2) - There is currently no way to specify a colon as part of the name of a - concept, in the attribute editor.

- - - - - - -
-

WARNING: - Editing the gram or sys - hierarchies manually can easily corrupt the KB, so please make a backup - copy of your analyzer project before experimenting.  The - Gram Tab is the safest way to edit the Gram hierarchy.

- -

Quick Edit Table

- -

The Quick Edit Table is used to store commonly used attribute-value - pairs.  Attributes - can be added to the selected concept by double-clicking the attribute-value - line in the Quick Edit Table or by selecting the line in the table and - clicking on the << KB Edit - button.  Add - in the Action column indicates the attribute-value pair will be added - to the concept.  Once - a concept is added, the Action column changes to Del (for delete).  Double-clicking - on a Del line will delete the attribute-value pair from the concept.

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Item

 -

Description

-

Attribute

 -

Input box to add attribute to Quick Edit Table. -  Added attribute - appears in the Attribute column.

-

Value

 -

Input box to add value for given attribute - to Quick Edit Table.  Added - value appears in the Value column.

-

Add

 -

Adds attribute-value pair to the Quick Edit - Table.

-

Remove

 -

Removes selected attribute-value pair from - the Quick Edit Table.  To - remove multiple attribute-value pairs at the same time, hold down the - shift key and select the attribute-value lines in the Quick Edit Table - you want to delete.

-

Update

 -

Updates changes made to attribute-value pair - in the Quick Edit Table.

-

Help

 -

Launches instructions on using the Quick Edit - Table.

-

<< KB Edit

 -

Assigns selected attribute-value pair to the selected concept. To assign - multiple attribute-value pairs at the same time, hold down the shift key - and select the appropriate attribute-value lines in the Quick Edit Table.

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Tools/EmailTool.gif b/Help/helps/VisualText_Interface/Tools/EmailTool.gif deleted file mode 100644 index 13b124a..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/EmailTool.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/HexViewer.gif b/Help/helps/VisualText_Interface/Tools/HexViewer.gif deleted file mode 100644 index 469eb8c..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/HexViewer.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/KB_Editor.htm b/Help/helps/VisualText_Interface/Tools/KB_Editor.htm deleted file mode 100644 index 8d146e5..0000000 --- a/Help/helps/VisualText_Interface/Tools/KB_Editor.htm +++ /dev/null @@ -1,946 +0,0 @@ - - - - - - - - - - -KB Editor - - - - - - - - - - - - - - - - - - - - - - - - - - - -

KB Editor

- -

Function

- -

The KB Editor displays the Knowledge - Base and allows you to manipulate portions of it.

- -

Accessing

- -

The KB Editor can be accessed by selecting the KB - Editor button on the Workspace Toolbar, or by selecting - KB Editor from the KB Menu.  You - can also access the KB Editor from the Attribute - Editor.

- -

- -

Concept Hierarchy Display

- -

The KB Editor displays a hierarchy of concepts. The major organization - of this Conceptual Grammar Knowledge Base Management System is a hierarchy - of concepts.

- -

concept is the root of the concept - hierarchy. It is the topmost concept in the knowledge base.  The - child concept gram is where the - Sample Hierarchy (or Gram Hierarchy) is implemented. Other VisualText - and KB internals are stored in sys. - The dictionary hierarchy (dict), - the analyzer sequence and a mirror for input files are stored under sys.

- -

Concept Icons

- -

Concept icons in the KB Editor are color-coded.  You - can tell several things about a concept just by looking at the color of - the icon.

- -

- - - - - - - - - - - - - - - - - - - - - - - -
-

Color

 -

Description

-

Blue

 -

Concept in selected.

-

Yellow

 -

Concept has an associated phrase.

-

Grey

 -

Concept has attributes

-

Clear (hollow)

 -

Concept has no attributes.

- -

 

- -

Using the KB Editor

- -

The KB Editor is used to modify current concepts and to add new concepts - to the concept hierarchy when building representation schemes. The existing - hierarchies may be modified, but it is not recommended to edit system - attributes of concepts. Also, note that the gram - and sys subhierarchies - should not be hand-edited using the KB Editor.   

- -

The Attribute Editor can be used to manually add or modify attributes. -  However, - whenever feasible users should edit concepts using the Gram - Concept Properties dialog. The Gram Concept Properties dialog is accessed - by selecting Properties from the Gram - Tab Popup Menu.

- -

KB Editor Popup Menu

- -

To edit using the KB Editor, select a node and right click to bring - up the KB Editor Popup Menu:

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Item

 -

Description

-

Attributes

 -

Launches the Attribute - Editor dialog and displays the attributes belonging to the selected - concept in the KB. User has option to delete, add or change names of attributes - and values.

-

Display as - Root

 -

Displays selected concept as the root of the - concept hierarchy.

-

Display Entire - KB

 -

Displays the entire knowledge base starting - at root.

-

Display Parent - as Root

 -

Displays parent of concept currently displayed - as root.

-

Phrase - Tool

 -

Launches the Phrase Tool. Displays the phrase, if any, of a selected - concept in the Knowledge Base. User can add edit or delete nodes of a - concept's phrase. (See below.)

-

Refresh

 -

Refreshes the KB Editor display.

-

Add Concept(s)

 -

Adds a new concept under the selected concept.

-

Add Node

 -

Adds a new node to the concept's phrase.

-

Delete

 -

Deletes the selected concept and its subtree.

-

Delete Children

 -

Deletes the subtree of a selected concept, - but retains the concept itself.

-

Modular

 -

Submenu for saving and loading modular KB - files.  (See - below.)   

-

Save

 -

Saves the KB.

- - - - - - -
-

Note: - To prevent users from inadvertently corrupting the knowledge base, analyzers - are loaded in KB Safe Edit Mode. -  With KB - Safe Edit Mode enabled, Add Concept, - Add Node, Delete, - Delete Children options on the - KB Editor Popup Menu are not available when either the gram concept or - the sys concept is selected.  You - may disable KB Safe Edit Mode by unchecking the KB - Safe Edit Mode box on the General Preferences tab.  Users - are advised to create backups before using the KB Editor when KB Safe - Edit Mode is turned off.

- -

 

- -

Phrase Tool

- -

In the Conceptual Grammar knowledge base, each concept may have one - associated phrase. A phrase consists - of nodes, which are basically - concepts that are not attached directly to the KB hierarchy.

- -

A phrase is a convenient way to embed sequential - information of any type in the knowledge base. For example, if you browse - the concept > gram subhierarchy - of the KB, you will find that the sample hierarchy stores user-highlighted - samples as phrases (each sample is actually a node - of the phrase belonging to a rule concept).

- -

The Phrase Tool lets you view and edit the phrase - that belongs to the currently selected concept. Each line in the Phrase - Tool represents one node. Only the name of the node is visible, but double-clicking - a node or selecting the Attributes button brings up the Attribute Editor - for the current node.

- -

- -

Phrase Tool Description

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Item

 -

Description

-

Delete

 -

Deletes the selected node.

-

Attributes

 -

Launches the Attribute Editor for the selected - node.

-

Add

 -

Launches the Add Node dialog to add a node - to the current phrase. (See below.)

-

 -

Moves the selected node up one position in - the phrase.

-

 -

Moves the selected node down one position - in the phrase.

-

Done

 -

Saves changes to the phrase and closes the - Phrase dialog box.

-

Cancel

 -

Cancels the changes to the phrase.

- - - - - - -
-

Note: - If the analyzer has been loaded in KB - Edit Safe Mode, the only - option available in the Phrase Tool when viewing gram or sys concepts - is the Attributes button.    

- -

 

- -

Add Node Dialog

- -

The Add Node - dialog appears when adding a node to a concept's phrase, either from the - Phrase Tool or the KB Editor Popup Menu > Add Node menu item. To add - a node, you merely specify a name for it. You may edit an existing node's - attributes with the Attribute Editor.

- -

- -

Modular Submenu

- - - - - - - - - - - - - - - -
-

Item

 -

Description

-

Load

 -

 

-

Save

 -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Tools/Mask.gif b/Help/helps/VisualText_Interface/Tools/Mask.gif deleted file mode 100644 index 9d7bd8c..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/Mask.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/TextFileInfo.gif b/Help/helps/VisualText_Interface/Tools/TextFileInfo.gif deleted file mode 100644 index eac0662..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/TextFileInfo.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/Tools_ASCII_Table.htm b/Help/helps/VisualText_Interface/Tools/Tools_ASCII_Table.htm deleted file mode 100644 index 4703c11..0000000 --- a/Help/helps/VisualText_Interface/Tools/Tools_ASCII_Table.htm +++ /dev/null @@ -1,78 +0,0 @@ - - - - - - - - - - -ASCII Table - - - - - - - - - - - - - - - - - - - - - - - - - - - -

ASCII Table

- -

Function

- -

The ASCII Table is used to display the decimal and hexadecimal numbers - and their corresponding ASCII characters.

- -

Accessing

- -

The ASCII Table can be displayed from several places within VisualText. -  It can - be accessed from the main Tools Menu, - the Text Tab Popup Menu under - Tools, and from the Tools submenu in the Text - File Popup Menu.

- -

ASCII Table Display

- -

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Tools/Tools_Character_Viewer.htm b/Help/helps/VisualText_Interface/Tools/Tools_Character_Viewer.htm deleted file mode 100644 index 2a762bf..0000000 --- a/Help/helps/VisualText_Interface/Tools/Tools_Character_Viewer.htm +++ /dev/null @@ -1,88 +0,0 @@ - - - - - - - - - - -Character Viewer - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Character Viewer

- -

Function

- -

The Character Viewer Tool is used to get information on lines of text - and the characters used in each line of text.  

- -

Accessing

- -

The Character Viewer Tool can be displayed from several places within - VisualText.  It - can be accessed from the main Tools Menu, - the Text Tab Popup Menu under - Tools, and from the Tools submenu in the Text - File Popup Menu.

- -

The Character Viewer Tool can be launched for an open file in the Workspace. -  If no file - is open in the Workspace, an Open dialog is launched to navigate to the - desired text file.

- -

Character Viewer Tool Display

- -

The Character Viewer Tool shows the line number, a cumulative count - of total characters seen per line, the characters on a line, and the ASCII - characters for each line of text in selected file. Under each line is - the hexadecimal value for each character above it.  

- -

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Tools/Tools_DictionaryLookup.htm b/Help/helps/VisualText_Interface/Tools/Tools_DictionaryLookup.htm deleted file mode 100644 index 12a2855..0000000 --- a/Help/helps/VisualText_Interface/Tools/Tools_DictionaryLookup.htm +++ /dev/null @@ -1,151 +0,0 @@ - - - - - - - - - - -Dictionary Lookup - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Dictionary Lookup

- -

Function

- -

The Dictionary Lookup Tool allows you to access and search specified - online dictionaries.  It - can be a handy tool when doing lexical work.  You - can feed Dictionary Lookup a list of words or phrases and it will return - html files containing the results for the search for each of the words - on the list.  The - html files are placed in folders in the Text Tab.  You - can also use Dictionary Lookup as a reference source to search for words - in the online dictionary directly from VisualText.  

- -

Accessing

- -

Dictionary Lookup can be accessed from several places within VisualText. -  It can - be accessed from the main Tools Menu, - the Text Tab Popup Menu under - Tools, and from the Tools menu in the Text - File Popup Menu. Dictionaries can also - be accessed from the Attribute Editor. -

- -

Setting Dictionary Lookup Preferences

- -

Access to the online ontology WordNet is provided by default.  To - add other online dictionaries to Dictionary Lookup, you must specify the - dictionary path in the Dictionary - Lookup tab in VisualText preferences located under the File Menu. -  For more - information on creating path names to online dictionary sources, see URLs for Dictionary - Lookup.

- - - - - - -
-

Note: - Access to WordNet relies on a fixed webpage CGI to WordNet. If the WordNet - page changes, the path for WordNet must be changed in the preferences - dialog.

- -

 

- -

Using Dictionary Lookup

- -

To create dictionary files from a word - list:

- -

Dictionary Lookup expects a file that consists of a list of words, one - per line with no spaces after the word.

- -
    - -

  1. Select the file - in the Text Tab.

    2. - -

  2. Right-click and select Dictionary - > Lookup from the Text Tab - Popup, and choose the online dictionary from the list.  Html - files for each of the words on the list are created in a folder in the - input directory.  The - name of the folder is filename_DictionaryName.  

    3. -
- -

To use Dictionary Lookup to find words - in an online dictionary:

- -
    - -

  1. Highlight a word - in an open text file in the Workspace.  (If - you do not highlight a word, you will be prompted with a dialog box.)

    2. - -

  2. Right-click and select Dictionary - > Lookup from the Text File - Popup, and choose the online dictionary from the list.  The - word is displayed in a browser window.  

    3. -
- -

See Also

- -

Dictionary - Lookup Preferences

- -

URLs - for Dictionary Lookup

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Tools/Tools_Dictionize.htm b/Help/helps/VisualText_Interface/Tools/Tools_Dictionize.htm deleted file mode 100644 index 935fa92..0000000 --- a/Help/helps/VisualText_Interface/Tools/Tools_Dictionize.htm +++ /dev/null @@ -1,92 +0,0 @@ - - - - - - - - - - -Dictionize - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Dictionize Tool

- -

Function

- -

The Dictionize Tool breaks up a file of alphabetized words or phrases - into a set of word files.  The - Dictionize Tool is a batch tool that can be used when adding new lexical - items to the knowledge base.  

- -

Accessing

- -

The Dictionize Tool can be accessed from several places within VisualText. -  It can - be accessed from the Text Tab Popup - Menu under Tools, and from the Tools submenu in the Text - File Popup Menu.

- -

Using the Dictionize Tool

- -

The Dictionize Tool expects an alphabetized list of words or phrases. -  There should - only be one word or phrase per line.  

- -

To use the Dictionize Tool:

- -
    - -

  1. Select the file containing the words or phrases - in the Text Tab.

    2. - -

  2. Right-click and select - Tools > Dictionize. -  A set of - folders and files corresponding to the indexes in the KB dictionary hierarchy - are created in the same folder as the text file.  (You - can also open the file in the Workspace and select Tools > Dictionize - from the right-click menu.)

    3. -
- -

 

- -

 

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Tools/Tools_Email_Tool.htm b/Help/helps/VisualText_Interface/Tools/Tools_Email_Tool.htm deleted file mode 100644 index f814625..0000000 --- a/Help/helps/VisualText_Interface/Tools/Tools_Email_Tool.htm +++ /dev/null @@ -1,275 +0,0 @@ - - - - - - - - - - -Email Tool - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Email Tool

- -

Function

- -

The Email Tool enables you to send email directly from the VisualText - interface.

- -

Accessing

- -

The Email Tool can be accessed from several places within VisualText. -  It can - be accessed from the main Tools Menu, - the Text Tab Popup Menu under - Tools, and from the Tools submenu in the Text - File Popup Menu.

- -

Setting Email Tool Preferences

- -

Preferences for the Email Tool can be set under the Email Tab in VisualText - preferences under the File Menu.  SMTP - Server and From Email must be set before the Email Tool can be used.

- -

Email Tool Display

- -

When the Email Tool is selected, the content of the selected file is - automatically filled into the body of the message.

- -

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Menu Item

 -

Description

-

To:

 -

Field to enter email recipient. If text file - contains "to: person@emailaddress" field is automatically entered - with information from file.

-

Subject:

 -

Field to enter email subject. If text file - contains "subject: subject of email" field is automatically - entered with information from file.

-

Email List - File:

 -

Allows user to browse to a folder of email - files to be sent.

-

<<

 -

Allows user to move to previous file in the - email folder.

-

>>

 -

Allows user to move to the next file in the - email folder.

-

Send Current

 -

Sends the current email.

-

Send All

 -

Emails all the files in the current folder.

-

Done

 -

Closes the Email Tool display.

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Tools/Tools_Hex_Viewer.htm b/Help/helps/VisualText_Interface/Tools/Tools_Hex_Viewer.htm deleted file mode 100644 index cc6501e..0000000 --- a/Help/helps/VisualText_Interface/Tools/Tools_Hex_Viewer.htm +++ /dev/null @@ -1,74 +0,0 @@ - - - - - - - - - - -Hex Viewer - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Hex Viewer

- -

Function

- -

The Hex Viewer provides various views of the characters in a text file. -

- -

Accessing

- -

The Hex Viewer can be accessed from several places within VisualText. -  It can - be accessed from the main Tools Menu, - the Text Tab Popup Menu under Tools, - and from the Tools submenu in the Text - File Popup Menu.

- -

Hex Viewer Display

- -

The ASCII Table button displays the table of ASCII codes.  You - can also see the address, hexcodes, equivalent ASCII characters, and full - 8-digit addresses for each character in the text. The slider determines - the number of characters displayed per line.

- -

- - - - diff --git a/Help/helps/VisualText_Interface/Tools/Tools_Jumble.htm b/Help/helps/VisualText_Interface/Tools/Tools_Jumble.htm deleted file mode 100644 index 9a9e3f7..0000000 --- a/Help/helps/VisualText_Interface/Tools/Tools_Jumble.htm +++ /dev/null @@ -1,95 +0,0 @@ - - - - - - - - - - -Jumble Tool - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Jumble Tool

- -

Function

- -

Jumble is a simple non-graphic tool that converts a text file to a version - that facilitates studying the format of the file. Jumble randomly changes - the order of letters in the text file leaving punctuation and capitalization - intact.

- -

Accessing

- -

The Jumble Tool can be accessed from several places within VisualText. -  It can - be accessed from the Text Tab Popup - Menu under Tools, and from the Tools submenu in the Text - File Popup Menu.

- -

Using the Jumble Tool

- -

To use the Jumble Tool:

- -
    - -

  1. Select the text - in the Text Tab.

    2. - -

  2. From the right-click Menu, select Tools - > Jumble.  A - .jmb (for jumble) file for the selected text is created in the same folder - as the text.  (You - can also open the text file in the Workspace and select Tools > Jumble - from the right-click menu.)

    3. -
- -

Jumble Example

- -

- - - - diff --git a/Help/helps/VisualText_Interface/Tools/Tools_Mask_Tool.htm b/Help/helps/VisualText_Interface/Tools/Tools_Mask_Tool.htm deleted file mode 100644 index 3959066..0000000 --- a/Help/helps/VisualText_Interface/Tools/Tools_Mask_Tool.htm +++ /dev/null @@ -1,95 +0,0 @@ - - - - - - - - - - -Mask Tool - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Mask Tool

- -

Function

- -

Mask is a simple non-graphic tool that converts a text file to a version - that facilitates studying the format of the file. All alphabetic characters - are converted to "a" or "A", retaining capitalization - of the input text.

- -

Accessing

- -

The Mask Tool can be accessed from several places within VisualText. -  It can - be accessed from the Text Tab Popup - Menu under Tools, and from the Tools submenu in the Text - File Popup Menu.

- -

Using the Mask Tool

- -

To use the Mask Tool:

- -
    - -

  1. Select the text - in the Text Tab.

    2. - -

  2. From the right-click Menu, select Tools - > Mask.  A - .msk (for mask) file for the selected text is created in the same folder - as the text.  (You - can also open the text file in the Workspace and select Tools > Mask - from the right-click menu.)

    3. -
- -

Mask Example

- -

- - - - diff --git a/Help/helps/VisualText_Interface/Tools/Tools_Text_File_Info.htm b/Help/helps/VisualText_Interface/Tools/Tools_Text_File_Info.htm deleted file mode 100644 index 8ae9711..0000000 --- a/Help/helps/VisualText_Interface/Tools/Tools_Text_File_Info.htm +++ /dev/null @@ -1,72 +0,0 @@ - - - - - - - - - - -Text File Info - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Text File Info

- -

Function

- -

The Text File Info dialog displays information about a text file.

- -

Accessing

- -

The Text File Info dialog can be displayed from several places within - VisualText.  It - can be accessed from the main Tools Menu, - the Text Tab Popup Menu under Tools, - and from the Tools submenu in the Text - File Popup Menu.

- -

Text File Info Display

- -

The Text File Info dialog provides the location of the selected file, - creation and modification dates, line count, word count, character count, - and file size.

- -

- - - - diff --git a/Help/helps/VisualText_Interface/Tools/asciiTable.gif b/Help/helps/VisualText_Interface/Tools/asciiTable.gif deleted file mode 100644 index 7289a35..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/asciiTable.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/attr_editor_020619.gif b/Help/helps/VisualText_Interface/Tools/attr_editor_020619.gif deleted file mode 100644 index f4afa02..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/attr_editor_020619.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/attrib_edit0724_table.gif b/Help/helps/VisualText_Interface/Tools/attrib_edit0724_table.gif deleted file mode 100644 index 828e64a..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/attrib_edit0724_table.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/charViewer.gif b/Help/helps/VisualText_Interface/Tools/charViewer.gif deleted file mode 100644 index 559f62b..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/charViewer.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/concept_phrase_tool_combo.gif b/Help/helps/VisualText_Interface/Tools/concept_phrase_tool_combo.gif deleted file mode 100644 index 66697ed..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/concept_phrase_tool_combo.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/dict_w_panel.gif b/Help/helps/VisualText_Interface/Tools/dict_w_panel.gif deleted file mode 100644 index 91d3318..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/dict_w_panel.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/jumpleexam.gif b/Help/helps/VisualText_Interface/Tools/jumpleexam.gif deleted file mode 100644 index b3057ac..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/jumpleexam.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/kb_editor0814.gif b/Help/helps/VisualText_Interface/Tools/kb_editor0814.gif deleted file mode 100644 index 95bbd0a..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/kb_editor0814.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/kb_editorIcons.gif b/Help/helps/VisualText_Interface/Tools/kb_editorIcons.gif deleted file mode 100644 index ce1bb14..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/kb_editorIcons.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/kb_editor_gramtab.gif b/Help/helps/VisualText_Interface/Tools/kb_editor_gramtab.gif deleted file mode 100644 index b776c12..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/kb_editor_gramtab.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/phrase_tool_add_node.gif b/Help/helps/VisualText_Interface/Tools/phrase_tool_add_node.gif deleted file mode 100644 index 0e79075..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/phrase_tool_add_node.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/phrase_tool_attribute_combo.gif b/Help/helps/VisualText_Interface/Tools/phrase_tool_attribute_combo.gif deleted file mode 100644 index 558310b..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/phrase_tool_attribute_combo.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Tools/phrase_tool_dialog.gif b/Help/helps/VisualText_Interface/Tools/phrase_tool_dialog.gif deleted file mode 100644 index c837056..0000000 Binary files a/Help/helps/VisualText_Interface/Tools/phrase_tool_dialog.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Windows/Find_in_Files_Dialog.htm b/Help/helps/VisualText_Interface/Windows/Find_in_Files_Dialog.htm deleted file mode 100644 index fd32b04..0000000 --- a/Help/helps/VisualText_Interface/Windows/Find_in_Files_Dialog.htm +++ /dev/null @@ -1,361 +0,0 @@ - - - - - - - - - - -Find in Files Dialog - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Find in Files Dialog

- -

The Find in Files dialog allows you to - find text in multiple files.

- -

Accessing

- -

The Find in Files dialog is - launched by selecting Find in Files from the Edit Menu or by selecting - the Find in Files icon.  

- -

You can constrain the search by selecting specific file types or by - selecting another analyzer project.  Search - results are displayed in the Find Window.

- -

- -

 

- - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Dialog Item

 -

Description

-

Find

 -

Input panel to type text to be searched.

-

Match Case

 -

Searches text matching exact case of search term.

-

Append

 -

Appends search results to the Find Window.  When - Append is not checked, Find Window results are cleared before results - from new search are displayed.

-

Pass Files

 -

Searches for search term in pass files.

-

All Text Files

 -

Searches for search term in all text files.

-

Concepts

 -

Searches for search term in Gram Tab concepts.

-

Output Files

 -

Searches for search term in output files.

-

Log Files

 -

Searches for search term in log files.

-

Other Analyzer Pass Files

 -

Shows other analyzer projects.  Selecting - analyzer project enables search in pass files of specified project.

-

OK

 -

Starts the search.

-

Cancel

 -

Closes the Find in Finds dialog box.

- -

 

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Windows/Replace_in_Files_Dialog.htm b/Help/helps/VisualText_Interface/Windows/Replace_in_Files_Dialog.htm deleted file mode 100644 index 9620f3d..0000000 --- a/Help/helps/VisualText_Interface/Windows/Replace_in_Files_Dialog.htm +++ /dev/null @@ -1,391 +0,0 @@ - - - - - - - - - - -Replace in Files Dialog - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Replace in Files Dialog

- -

The Replace in Files dialog allows you to - replace text in multiple files.

- -

The Replace in Files dialog - is launched by selecting Replace in Files - from the Edit Menu.  You - may specify the files where text should be replaced.   Search - and replace results are displayed in the Find - Window.

- -

- - - --- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

Dialog Item

 -

Description

-

Find

 -

Input panel to type search term.

-

Replace With

 -

Input panel to specify what found text should be replaced with.

-

Replace

 -

Starts the search and replace function.  Results - of Replace displayed in Find Window.

-

Cancel

 -

Closes the Replace in Files dialog box.

-

Match Whole Word Only

 -

Searches text matching whole words only.

-

Match Case

 -

Searches text matching exact case of search term.

-

All Pass Files

 -

Searches for search term in all pass files.

-

All Text Files

 -

Searches for search term in all text files.

-

Replace in Pass Subfolders

 -

Replaces instances of found search term in all pass file subfolders.

-

Replace in Text Subfolders

 -

Replaces instances of found search term in all text file subfolders.

-

Directory

 -

Input panel to specify directory to perform search.

-

File Spec

 -

Specifies files to search by file extension.

-

Replace in Subfolders

 -

Replaces instances of found search term in all subfolders.

- -

 

- - - - diff --git a/Help/helps/VisualText_Interface/Windows/ana_tab.gif b/Help/helps/VisualText_Interface/Windows/ana_tab.gif deleted file mode 100644 index 9932bcd..0000000 Binary files a/Help/helps/VisualText_Interface/Windows/ana_tab.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Windows/findInFilesdialog.gif b/Help/helps/VisualText_Interface/Windows/findInFilesdialog.gif deleted file mode 100644 index cb8137c..0000000 Binary files a/Help/helps/VisualText_Interface/Windows/findInFilesdialog.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Windows/find_in_files_wtitles.gif b/Help/helps/VisualText_Interface/Windows/find_in_files_wtitles.gif deleted file mode 100644 index ebd8453..0000000 Binary files a/Help/helps/VisualText_Interface/Windows/find_in_files_wtitles.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Windows/gram_tab.gif b/Help/helps/VisualText_Interface/Windows/gram_tab.gif deleted file mode 100644 index 3c9dde1..0000000 Binary files a/Help/helps/VisualText_Interface/Windows/gram_tab.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Windows/log_window_wtitles.gif b/Help/helps/VisualText_Interface/Windows/log_window_wtitles.gif deleted file mode 100644 index 433ab56..0000000 Binary files a/Help/helps/VisualText_Interface/Windows/log_window_wtitles.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Windows/parse_yellow.gif b/Help/helps/VisualText_Interface/Windows/parse_yellow.gif deleted file mode 100644 index c73ef2d..0000000 Binary files a/Help/helps/VisualText_Interface/Windows/parse_yellow.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Windows/replaceInFilesDialog.gif b/Help/helps/VisualText_Interface/Windows/replaceInFilesDialog.gif deleted file mode 100644 index 2b19293..0000000 Binary files a/Help/helps/VisualText_Interface/Windows/replaceInFilesDialog.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/Windows/text_tab.gif b/Help/helps/VisualText_Interface/Windows/text_tab.gif deleted file mode 100644 index bd5ffd1..0000000 Binary files a/Help/helps/VisualText_Interface/Windows/text_tab.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/anaDependent_pref.gif b/Help/helps/VisualText_Interface/anaDependent_pref.gif deleted file mode 100644 index f6c7620..0000000 Binary files a/Help/helps/VisualText_Interface/anaDependent_pref.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/dictionaryLookup_pref.gif b/Help/helps/VisualText_Interface/dictionaryLookup_pref.gif deleted file mode 100644 index 8244375..0000000 Binary files a/Help/helps/VisualText_Interface/dictionaryLookup_pref.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/viewIcons.gif b/Help/helps/VisualText_Interface/viewIcons.gif deleted file mode 100644 index 807a317..0000000 Binary files a/Help/helps/VisualText_Interface/viewIcons.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/vt_preferences_display.gif b/Help/helps/VisualText_Interface/vt_preferences_display.gif deleted file mode 100644 index 04e3b19..0000000 Binary files a/Help/helps/VisualText_Interface/vt_preferences_display.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/vt_preferences_gen.gif b/Help/helps/VisualText_Interface/vt_preferences_gen.gif deleted file mode 100644 index 061f170..0000000 Binary files a/Help/helps/VisualText_Interface/vt_preferences_gen.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/vt_prefereneces_email.gif b/Help/helps/VisualText_Interface/vt_prefereneces_email.gif deleted file mode 100644 index 43deeec..0000000 Binary files a/Help/helps/VisualText_Interface/vt_prefereneces_email.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/vt_preferneces_archiving.gif b/Help/helps/VisualText_Interface/vt_preferneces_archiving.gif deleted file mode 100644 index 3cbe8c5..0000000 Binary files a/Help/helps/VisualText_Interface/vt_preferneces_archiving.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/windowToggles.gif b/Help/helps/VisualText_Interface/windowToggles.gif deleted file mode 100644 index 7948e8b..0000000 Binary files a/Help/helps/VisualText_Interface/windowToggles.gif and /dev/null differ diff --git a/Help/helps/VisualText_Interface/window_pulldown.gif b/Help/helps/VisualText_Interface/window_pulldown.gif deleted file mode 100644 index 58ee293..0000000 Binary files a/Help/helps/VisualText_Interface/window_pulldown.gif and /dev/null differ diff --git a/Help/helps/Working_with_Unicode.htm b/Help/helps/Working_with_Unicode.htm deleted file mode 100644 index 947d4fb..0000000 --- a/Help/helps/Working_with_Unicode.htm +++ /dev/null @@ -1,133 +0,0 @@ - - - - - - - - - - -Working_with_Unicode - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Working with Unicode

- -

Starting with VisualText 2.0.1, a Unicode VisualText executable is available - with every release.

- -

Unicode is well-documented on the web.  Basically, - Unicode is a character set that supports the major languages and character - sets.  To - do this, characters consist of two bytes, rather than the ANSI or ASCII - representation of characters with one byte.

- -

While the advantages of Unicode are tremendous, it also adds substantial - complexity and headaches to programs that use it.  Also, - Unicode is necessarily processed more slowly than single-byte character - representations.

- -

VisualText support for Unicode is an ongoing enterprise.  The - 2.0.1 release provides default tokenization for the world's character - sets (or "scripts"), as well as supporting Unicode characters - in NLP++ strings, in the GUI, in knowledge bases, compiled analyzers, - and the like.

- -

In subsequent releases, we plan to add NLP++ functions that will help - characterize and manipulate Unicode constructs in text, parse trees, strings, - and the knowledge base.

- -

Windows and Unicode

- -

To display and input Unicode fonts, you'll need to consult your Microsoft - Windows documentation.  For - example, in Windows XP, select

- -

Start Button > Settings > Control Panel > Regional and Language - Settings

- -

The Languages tab there lets you install fonts for complex and East - Asian character sets.  The - Details button within the Languages tab lets you set up input in various - character sets, and so on.

- -

NOTES

- -

[1] If using English only or small Western alphabets (e.g., French, - Spanish, German), consider staying with ANSI VisualText.

- -

[2] An analyzer should be used with Unicode or ANSI VisualText, but - not both.

- -

[3] Unicode VisualText writes files out using the UTF-8 format and expects - to read files in the same format.  If - an input file contains Unicode characters, then VisualText expects it - to have the Byte Order Marker (BOM) as required of Windows files.  A - nice thing about UTF-8 is that normal ANSI or ASCII files are automatically - in the right format, with or without a BOM header.

- -

If a file does not appear to be read correctly by VisualText, - then it may have come from a non-Windows source.  Try - opening the file in an application such as WordPad, add a character, then - remove the character, then save the file as UTF-8.  This - should automatically add the BOM head to the beginning of the file.

- -

[4] KB dictionary indexing.  We'll - be providing analyzer templates specific for Unicode.  You - can also manually add dictionary indexes to the knowledge base (by editing - the kb\user\hier.kb file).  Contact - us at support@textanalysis.com for help with these issues.

- -

add hier "concept" "sys" "dict" "a" - "z" "zy"
-add hier "concept" "sys" "dict" "a" - "あ"
-add hier "concept" "sys" "dict" "a" - "あ" "あ"

- -

Basically, you'll want to add lines like the second and third above, - in order to have Unicode characters distributed well within the KB dictionary.

- -

[5] No particular updates have been made yet to RUG (Automated Rule - Generation) for Unicode VisualText.

- -

 

- - - - diff --git a/Help/helps/X.htm b/Help/helps/X.htm deleted file mode 100644 index 406745b..0000000 --- a/Help/helps/X.htm +++ /dev/null @@ -1,46 +0,0 @@ - - - - - - - - - - -X - - - - - - - - - - - - - - - - - - - - - - - - - - - -

X

- -

Reference to a parse tree node or its variable.  See - Variables.

- - - - diff --git a/Help/helps/_xALPHA.htm b/Help/helps/_xALPHA.htm deleted file mode 100644 index 16f8f29..0000000 --- a/Help/helps/_xALPHA.htm +++ /dev/null @@ -1,57 +0,0 @@ - - - - - - - - - - -_xALPHA - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xALPHA

- -

Purpose

- -

Match an alphabetic token, including accented and other extended ANSI - chars.

- -

Remarks

- -

Example

- -

See Also

- -

_xCAP, _xNUM, _xCTRL, - Special Rule Elements

- - - - diff --git a/Help/helps/_xANY.htm b/Help/helps/_xANY.htm deleted file mode 100644 index cb34b65..0000000 --- a/Help/helps/_xANY.htm +++ /dev/null @@ -1,62 +0,0 @@ - - - - - - - - - - -_xANY - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xANY

- -

Purpose

- -

Match any single node.

- -

Remarks

- -

A rule written with _xANY [max=0] will not work.  _xANY - is not implemented as a wildcard.  Instead, - the near equivalent _xANY _xWILD should be used.

- -

Example

- -

See Also

- -

min, - max, Special Rule Elements, - Phrase Element - Modifiers

- - - - diff --git a/Help/helps/_xBLANK.htm b/Help/helps/_xBLANK.htm deleted file mode 100644 index b7dd4c5..0000000 --- a/Help/helps/_xBLANK.htm +++ /dev/null @@ -1,58 +0,0 @@ - - - - - - - - - - -_xBLANK - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xBLANK

- -

Purpose

- -

Match a whitespace token, excluding newline.

- -

Remarks

- -

_xBLANK is equivalent to _xWILD [match=(\ \t)].

- -

Example

- -

See Also

- -

_xWHITE, Special - Rule Elements

- - - - diff --git a/Help/helps/_xCAP.htm b/Help/helps/_xCAP.htm deleted file mode 100644 index 399ddb0..0000000 --- a/Help/helps/_xCAP.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -_xCAP - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xCAP

- -

Purpose

- -

Match an alphabetic with uppercase first letter.

- -

Remarks

- -

Example

- -

See Also

- -

_xALPHA, Special - Rule Elements

- - - - diff --git a/Help/helps/_xCAPLET.htm b/Help/helps/_xCAPLET.htm deleted file mode 100644 index c388daf..0000000 --- a/Help/helps/_xCAPLET.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -_xCAPLET - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xCAPLET

- -

Purpose

- -

Match an alphabetic consisting of a single uppercase letter.

- -

Remarks

- -

Example

- -

See Also

- -

_xLET, Special - Rule Elements

- - - - diff --git a/Help/helps/_xCTRL.htm b/Help/helps/_xCTRL.htm deleted file mode 100644 index 782dbb9..0000000 --- a/Help/helps/_xCTRL.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -_xCTRL - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xCTRL

- -

Purpose

- -

Match control and nonalphabetic extended ANSI characters.

- -

Remarks

- -

Example

- -

See Also

- -

_xALPHA,  Special - Rule Elements

- - - - diff --git a/Help/helps/_xEND.htm b/Help/helps/_xEND.htm deleted file mode 100644 index 4f7cf2a..0000000 --- a/Help/helps/_xEND.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -_xEND - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xEND

- -

Purpose

- -

Match if at the end of a phrase (or "segment").

- -

Remarks

- -

Example

- -

See Also

- -

_xSTART, Special - Rule Elements

- - - - diff --git a/Help/helps/_xEOF.htm b/Help/helps/_xEOF.htm deleted file mode 100644 index 243d6bf..0000000 --- a/Help/helps/_xEOF.htm +++ /dev/null @@ -1,55 +0,0 @@ - - - - - - - - - - -_xEOF - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xEOF

- -

Purpose

- -

Match the end of file.

- -

Remarks

- -

Example

- -

See Also

- -

Special Rule Elements

- - - - diff --git a/Help/helps/_xLET.htm b/Help/helps/_xLET.htm deleted file mode 100644 index f400a31..0000000 --- a/Help/helps/_xLET.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -_xLET - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xLET

- -

Purpose

- -

Match an alphabetic consisting of a single letter.

- -

Remarks

- -

Example

- -

See Also

- -

_xCAPLET, Special - Rule Elements

- - - - diff --git a/Help/helps/_xNIL.htm b/Help/helps/_xNIL.htm deleted file mode 100644 index 2a8d77b..0000000 --- a/Help/helps/_xNIL.htm +++ /dev/null @@ -1,61 +0,0 @@ - - - - - - - - - - -_xNIL - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xNIL

- -

Purpose

- -

Match nothing.  Designates - a suggested element when the rule performs a special action, such as removing - the matched nodes from the parse tree.

- -

Remarks

- -

_xNIL has no special action, but serves as documentation for the rule - writer.

- -

Example

- -

See Also

- -

_xWILD, Special - Rule Elements

- - - - diff --git a/Help/helps/_xNUM.htm b/Help/helps/_xNUM.htm deleted file mode 100644 index cf74a41..0000000 --- a/Help/helps/_xNUM.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -_xNUM - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xNUM

- -

Purpose

- -

Match a numeric token.

- -

Remarks

- -

Example

- -

See Also

- -

_xALPHA, Special - Rule Elements

- - - - diff --git a/Help/helps/_xPUNCT.htm b/Help/helps/_xPUNCT.htm deleted file mode 100644 index 257fb59..0000000 --- a/Help/helps/_xPUNCT.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -_xPUNCT - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xPUNCT

- -

Purpose

- -

Match a punctuation token.

- -

Remarks

- -

Example

- -

See Also

- -

_ALPHA, Special - Rule Elements

- - - - diff --git a/Help/helps/_xSTART.htm b/Help/helps/_xSTART.htm deleted file mode 100644 index c2f318d..0000000 --- a/Help/helps/_xSTART.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -_xSTART - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xSTART

- -

Purpose

- -

Match if at the start of a phrase (or "segment").

- -

Remarks

- -

Example

- -

See Also

- -

_xEND, Special - Rule Elements

- - - - diff --git a/Help/helps/_xWHITE.htm b/Help/helps/_xWHITE.htm deleted file mode 100644 index 94a2792..0000000 --- a/Help/helps/_xWHITE.htm +++ /dev/null @@ -1,56 +0,0 @@ - - - - - - - - - - -_xWHITE - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xWHITE

- -

Purpose

- -

Match a whitespace token, including newline.

- -

Remarks

- -

Example

- -

See Also

- -

_xBLANK, Special - Rule Elements

- - - - diff --git a/Help/helps/_xWILD.htm b/Help/helps/_xWILD.htm deleted file mode 100644 index 85565f8..0000000 --- a/Help/helps/_xWILD.htm +++ /dev/null @@ -1,64 +0,0 @@ - - - - - - - - - - -_xWILD - - - - - - - - - - - - - - - - - - - - - - - - - - - -

_xWILD

- -

Purpose

- -

Match anything.  _xWILD - is an unrestricted wildcard. Key-value pairs may add restrictions on number - of nodes matched and on what is matched.

- -

Remarks

- -

With a match or fail list, _xWILD becomes an "OR" matching - function.

- -

Example

- -

See Also

- -

min, - max, matches, - Phrase Element - Modifiers, Special - Rule Elements

- - - - diff --git a/Help/helps/addattr.htm b/Help/helps/addattr.htm deleted file mode 100644 index d0bb9a1..0000000 --- a/Help/helps/addattr.htm +++ /dev/null @@ -1,120 +0,0 @@ - - - - - - - - - - -addattr - - - - - - - - - - - - - - - - - - - - - - - - - - - -

addattr

- -

Purpose

- -

Add an empty attribute named nameString - (one with no value) to the concept named concept.

- -

Syntax

- -

returnedAttribute = addattr(concept, nameString);

- -

returnedAttribute - type: attr
-concept - type: con
-nameString - type: str

- -

Returns

- -

A handle on the newly created attribute.

- -

Remarks

- -

You may know an attribute before you know its value. With this function, - you can create the attribute without mentioning a value. If a bad concept - is passed, a warning is written to output log file - "Warning: Given - no concept." Attribute names are unique, so if you try to add the - same name more than once, i.e. if called repeatedly with the same arguments, - the repeated call is ignored and only one attribute is added.

- -

Example

- -

@CODE
-G("myConcept") = makeconcept(findroot(),"a concept");
-G("myAttr") = addattr(G("myConcept"),"an attribute");
-"output.txt" << attrname(G("myAttr")) << "\n";

- -

This prints out the following to output.txt:

- -

an attribute

- -

See Also

- -

addconval, addnumval, - addstrval, addsval, - Knowledge Base Functions

- -

 

- - - - diff --git a/Help/helps/addcnode.htm b/Help/helps/addcnode.htm deleted file mode 100644 index 5faa64a..0000000 --- a/Help/helps/addcnode.htm +++ /dev/null @@ -1,81 +0,0 @@ - - - - - - - - - - -addcnode - - - - - - - - - - - - - - - - - - - - - - - - - - - -

addcnode

- -

Purpose

- -

Add the named node nodeName - to the end of the concept aConcept's - phrase.

- -

Syntax

- -

returnedConcept = addcnode(aConcept, nodeName)

- -

returnedConcpet - type: con
-aConcept - type:  con
-nodeName - type:

- -

Returns

- -

Remarks

- -

Example

- -

See Also

- -

addnode, makephrase, - Knowledge Base Functions

- - - - diff --git a/Help/helps/addconval.htm b/Help/helps/addconval.htm deleted file mode 100644 index e85607e..0000000 --- a/Help/helps/addconval.htm +++ /dev/null @@ -1,157 +0,0 @@ - - - - - - - - - - -addconval - - - - - - - - - - - - - - - - - - - - - - - - - - - -

addconval

- -

Purpose

- -

Add the concept value valueConcept - to the attribute attrNameString - belonging to the concept concept.

- -

Syntax

- -

addconval(concept, attrNameString, valueConcept)

- -

concept - type: con
-attrNameString - type: str
-valueConcept - type: con

- -

Returns

- -

Nothing

- -

Remarks

- -

Concepts can have zero or more attributes, and attributes can have zero - or more values. Values of attributes can be strings, numbers or other - concepts. This function adds a concept to the value of other another concept's - attribute. All concepts need to be rooted in the concept hierarchy, so - first you need to create them before you can assign them as values of - attributes to other concepts.

- -

If the attribute doesn't exist yet, addconval will create it. If the - concept or the value concept is bad, an error appears in the log output - window.

- -

Example

- -

In this example, we show how we can create a concept like noun and assign - it as the value of a word's attribute. We create the concepts named words and noun - as children to the root of the KB (concept), - and then make the concept book - a child of words.

- -

G("words") = makeconcept(findroot(), "words");
-G("noun") = makeconcept(findroot(),"noun");
-G("noun_book") = makeconcept(G("words"),"book");

- -

The KB editor at this point looks like:

- -

- -

Next, give book - the attribute "is a" - and make the concept noun be the - value of book's attribute "is a":

- -

G("an_attr") = addattr(G("noun_book"),"is a");
-addconval(G("noun_book"), "is a", G("noun"));

- -

The remaining code finds the first child of words - named book, - gets the concept which is the value of the attribute "is - a" and prints out its name - ('noun') to output.txt:

- -

G("aConcept") = getconcept(G("words"),"book");
-G("anotherConcept") = getconval(findvals(G("aConcept"),"is a"));
-"output.txt" << conceptname(G("anotherConcept")) << "\n";

- -

See Also

- -

addsval, addstrval, - addnumval, Knowledge - Base Functions

- - - - diff --git a/Help/helps/addnode.htm b/Help/helps/addnode.htm deleted file mode 100644 index b0c274a..0000000 --- a/Help/helps/addnode.htm +++ /dev/null @@ -1,82 +0,0 @@ - - - - - - - - - - -addnode - - - - - - - - - - - - - - - - - - - - - - - - - - - -

addnode

- -

Purpose

- -

Add named node nodeName as - the numth node number in a phrase - aPhrase.

- -

Syntax

- -

returnedNode = addnode(aPhrase, nodeName, number);

- -

returnedNode - type: con
-aPhrase - type: phr
-nodeName - type: str
-number - type: int

- -

Returns

- -

Remarks

- -

Example

- -

See Also

- -

addcnode, makephrase, - Knowledge Base Functions

- - - - diff --git a/Help/helps/addnumval.htm b/Help/helps/addnumval.htm deleted file mode 100644 index 91daffc..0000000 --- a/Help/helps/addnumval.htm +++ /dev/null @@ -1,108 +0,0 @@ - - - - - - - - - - -addnumval - - - - - - - - - - - - - - - - - - - - - - - - - - - -

addnumval

- -

Purpose

- -

Add the number valueNumber - as a numeric value to the concept concept's - attribute that has name attrNameString.

- -

Syntax

- -

addnumval(concept, attrNameString, valueNumber)

- -

concept - type: con
-attrNameString - type: str
-valueNumber - type: int

- -

Returns

- -

Nothing

- -

Remarks

- -

If the attribute doesn't exist yet, addnumval - will create it. If the concept is bad, an error appears in the log output - window at runtime.

- -

Example

- -

@CODE
-G("Malibu") = makeconcept(findroot(), "Malibu");
-addstrval(G("Malibu"),"Latitude Direction","North");
-addstrval(G("Malibu"),"Longitude Direction","West");
-addnumval(G("Malibu"),"Latitude value",33);
-addnumval(G("Malibu"),"Longitude value",118);
-addsval(G("Malibu"),"Route",1+2);

- -

See Also

- -

addattr, addsval, - addstrval, Knowledge - Base Functions

- - - - diff --git a/Help/helps/addstrval.htm b/Help/helps/addstrval.htm deleted file mode 100644 index 1134344..0000000 --- a/Help/helps/addstrval.htm +++ /dev/null @@ -1,116 +0,0 @@ - - - - - - - - - - -addstrval - - - - - - - - - - - - - - - - - - - - - - - - - - - -

addstrval

- -

Purpose

- -

Add the string valueString - as a string value to the concept concept's - attribute that has name nameString.

- -

Syntax

- -

addstrval(concept, nameString, valueString)

- -

concept - type: con
-nameString - type: str
-valueString - type: str

- -

Returns

- -

Nothing

- -

Remarks

- -

If the attribute doesn't exist yet, addstrval - will create it. If the concept is bad, an error appears in the log output - window at runtime.

- -

Example

- -

G("Malibu") = makeconcept(findroot(), "Malibu");
-addstrval(G("Malibu"),"Latitude Direction","North");
-addstrval(G("Malibu"),"Longitude Direction","West");
-addnumval(G("Malibu"),"Latitude value",33);
-addnumval(G("Malibu"),"Longitude value",118);
-addsval(G("Malibu"),"Route",1+2);

- -

See Also

- -

addconval, addsval, addnumval, Knowledge Base Functions

- -

 

- - - - diff --git a/Help/helps/addsval.htm b/Help/helps/addsval.htm deleted file mode 100644 index b74d35e..0000000 --- a/Help/helps/addsval.htm +++ /dev/null @@ -1,117 +0,0 @@ - - - - - - - - - - -addsval - - - - - - - - - - - - - - - - - - - - - - - - - - - -

addsval

- -

Purpose

- -

Add numeric value number as - a string to concept concept's - attribute named nameString.

- -

Syntax

- -

addsval(concept, nameString, number)

- -

concept - type: con
-nameString - type: str
-number - type: int

- -

Returns

- -

Nothing.

- -

Remarks

- -

If the attribute has not been created yet, addsval - will create it.

- -

Sometimes it is convenient to treat a number like a string. In this - case, you can perform numeric operations in the number field of this function, - but then have the result stored in the KB as a string.

- -

Example

- -

@CODE

- -

G("Malibu") = makeconcept(findroot(), "Malibu");
-addsval(G("Malibu"),"Route",1+2);

- -

See Also

- -

addstrval, addnumval, addconval, Knowledge Base Functions

- -

 

- - - - diff --git a/Help/helps/addword.htm b/Help/helps/addword.htm deleted file mode 100644 index c0ae25c..0000000 --- a/Help/helps/addword.htm +++ /dev/null @@ -1,121 +0,0 @@ - - - - - - - - - - -addword - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

addword

- -

Purpose

- -

Add dictionary concept wordString - to the KB dictionary hierarchy.  Return - a handle to concept.

- -

Syntax

- -

returnedWordConcept = addword(wordString)

- -

returnedWordConcept - type: con
-wordString - type: str

- -

Returns

- -

Handle to the newly created word concept. If the word cannot be added, - returns 0.

- -

Remarks

- -

The dictionary is a specialized part of the KB hierarchy used to store - and lookup words.

- -

Example

- -

@CODE
-"output.txt" << "1 " << conceptname(addword("hello")) << "\n";
-"output.txt" << "2 " << conceptname(wordindex("hello")) << "\n";
-"output.txt" << "3 " << findwordpath("hello") << "\n";
-"output.txt" << "4 " << findwordpath("olleh") << "\n";
-"output.txt" << "5 " << wordpath("foobaz") << "\n";
-"output.txt" << "6 " << conceptname(dictfindword("hello")) << \n";
-rmword("foobaz");

- -

Prints out:

- -

1 hello
-2 he
-3 "concept" "sys" "dict" "a" "h" "he" "hello"
-4
-5 "concept" "sys" "dict" "a" "f" "fo" "foobaz"
-6 hello

- -

See Also

- -

dictgetword, wordindex, - wordpath, rmword, - Knowledge Base Functions

- -

 

- - - - diff --git a/Help/helps/arraylength.htm b/Help/helps/arraylength.htm deleted file mode 100644 index 6eac675..0000000 --- a/Help/helps/arraylength.htm +++ /dev/null @@ -1,86 +0,0 @@ - - - - - - - - - - -arraylength - - - - - - - - - - - - - - - - - - - - - - - - - - - -

arraylength

- -

Purpose

- -

Count the number of values in a given variable.

- -

Syntax

- -

returnedNumber = arraylength(variableName)

- -

returnedNumber - type: int
-variableName - type: var

- -

Returns

- -

Remarks

- -

Note that the variableName must evaluate to a variable reference (G, - S, X, N).

- -

WARNING: arraylength() - never returns 0, since every variable is considered to have at least one - value.  This - may yield unexpected behaviors at times.

- -

Example

- -

See Also

- -

Parse Tree Functions -

- - - - diff --git a/Help/helps/attr.htm b/Help/helps/attr.htm deleted file mode 100644 index 5e960c8..0000000 --- a/Help/helps/attr.htm +++ /dev/null @@ -1,69 +0,0 @@ - - - - - - - - - - -attr - - - - - - - - - - - - - - - - - - - - - - - - - - - -

attr

- -

Purpose

- -

Layer additional attributes for the element in the parse tree, as "mini-reductions". -

- -

Remarks

- -

Use the names in the list to name nodes. Each node that matched current - rule element will be layered.  layer, layers, - and attrs are alternate forms - for attr.  

- -

attr can also be used as a suggested - element modifier.

- -

Example

- -

See Also

- -

Suggested - Element Modifiers, Phrase - Element Modifiers

- -

 

- - - - diff --git a/Help/helps/attrchange.htm b/Help/helps/attrchange.htm deleted file mode 100644 index ee8f264..0000000 --- a/Help/helps/attrchange.htm +++ /dev/null @@ -1,151 +0,0 @@ - - - - - - - - - - -attrchange - - - - - - - - - - - - - - - - - - - - - - - - - - - -

attrchange

- -

Purpose

- -

Replace all matching attribute-value pairs attrStr-oldValStr in the KB belonging to the - concept aConcept to new attribute-value - pairs attrStr-newValStr.

- -

Syntax

- -

returnedBoolean = attrchange(aConcept, attrStr, oldValStr, newValStr)

- -

returnedBoolean - type: bool
-aConcept - type: con
-attrStr - type: str
-oldValStr - type: str
-newValStr - type: str

- -

Returns

- -

1 if at least one pair is changed, 0 otherwise.

- -

Remarks

- -

Think of this as a global replace. If passed a bad concept, the function - writes an error to the output log.

- -

Example

- -

# if you find apples in the concept hierarchy
-if (findconcept(findroot(),"apple"))
-# kill them (to start fresh)
-rmconcept(findconcept(findroot(),"apple"));
-# Create the apple concept
-G("apple") = makeconcept(findroot(),"apple");
-# Apple's color's red
-addstrval(G("apple"),"color","red");
-# MacIntosh apples are also red
-addstrval(makeconcept(G("apple"),"MacIntosh"),"color","red");
-# Print out apple's attribute and value
-if(attrname(findattrs(G("apple")))){
-"output.txt" << "1) apples " << attrname(findattrs(G("apple")));
-"output.txt" << " "
-<< getstrval(attrvals(findattrs(G("apple")))) << "\n";
-}
-# Test for a given attribute and value
-"output.txt" << "2) are apples red? "
-<< attrwithval(G("apple"), "color", "red") << "\n";
-"output.txt" << "3) are apples blue? " <<
-attrwithval(G("apple"), "color", "blue") << "\n";
-# Change a value of a given attribute to another value.
-attrchange(findroot(), "color", "red", "blue");
-"output.txt" << "4) are apples blue now? " <<
-attrwithval(G("apple"), "color", "blue") << "\n";

- -

The code above writes the following to output.txt:

- -

1) apples color red
-2) are apples red? 1
-3) are apples blue? 0
-4) are apples blue now? 1

- -

Since attrchange operates on all children of the parent concept, the - resulting KB looks like:

- -

- -

See Also

- -

replaceval, Knowledge Base Functions -

- - - - diff --git a/Help/helps/attrexists.htm b/Help/helps/attrexists.htm deleted file mode 100644 index ccce57f..0000000 --- a/Help/helps/attrexists.htm +++ /dev/null @@ -1,138 +0,0 @@ - - - - - - - - - - -attrexists - - - - - - - - - - - - - - - - - - - - - - - - - - - -

attrexists

- -

Purpose

- -

Find attribute attrStr with - value valStr, searching in the - subhierarchy of concept parentConcept. -  

- -

Syntax

- -

returnedBoolean = attrexists(parentConcept, attrStr, valStr)

- -

returnedBoolean - type: bool
-parentConcept - type: con
-attrStr - type: str
-valStr - type: str

- -

Returns

- -

1 if any concept below parentConcept - has the attribute-value pair, 0 otherwise. Passed a bad concept, the function - writes an error to the output log.

- -

Remarks

- -

The search starts at the named concept, and continues down its entire - subhierarchy. Contrast this with attrwithval, - which is limited to only the argument concept.

- -

Example

- -

# if you find apples in the concept hierarchy
-if (findconcept(findroot(),"apple"))
-# kill them (to start fresh)
-rmconcept(findconcept(findroot(),"apple"));
-# Create the apple concept
-G("apple") = makeconcept(findroot(),"apple");
-# Apple's color's red
-addstrval(G("apple"),"color","red");
-# MacIntosh apples are also red
-addstrval(makeconcept(G("apple"),"MacIntosh"),"color","red");
-# Print out apple's attribute and value
-if(attrname(findattrs(G("apple")))){
-"output.txt" << "1) apples " << attrname(findattrs(G("apple")));
-"output.txt" << " "
-<< getstrval(attrvals(findattrs(G("apple")))) << "\n";
-}
-# Test for a given attribute and value
-"output.txt" << "2) are apples red? " <<
-attrwithval(G("apple"), "color", "red") << "\n";
-"output.txt" << "3) are apples blue? " <<
-attrwithval(G("apple"), "color", "blue") << "\n";
-# Change a value of a given attribute to another value.
-attrchange(findroot(), "color", "red", "blue");
-"output.txt" << "4) are apples blue now? " <<
-attrwithval(G("apple"), "color", "blue") << "\n";
-"output.txt" << "5) " << attrexists(findroot(), "color", "blue");

- -

The code above writes the following to output.txt:

- -

1) apples color red
-2) are apples red? 1
-3) are apples blue? 0
-4) are apples blue now? 1
-5) 1

- -

See Also

- -

attrwithval, - Knowledge Base Functions

- - - - diff --git a/Help/helps/attrname.htm b/Help/helps/attrname.htm deleted file mode 100644 index 80e60a1..0000000 --- a/Help/helps/attrname.htm +++ /dev/null @@ -1,104 +0,0 @@ - - - - - - - - - - -attrname - - - - - - - - - - - - - - - - - - - - - - - - - - - -

attrname

- -

Purpose

- -

Fetch the name of an knowledge base attribute.

- -

Syntax

- -

returnedString = attrname(anAttribute);

- -

returnedString - type: str
-anAttribute - type: con

- -

Returns

- -

A string which is the name of the passed attribute.

- -

Remarks

- -

If you try to print out an attribute directly, you'll get an error. - This function returns the string value name of an attribute. If a bad - attribute is passed, an error is written to output log file.

- -

Example

- -

G("myConcept") = makeconcept(findroot(),"a concept");
-G("myAttr") = addattr(G("myConcept"),"an attribute");
-"output.txt" << attrname(G("myAttr")) << "\n";

- -

This prints out

- -

an attribute

- -

See Also

- -

conceptname, - Knowledge Base Functions

- - - - diff --git a/Help/helps/attrs.htm b/Help/helps/attrs.htm deleted file mode 100644 index 91e8cdc..0000000 --- a/Help/helps/attrs.htm +++ /dev/null @@ -1,69 +0,0 @@ - - - - - - - - - - -attrs - - - - - - - - - - - - - - - - - - - - - - - - - - - -

attrs

- -

Purpose

- -

Layer additional attributes for the element in the parse tree, as "mini-reductions". -

- -

Remarks

- -

Use the names in the list to name nodes. Each node that matched current - rule element will be layered.  layer, layers, - and attr are alternate forms for - attrs.  

- -

attrs can also be used as a - suggested element modifier.

- -

Example

- -

See Also

- -

Suggested - Element Modifiers, Phrase - Element Modifiers

- -

 

- - - - diff --git a/Help/helps/attrtype.htm b/Help/helps/attrtype.htm deleted file mode 100644 index 5ab108b..0000000 --- a/Help/helps/attrtype.htm +++ /dev/null @@ -1,76 +0,0 @@ - - - - - - - - - - -attrtype - - - - - - - - - - - - - - - - - - - - - - - - - - - -

attrtype

- -

Purpose

- -

Fetch an attribute's type

- -

Syntax

- -

type = attrtype(concept,attr_name);

- -

returnedValue(s) - type: number
-anAttribute - type: attr

- -

Returns

- -

Returns 0 for string, 1 for number, 2 for concept, and 3 for float.

- -

Remarks

- -

This allows for calling the correct function, either getstrval, getnumval, or getconval.

- -

Example

- - - diff --git a/Help/helps/attrvals.htm b/Help/helps/attrvals.htm deleted file mode 100644 index 6888316..0000000 --- a/Help/helps/attrvals.htm +++ /dev/null @@ -1,188 +0,0 @@ - - - - - - - - - - -attrvals - - - - - - - - - - - - - - - - - - - - - - - - - - - -

attrvals

- -

Purpose

- -

Fetch attribute's list of values.

- -

Syntax

- -

returnedValue(s) = attrvals(anAttribute);

- -

returnedValue(s) - type: val
-anAttribute - type: attr

- -

Returns

- -

A list of values.

- -

Remarks

- -

Concepts can have zero or more attributes. Each attribute can zero or - more values. This function takes an attribute as argument and returns - the list of values associated with the attribute. If there are no values - of the attribute, the function returns NULL.

- -

Example

- -

To demonstrate attrvals, we first need to build a KB:

- -

@CODE
-# if you find apples in the concept hierarchy
-if (findconcept(findroot(),"apple"))
-# kill them (to start fresh)
-rmconcept(findconcept(findroot(),"apple"));
-# Create the apple concept
-G("apple") = makeconcept(findroot(),"apple");
-# Apples have color
-addstrval(G("apple"),"have","color");
-# Apple's color is red
-addstrval(G("apple"),"color","red");
-# Apple's color is also green and yellow
-addstrval(G("apple"),"color","green and yellow");

- -

The code creates a KB like this:

- -

- -

The following code accesses the KB's attributes and values:

- -

# Find apple's attribute list
-G("attrList") = findattrs(G("apple"));
-# Find the list of values of the first attribute
-G("valList") = attrvals(G("attrList"));
-# print out the first attribute's name and value
-if(attrname(G("attrList"))){
-"output.txt" << "1) first attribute of apple is: "
-<< attrname(G("attrList")) << "\n";
-"output.txt" << "2) first value of that attribute is: "
-<< getstrval(G("valList")) << "\n";}
-# get the next attribute
-G("nextAttr") = nextattr(G("attrList"));
-if(attrname(nextattr(G("attrList"))))
-"output.txt" << "3) next attribute of apple is: "
-<< attrname(G("nextAttr")) << "\n";
-# get the list of values of the second attribute
-G("valList") = attrvals(G("nextAttr"));
-# print the first value's name
-"output.txt" << "4) first value of that attribute is: "
-<< getstrval(G("valList")) << "\n";
-# print the second value of the second attribute
-"output.txt" << "5) second value of that attribute is: "
-<< getstrval(nextval(G("valList"))) << "\n";

- -

The output looks like this:

- -

- -

See Also

- -

attrexists, attrwithval, findvals, - Knowledge Base Functions

- - - - diff --git a/Help/helps/attrwithval.htm b/Help/helps/attrwithval.htm deleted file mode 100644 index 84d8088..0000000 --- a/Help/helps/attrwithval.htm +++ /dev/null @@ -1,138 +0,0 @@ - - - - - - - - - - -attrwithval - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

attrwithval

- -

Purpose

- -

Check concept to see if it - has attribute attrStr with value - valStr.

- -

Syntax

- -

returnedBoolean = attrwithval(concept, attrStr, valStr)

- -

returnedBoolean - type: bool (1 or 0)
-concept - type: con
-attrStr - type: str
-valStr - type: str

- -

Returns

- -

1 if the concept has the attribute-value pair, 0 otherwise. Passed a - bad concept, the function writes an error to the output log.

- -

Remarks

- -

If an attribute has multiple values, returns true if one of the values - matches valStr.  Limited - to string-valued attributes.

- -

Example

- -

# if you find apples in the concept hierarchy
-if (findconcept(findroot(),"apple"))
-# kill them (to start fresh)
-rmconcept(findconcept(findroot(),"apple"));
-# Create the apple concept
-G("apple") = makeconcept(findroot(),"apple");
-# Apple's color's red
-addstrval(G("apple"),"color","red");
-# MacIntosh apples are also red
-addstrval(makeconcept(G("apple"),"MacIntosh"),"color","red");
-# Print out apple's attribute and value
-if(attrname(findattrs(G("apple")))){
-"output.txt" << "1) apples " << attrname(findattrs(G("apple")));
-"output.txt" << " "
-<< getstrval(attrvals(findattrs(G("apple")))) << "\n";
-}
-# Test for a given attribute and value
-"output.txt" << "2) are apples red? " <<
-attrwithval(G("apple"), "color", "red") << "\n";
-"output.txt" << "3) are apples blue? " <<
-attrwithval(G("apple"), "color", "blue") << "\n";
-# Change a value of a given attribute to another value.
-attrchange(findroot(), "color", "red", "blue");
-"output.txt" << "4) are apples blue now? " <<
-attrwithval(G("apple"), "color", "blue") << "\n";

- -

The code above writes the following to output.txt:

- -

1) apples color red
-2) are apples red? 1
-3) are apples blue? 0
-4) are apples blue now? 1

- -

See Also

- -

attrexists, attrvals, -  Knowledge - Base Functions

- -

 

- - - - diff --git a/Help/helps/base.htm b/Help/helps/base.htm deleted file mode 100644 index f2a978b..0000000 --- a/Help/helps/base.htm +++ /dev/null @@ -1,79 +0,0 @@ - - - - - - - - - - -base - - - - - - - - - - - - - - - - - - - - - - - - - - - -

base

- -

Purpose

- -

The base suggested-element modifier specifies that the reduce node will - be the bottom-most node in a singlet - chain.  That - is, subsequent rules using the singlet element modifier will not look - "down" past a node labeled with the base attribute.

- -

Remarks

- -

Base and singlet provide a mechanism for layering multiple attributes - and syntactic and semantic ambiguities directly into the parse tree.

- -

The example will help clarify the use of the base modifier.

- -

Example

- -

# The rule below reduces a node labeled _noun to an _np.  It - places an internal flag called "base" on the _np node.
-# Subsequent rules (e.g., in subsequent passes) that are searching for - _noun will not find it under an _np created with the current rule.
-@RULES
-_np [base] <- _noun @@

- -

# Say a subsequent pass has this rule.  Even - though the s (or singlet) modifier says to look "within" a node - for _noun, it won't find it within an _np
-# labeled with a base flag, as created by the rule above.
-@RULES
-_np <- _adj _noun [s] @@

- -

 

- -

Suggested - Element Modifiers.

- - - - diff --git a/Help/helps/batchstart.htm b/Help/helps/batchstart.htm deleted file mode 100644 index 8075ef2..0000000 --- a/Help/helps/batchstart.htm +++ /dev/null @@ -1,91 +0,0 @@ - - - - - - - - - - -batchstart - - - - - - - - - - - - - - - - - - - - - - - - - - - -

batchstart

- -

Purpose

- -

Check if a batch analysis process has just begun.

- -

Syntax

- -

returnedBool = batchstart()

- -

returnedBool - type: bool

- -

Returns

- -

True if we're processing the first file in a batch, as signaled by the - Batch Toggle in the VisualText GUI.

- -

Remarks

- -

Example

- -

@CODE
-G("startflag") = batchstart();
-@@CODE

- -

See Also

- -

Special Functions

- -

 

- - - - diff --git a/Help/helps/cap.htm b/Help/helps/cap.htm deleted file mode 100644 index 543b517..0000000 --- a/Help/helps/cap.htm +++ /dev/null @@ -1,95 +0,0 @@ - - - - - - - - - - -cap - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

cap

- -

Purpose

- -

Pre action that succeeds if leaf token has capitalized first letter.

- -

Syntax

- -

<fromRuleEltNumber,toRuleEltNumber>cap();

- -

fromRuleEltNumber - type: int
-toRuleEltNumber - type: int

- -

Returns

- -

Remarks

- -

Example

- -

@PRE
-<1,1>cap();
-@RULES
-_noun <- _xALPHA@@

- -

See Also

- -

uppercase, lowercase, - PRE Actions -

- -

 

- -

 

- - - - diff --git a/Help/helps/cbuf.htm b/Help/helps/cbuf.htm deleted file mode 100644 index 0fb8845..0000000 --- a/Help/helps/cbuf.htm +++ /dev/null @@ -1,143 +0,0 @@ - - - - - - - - - - -cbuf - - - - - - - - - - - - - - - - - - - - - - - - - - - -

cbuf

- -

Purpose

- -

Send output to user-supplied buffer.

- -

Syntax

- -

returnedOstream = cbuf()

- -

returnedOstream - type: ostream

- -

Returns

- -

Remarks

- -

For use in embedded or standalone analyzers.  User - may supply an output buffer to an API call such as nlp->analyze(). -  Then sending - output to cbuf() will write to the user-supplied output stream.

- -

Example

- -

@CODE

- -

# In VisualText, output to a file.  Outside VisualText, output to user-supplied buffer.

- -

if (interactive())

- -

   G("out") = "buf.txt";

- -

else

- -

   G("out") = cbuf();

- -

G("out") << "Hello output buffer!" << "\n";

- -

@@CODE

- -

See Also

- - - - diff --git a/Help/helps/closefile.htm b/Help/helps/closefile.htm deleted file mode 100644 index 7354f35..0000000 --- a/Help/helps/closefile.htm +++ /dev/null @@ -1,95 +0,0 @@ - - - - - - - - - - -closefile - - - - - - - - - - - - - - - - - - - - - - - - - - - -

closefile

- -

Purpose

- -

Close the output file stream.

- -

Syntax

- -

returnedBool = closefile(fileOstream)

- -

returnedBool - type: bool
-fileOstream - type: ostream

- -

Returns

- -

Remarks

- -

Example

- -

@CODE
-G("file") = openfile("c:\\tmp.txt");
-closefile(G("file"));
-@@CODE

- -

See Also

- -

openfile, Formatting - and I/O Functions

- -

 

- - - - diff --git a/Help/helps/conceptname.htm b/Help/helps/conceptname.htm deleted file mode 100644 index 821fa2c..0000000 --- a/Help/helps/conceptname.htm +++ /dev/null @@ -1,111 +0,0 @@ - - - - - - - - - - -conceptname - - - - - - - - - - - - - - - - - - - - - - - - - - - -

conceptname

- -

Purpose

- -

Fetch the string value name of a concept.

- -

Syntax

- -

returnedString = conceptname(concept)

- -

returnedString - type: str
-concept - type: con

- -

Returns

- -

Name of concept as a string.

- -

Remarks

- -

You cannot directly print out the name of a concept by referring to - its handle, but instead must use this function to access its string valued - name. If the concept is bad, an error appears in the log output window.

- -

Example

- -

@CODE

- -

"output.txt" << conceptname(makeconcept(findroot(),"test"));

- -

Prints out the following in output.txt

- -

test

- -

See Also

- -

attrname, Knowledge - Base Functions

- - - - diff --git a/Help/helps/conceptpath.htm b/Help/helps/conceptpath.htm deleted file mode 100644 index 2d36453..0000000 --- a/Help/helps/conceptpath.htm +++ /dev/null @@ -1,105 +0,0 @@ - - - - - - - - - - -conceptpath - - - - - - - - - - - - - - - - - - - - - - - - - - - -

conceptpath

- -

Purpose

- -

 Fetch entire path of given concept as a string.

- -

Syntax

- -

returnedString = conceptpath(concept)

- -

returnedString - type: str
-concept - type: con

- -

Returns

- -

Path of concept as a string.

- -

Remarks

- -

If the concept is bad, an error appears in the output log window.

- -

Example

- -

G("childConcept") = makeconcept(makeconcept(findroot(), "parent"), "child");
-"output.txt" << conceptpath(G("childConcept"));

- -

Writes the following to output.txt:

- -

"concept" "parent" "child"

- -

See Also

- -

attrname, findwordpath, - wordpath, Knowledge - Base Functions

- - - - diff --git a/Help/helps/conval.htm b/Help/helps/conval.htm deleted file mode 100644 index 8882601..0000000 --- a/Help/helps/conval.htm +++ /dev/null @@ -1,162 +0,0 @@ - - - - - - - - - - -conval - - - - - - - - - - - - - - - - - - - - - - - - - - - -

conval

- -

Purpose

- -

Fetch concept-value of attribute attrNameString - (must be first) belonging to concept concept.

- -

Syntax

- -

returnedConcept = conval(concept, attrNameString)

- -

returnedConcept - type: con
-concept - type: con
-attrNameString - type: str

- -

Returns

- -

returnedConcept - type: con

- -

Remarks

- -

An attribute's value can be a concept, and conval returns attribute - values of this type.  If - an attribute has multiple values, it only fetches the first.  If - it can't fetch a concept value, it returns 0.

- -

Example

- -

First, create the concepts named words - and noun as children to the - root of the KB, then make the concept book - a child of words.

- -

G("words") = makeconcept(findroot(), "words");
-G("noun") = makeconcept(findroot(),"noun");
-G("noun_book") = makeconcept(G("words"),"book");

- -

The KB editor at this point looks like:

- -

- -

Next, give book - the attribute "is a" - and make the concept noun be - the value of book's attribute - "is a":

- -

G("an_attr") = addattr(G("noun_book"),"is a");
-addconval(G("noun_book"), "is a", G("noun"));

- -

The remaining code fetches the value of the "is a" attribute, - i.e., the noun concept, and prints its name:

- -

G("bConcept") = conval(G("noun_book"),"is a");

- -

"output.txt" << conceptname(G("bConcept")) << "\n";

- -

This prints the following to output.txt:

- -

noun

- -

See Also

- -

numval, strval, - getconval, Knowledge - Base Functions

- - - - diff --git a/Help/helps/cout.htm b/Help/helps/cout.htm deleted file mode 100644 index 50a23b2..0000000 --- a/Help/helps/cout.htm +++ /dev/null @@ -1,143 +0,0 @@ - - - - - - - - - - -cout - - - - - - - - - - - - - - - - - - - - - - - - - - - -

cout

- -

Purpose

- -

Send output to standard output or user-supplied stream.

- -

Syntax

- -

returnedOstream = cout()

- -

returnedOstream - type: ostream

- -

Returns

- -

Remarks

- -

For use in embedded or standalone analyzers.  User - may supply an output stream to an API call such as nlp->analyze(). -  Then sending - output to cout() will write to the user-supplied output stream.

- -

Example

- -

@CODE

- -

# In VisualText, output to a file.  Outside VisualText, output to user-supplied stream.

- -

if (interactive())

- -

   G("out") = "out.txt";

- -

else

- -

   G("out") = cout();

- -

G("out") << "Hello output stream!" << "\n";

- -

@@CODE

- -

See Also

- - - - diff --git a/Help/helps/coutreset.htm b/Help/helps/coutreset.htm deleted file mode 100644 index 0dd1b6f..0000000 --- a/Help/helps/coutreset.htm +++ /dev/null @@ -1,133 +0,0 @@ - - - - - - - - - - -coutreset - - - - - - - - - - - - - - - - - - - - - - - - - - - -

coutreset

- -

Purpose

- -

Rebind the output stream cout to a given file.

- -

Syntax

- -

returnedOstream = coutreset(fileNameStr)

- -

returnedOstream - type: ostream

- -

fileNameStr - type: string

- -

Returns

- -

Remarks

- -

In VisualText, a way to bind where outputs to cout() are sent.

- -

Example

- -

@CODE

- -

# In VisualText, output to a file.  Outside VisualText, output to user-supplied stream.

- -

if (interactive())

- -

   coutreset("out.txt");

- -

cout() << "Hello output stream!" << "\n";

- -

@@CODE

- -

See Also

- - - - diff --git a/Help/helps/dballocstmt.htm b/Help/helps/dballocstmt.htm deleted file mode 100644 index fb737c2..0000000 --- a/Help/helps/dballocstmt.htm +++ /dev/null @@ -1,151 +0,0 @@ - - - - - - - - - - -dballocstmt - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dballocstmt

- -

Purpose

- -

Allocate statement handle for the currently open database.

- -

Syntax

- -

returnedBool = dballocstmt()

- -

returnedBool - type: bool
 

- -

Returns

- -

True if successful.

- -

Remarks

- -

NLP++ database functions assume only one statement handle is open at - a time.  In - addition, one can call dbexec() - to execute a statement in a separate fashion.

- -

Example

- -

@CODE

- -

dbopen("test","root","mypassword");

- -

dballocstmt();

- -

dbexecstmt("INSERT INTO employee (name, age) VALUES('John Smith','32');");

- -

dbfreestmt();

- -

dbclose();

- -

@@CODE

- -

 

- -

See Also

- -

Table of Database Functions

- -

Database1 Sample Analyzer

- -

 

- - - - diff --git a/Help/helps/dbbindcol.htm b/Help/helps/dbbindcol.htm deleted file mode 100644 index 5beb991..0000000 --- a/Help/helps/dbbindcol.htm +++ /dev/null @@ -1,1494 +0,0 @@ - - - - - - - - - - -dbbindcol - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dbbindcol

- -

Purpose

- -

Bind a database table's column to an NLP++ variable.

- -

Syntax

- -

return_bool = dbbindcol(column_num, type_str, size_num, &var, &flag_var)

- -

return_bool - True if successful, else false

- -

column_num - Database column number

- -

type_str - Database type for the given column.  See table below

- -

size_num - For some database types, indicates the required storage size for retrieving the data

- -

&var - (call by reference). NLP++ variable to receive the fetched column value.

- -

&flag_var - (call by reference).  NLP++ variable to indicate whether the column had a value.
- 

- -

Returns

- -

True if successful.

- -

Remarks

- -

The associated function dbfetch - fetches a database table row, filling all the bound variables.

- -

Database Types and Conversion

- -

The supported database types (those not in parentheses) and their conversion - to NLP++ types are listed here:

- - - ---- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-

DATABASE TYPE

 -

NLP++ TYPE

 -

SPECIFY SIZE?

-

BIGINT

 -

NUM

 -

no

-

BINARY

 -

STR

 -

yes

-

BIT

 -

NUM

 -

no

-

(BLOB)

 -

---

 -

---

-

CHAR

 -

STR

 -

yes

-

DATE

 -

STR

 -

no

-

DATETIME

 -

STR

 -

no

-

DECIMAL

 -

FLOAT

 -

no

-

DOUBLE

 -

FLOAT

 -

no

-

ENUM

 -

STR

 -

yes

-

FLOAT

 -

FLOAT

 -

no

-

INT

 -

NUM

 -

no

-

INTEGER

 -

NUM

 -

no

-

(LONGBLOB)

 -

---

 -

---

-

LONGVARBINARY

 -

STR

 -

yes

-

(MEDIUMBLOB)

 -

---

 -

---

-

MEDIUMINT

 -

NUM

 -

no

-

NUMERIC

 -

FLOAT

 -

no

-

REAL

 -

FLOAT

 -

no

-

SET

 -

STR

 -

yes

-

SMALLINT

 -

NUM

 -

no

-

TIME

 -

STR

 -

no

-

TIMESTAMP

 -

STR

 -

no

-

TINYBLOB

 -

STR

 -

yes

-

TINYINT

 -

NUM

 -

no

-

VARBINARY

 -

STR

 -

yes

-

VARCHAR

 -

STR

 -

yes

-

YEAR

 -

STR

 -

no

-

 

 -

 

 -

 

- -

Example

- -

@CODE

- -

dbopen("test","root","mypassword");

- -

dballocstmt();

- -

dbexecstmt("SELECT * FROM employee;");

- -

dbbindcol(1,"varchar",50,&G("employee name"),&G("result1"));

- -

while (dbfetch())

- -

   {

- -

   "output.txt" << "employee name: ";

- -

   if (G("result1"))

- -

      "output.txt" << G("employee name") << "\n";

- -

   else

- -

      "output.txt" << "NULL" << "\n";

- -

   }

- -

dbfreestmt();

- -

dbclose();

- -

@@CODE

- -

 

- -

See Also

- -

Table of Database Functions

- -

Database1 Analyzer

- -

 

- -

 

- - - - diff --git a/Help/helps/dbclose.htm b/Help/helps/dbclose.htm deleted file mode 100644 index 9f510e5..0000000 --- a/Help/helps/dbclose.htm +++ /dev/null @@ -1,235 +0,0 @@ - - - - - - - - - - -dbclose - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dbclose

- -

Purpose

- -

Close the currently open database.

- -

Syntax

- -

return_bool = dbclose()

- -

return_bool - True if successful, else false
- 

- -

Returns

- -

True if successful.

- -

Remarks

- -

Close the currently open database.  NLP++ - currently allows only one database to be open at a time.

- -

Example

- -

@CODE

- -

dbopen("test","root","mypassword");

- -

dballocstmt();

- -

dbexecstmt("SELECT * FROM table;");

- -

dbbindcol(1,"varchar",50,&G("employee name"),&G("result1"));

- -

while (dbfetch())

- -

   {

- -

   "output.txt" << "employee name: ";

- -

   if (G("result1"))

- -

      "output.txt" << G("employee name") << "\n";

- -

   else

- -

      "output.txt" << "NULL" << "\n";

- -

   }

- -

dbfreestmt();

- -

dbclose();

- -

@@CODE

- -

 

- -

See Also

- -

Table of Database Functions

- -

Database1 Analyzer

- -

 

- -

 

- - - - diff --git a/Help/helps/dbexec.htm b/Help/helps/dbexec.htm deleted file mode 100644 index 1d4661f..0000000 --- a/Help/helps/dbexec.htm +++ /dev/null @@ -1,153 +0,0 @@ - - - - - - - - - - -dbexec - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dbexec

- -

Purpose

- -

Execute a statement on the currently open database.

- -

Syntax

- -

return_bool = dbexec(command_str)

- -

return_bool - True if successful, else false

- -

command_str - Statement to send to the database 

- -

Returns

- -

True if successful.

- -

Remarks

- -

Unlike dbexecstmt(), dbexec() - executes a statement in modular fashion anytime a database is open.  It - allocates a statement handle, sends the statement, then closes its statement - handle.

- -

Example

- -

@CODE

- -

dbopen("test","root","mypassword");

- -

dbexec("INSERT INTO abc (name, number) VALUES('Joe','0013');");

- -

dbclose();

- -

@@CODE

- -

 

- -

See Also

- -

Table of Database Functions

- -

Database1 Analyzer

- -

 

- -

 

- - - - diff --git a/Help/helps/dbexecstmt.htm b/Help/helps/dbexecstmt.htm deleted file mode 100644 index 5ab8c41..0000000 --- a/Help/helps/dbexecstmt.htm +++ /dev/null @@ -1,203 +0,0 @@ - - - - - - - - - - -dbexecstmt - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dbexecstmt

- -

Purpose

- -

Execute a statement with the currently open statement handle.

- -

Syntax

- -

return_bool = dbexecstmt(command_str)

- -

return_bool - True if successful, else false

- -

command_str - Statement to send to the database 

- -  - -

 

- -

Returns

- -

True if successful.

- -

Remarks

- -

Unlike dbexec(), dbexecstmt() - assumes there is a currently open statement handle.  NLP++ - currently allows only a single statement handle to be active at a time.

- -

Example

- -

@CODE

- -

dbopen("test","root","mypassword");

- -

dballocstmt();

- -

dbexecstmt("INSERT INTO abc (name, number) VALUES('Jane','0011');");

- -

dbexecstmt("INSERT INTO abc (name, number) VALUES('Joe','0013');");

- -

dbfreestmt();

- -

dbclose();

- -

@@CODE

- -

 

- -

See Also

- -

Table of Database Functions

- -

Database1 Analyzer

- -

 

- -

 

- - - - diff --git a/Help/helps/dbfetch.htm b/Help/helps/dbfetch.htm deleted file mode 100644 index 3b84828..0000000 --- a/Help/helps/dbfetch.htm +++ /dev/null @@ -1,227 +0,0 @@ - - - - - - - - - - -dbfetch - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dbfetch

- -

Purpose

- -

Fetch a row from the current database result.

- -

Syntax

- -

return_bool = dbfetch()

- -

return_bool - True if successful, else false

- -

Returns

- -

True if successful.

- -

Remarks

- -

Assumes that preceding calls to dbbindcol() have bound NLP++ variables - to database columns of the table that has been queried.

- -

Example

- -

@CODE

- -

dbopen("test","root","mypassword");

- -

dballocstmt();

- -

dbexecstmt("SELECT * FROM employee;");

- -

dbbindcol(1,"varchar",50,&G("employee name"),&G("result1"));

- -

while (dbfetch())

- -

   {

- -

   "output.txt" << "employee name: ";

- -

   if (G("result1"))

- -

      "output.txt" << G("employee name") << "\n";

- -

   else

- -

      "output.txt" << "NULL" << "\n";

- -

   }

- -

dbfreestmt();

- -

dbclose();

- -

@@CODE

- -

 

- -

See Also

- -

Table of Database Functions

- -

Database1 Analyzer

- -

 

- -

 

- - - - diff --git a/Help/helps/dbfreestmt.htm b/Help/helps/dbfreestmt.htm deleted file mode 100644 index bcb9839..0000000 --- a/Help/helps/dbfreestmt.htm +++ /dev/null @@ -1,151 +0,0 @@ - - - - - - - - - - -dbfreestmt - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dbfreestmt

- -

Purpose

- -

Free the statement handle for the currently open database.

- -

Syntax

- -

returnedBool = dbfreestmt()

- -

returnedBool - type: bool
 

- -

Returns

- -

True if successful.

- -

Remarks

- -

NLP++ database functions assume only one statement handle is open at - a time.  In - addition, one can call dbexec() - to execute a statement in a separate fashion.

- -

Example

- -

@CODE

- -

dbopen("test","root","mypassword");

- -

dballocstmt();

- -

dbexecstmt("INSERT INTO employee (name, age) VALUES('John Smith','32');");

- -

dbfreestmt();

- -

dbclose();

- -

@@CODE

- -

 

- -

See Also

- -

Table of Database Functions

- -

Database1 Sample Analyzer

- -

 

- - - - diff --git a/Help/helps/dbopen.htm b/Help/helps/dbopen.htm deleted file mode 100644 index fde33af..0000000 --- a/Help/helps/dbopen.htm +++ /dev/null @@ -1,262 +0,0 @@ - - - - - - - - - - -dbopen - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dbopen

- -

Purpose

- -

Open a database.

- -

Syntax

- -

return_bool = dbopen(db_str, user_str, password_str)

- -

return_bool - True if successful, else false

- -

db_str - Name of the database, as known to the ODBC manager

- -

user_str - Name of database account (may be optional)

- -

password_str - Password for database account (may be optional)
- 

- -

Returns

- -

True if successful.

- -

Remarks

- -

Open a database connection.  NLP++ - currently allows only one database to be open at a time.

- -

Example

- -

@CODE

- -

dbopen("test","root","mypassword");

- -

dballocstmt();

- -

dbexecstmt("SELECT * FROM table;");

- -

dbbindcol(1,"varchar",50,&G("employee name"),&G("result1"));

- -

while (dbfetch())

- -

   {

- -

   "output.txt" << "employee name: ";

- -

   if (G("result1"))

- -

      "output.txt" << G("employee name") << "\n";

- -

   else

- -

      "output.txt" << "NULL" << "\n";

- -

   }

- -

dbfreestmt();

- -

dbclose();

- -

@@CODE

- -

 

- -

See Also

- -

Table of Database Functions

- -

Database1 Analyzer

- -

 

- -

 

- - - - diff --git a/Help/helps/deaccent.htm b/Help/helps/deaccent.htm deleted file mode 100644 index 6a3be62..0000000 --- a/Help/helps/deaccent.htm +++ /dev/null @@ -1,80 +0,0 @@ - - - - - - - - - - -deaccent - - - - - - - - - - - - - - - - - - - - - - - - - - - -

deaccent

- -

Purpose

- -

Convert string to remove accented characters.

- -

Syntax

- -

returnedString = deaccent(string)

- -

returnedString - type: str
-string - type: str

- -

Returns

- -

String with accented characters converted to non-accented variants.

- -

Remarks

- -

Example

- -

See Also

- -

String - Functions 

- - - - diff --git a/Help/helps/debug.htm b/Help/helps/debug.htm deleted file mode 100644 index 69825c2..0000000 --- a/Help/helps/debug.htm +++ /dev/null @@ -1,83 +0,0 @@ - - - - - - - - - - -debug - - - - - - - - - - - - - - - - - - - - - - - - - - - -

debug

- -

Purpose

- -

Enable C++ source code debugging for VisualText developers.

- -

Syntax

- -

NULL = debug()

- -

Returns

- -

Remarks

- -

Enables VisualText developers to set a breakpoint in a particular rule, - from a C++ development environment, e.g. MS Visual C++.

- -

Example

- -

@POST
-debug();
-@RULES
-_xNIL <- _xNIL @@

- -

See Also

- -

Special Functions

- -

 

- - - - diff --git a/Help/helps/default.css b/Help/helps/default.css deleted file mode 100644 index 70966b9..0000000 --- a/Help/helps/default.css +++ /dev/null @@ -1,205 +0,0 @@ -/*=(===============================================================)*/ - -/*=(===============================================================)*/ -/*=(Created with RoboEditor.)=======================================*/ -/*=(===============================================================)*/ - -BODY { - font-family: Arial; - background-color: #ffffff; -} - -H1 { - font-family: Arial; - font-size: 18.0pt; - color: #008080; -} - -/*=(Generated Code)=================================================*/ -/*=(WARNING: DO NOT EDIT OR DELETE THIS SECTION!)===================*/ -/*begin!kadov{{=====================================================*/ -LI.kadov-H1 { - font-family: Arial; - font-size: 18.0pt; - color: #008080; -} -/*}}end!kadov=======================================================*/ - - - -H2 { - font-family: Arial; - font-weight: normal; - font-size: 14.0pt; - color: #008080; -} - -/*=(Generated Code)=================================================*/ -/*=(WARNING: DO NOT EDIT OR DELETE THIS SECTION!)===================*/ -/*begin!kadov{{=====================================================*/ -LI.kadov-H2 { - font-family: Arial; - font-weight: normal; - font-size: 14.0pt; - color: #008080; -} -/*}}end!kadov=======================================================*/ - - - -H3 { - font-weight: bold; - font-size: 14.0pt; -} - -/*=(Generated Code)=================================================*/ -/*=(WARNING: DO NOT EDIT OR DELETE THIS SECTION!)===================*/ -/*begin!kadov{{=====================================================*/ -LI.kadov-H3 { - font-weight: bold; - font-size: 14.0pt; -} -/*}}end!kadov=======================================================*/ - - - -H4 { - font-weight: bold; - font-size: 12.0pt; -} - -/*=(Generated Code)=================================================*/ -/*=(WARNING: DO NOT EDIT OR DELETE THIS SECTION!)===================*/ -/*begin!kadov{{=====================================================*/ -LI.kadov-H4 { - font-weight: bold; - font-size: 12.0pt; -} -/*}}end!kadov=======================================================*/ - - - -H5 { - font-size: 12.0pt; -} - -/*=(Generated Code)=================================================*/ -/*=(WARNING: DO NOT EDIT OR DELETE THIS SECTION!)===================*/ -/*begin!kadov{{=====================================================*/ -LI.kadov-H5 { - font-size: 12.0pt; -} -/*}}end!kadov=======================================================*/ - - - -H6 { - font-size: 12.0pt; -} - -/*=(Generated Code)=================================================*/ -/*=(WARNING: DO NOT EDIT OR DELETE THIS SECTION!)===================*/ -/*begin!kadov{{=====================================================*/ -LI.kadov-H6 { - font-size: 12.0pt; -} -/*}}end!kadov=======================================================*/ - - - -P { - font-size: 10.0pt; -} - -/*=(Generated Code)=================================================*/ -/*=(WARNING: DO NOT EDIT OR DELETE THIS SECTION!)===================*/ -/*begin!kadov{{=====================================================*/ -LI.kadov-P { - font-size: 10.0pt; -} -/*}}end!kadov=======================================================*/ - - - -P.Description { - font-family: Arial; - font-size: 8.0pt; - text-transform: none; - color: #000000; - vertical-align: baseline; - word-spacing: normal; - letter-spacing: normal; -} - -/*=(Generated Code)=================================================*/ -/*=(WARNING: DO NOT EDIT OR DELETE THIS SECTION!)===================*/ -/*begin!kadov{{=====================================================*/ -LI.kadov-P-CDescription { - font-family: Arial; - font-size: 8.0pt; - text-transform: none; - color: #000000; - word-spacing: normal; - letter-spacing: normal; -} -/*}}end!kadov=======================================================*/ - - - -P.color-bullets { - list-style-type: disc; - color: #0000ff; -} - -/*=(Generated Code)=================================================*/ -/*=(WARNING: DO NOT EDIT OR DELETE THIS SECTION!)===================*/ -/*begin!kadov{{=====================================================*/ -LI.kadov-P-Ccolor-bullets { - list-style-type: disc; - color: #0000ff; -} -/*}}end!kadov=======================================================*/ - - - -A.expandspot { - cursor: hand; - color: #008000; - font-style: normal; -} - -SPAN.expandtext { - color: #ff0000; - font-weight: normal; - font-style: italic; -} - -A.dropspot { - cursor: hand; - font-size: 10pt; - font-weight: bold; - font-style: normal; - color: #0000ff; -} - -A.glossterm { - cursor: hand; - color: #800000; - font-style: normal; -} - -SPAN.glosstext { - color: #0000ff; - font-weight: normal; - font-style: italic; -} - -DIV.droptext { - font-size: 8pt; - color: #000000; - margin-left: 0.25in; - margin-right: 0in; -} - - -/*=(===============================================================)*/ diff --git a/Help/helps/dejunk.htm b/Help/helps/dejunk.htm deleted file mode 100644 index 0a49860..0000000 --- a/Help/helps/dejunk.htm +++ /dev/null @@ -1,86 +0,0 @@ - - - - - - - - - - -dejunk - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dejunk

- -

Purpose

- -

Convert string to remove accented and extended ASCII/ANSI characters.

- -

Syntax

- -

returnedString = dejunk(string)

- -

returnedString - type: str
-string - type: str

- -

Returns

- -

String with characters converted to non-accented variants or to question - marks if no ASCII/ANSI equivalent is readily apparent.  This - function helps clean up "junk" for things like XML readers that - break on control characters.  Some - chars replaced with question mark ('?'), quote chars, and so on.

- -

Remarks

- -

Example

- -

See Also

- -

String - Functions 

- -

 

- - - - diff --git a/Help/helps/dictfindword.htm b/Help/helps/dictfindword.htm deleted file mode 100644 index ad31bdd..0000000 --- a/Help/helps/dictfindword.htm +++ /dev/null @@ -1,121 +0,0 @@ - - - - - - - - - - -dictfindword - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dictfindword

- -

Purpose

- -

Find dictionary concept wordString - in the dictionary hierarchy of KB, return handle to concept.

- -

Syntax

- -

returnedConcept = dictfindword(wordString)

- -

returnedConcept - type: con
-wordString - type: str

- -

Returns

- -

Handle to the word concept if wordString - is found in the dictionary, NULL otherwise.

- -

Remarks

- -

The dictionary is a specialized part of the KB hierarchy used to store - and lookup word forms. You can see it with the KB editor, under concept > sys > dict.

- -

Example

- -

@CODE
-"output.txt" << "1 " << conceptname(addword("hello")) << "\n";
-"output.txt" << "2 " << conceptname(wordindex("hello")) << "\n";
-"output.txt" << "3 " << findwordpath("hello") << "\n";
-"output.txt" << "4 " << findwordpath("olleh") << "\n";
-"output.txt" << "5 " << wordpath("foobaz") << "\n";
-"output.txt" << "6 " << conceptname(dictfindword("hello")) << \n";
-rmword("foobaz");

- -

Prints out:

- -

1 hello
-2 he
-3 "concept" "sys" "dict" "a" "h" "he" "hello"
-4
-5 "concept" "sys" "dict" "a" "f" "fo" "foobaz"
-6 hello

- -

See Also

- -

findwordpath, wordindex, - wordpath, Knowledge - Base Functions

- -

 

- -

 

- - - - diff --git a/Help/helps/dictfirst.htm b/Help/helps/dictfirst.htm deleted file mode 100644 index ff769d6..0000000 --- a/Help/helps/dictfirst.htm +++ /dev/null @@ -1,146 +0,0 @@ - - - - - - - - - - -dictfirst - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dictfirst

- -

Purpose

- -

Fetch the first word in the KB dictionary hierarchy.

- -

Syntax

- -

wordConcept = dictfirst()

- -

wordConcept - type: con

- -

Returns

- -

The first word-concept in lexicographic traversal of the KB dictionary - hierarchy.

- -

Remarks

- -

The dictionary is a specialized part of the KB hierarchy used to store - and lookup words.  dictfirst - and dictnext provide a convenient way to traverse the dictionary without - worrying about the details of its implementation.  (I.e., - the presence of layers of dictionary indexing concepts.)

- -

Example

- -

@CODE

- -

L("con") = dictfirst();

- -

while (L("con"))

- -

  {

- -

  "output.txt" << conceptname(L("con") << "\n";

- -

  L("con") = dictnext(L("con"));

- -

  }

- -

Prints a list of the dictionary entries, one per line.

- -

See Also

- -

dictnext, Knowledge - Base Functions

- -

 

- - - - diff --git a/Help/helps/dictgetword.htm b/Help/helps/dictgetword.htm deleted file mode 100644 index 2e514e3..0000000 --- a/Help/helps/dictgetword.htm +++ /dev/null @@ -1,97 +0,0 @@ - - - - - - - - - - -dictgetword - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dictgetword

- -

Purpose

- -

Find else add dictionary concept wordString - to the KB dictionary hierarchy.  Return - the concept.

- -

Syntax

- -

returnedWordConcept = dictgetword(wordString)

- -

returnedWordConcept - type: con
-wordString - type: str

- -

Returns

- -

The newly created word concept. If the word cannot be added, returns - 0.

- -

Remarks

- -

The dictionary is a specialized part of the KB hierarchy used to store - and lookup words.

- -

Example

- -

See addword.

- -

See Also

- -

addword, wordindex, - wordpath, rmword, - Knowledge Base Functions

- -

 

- - - - diff --git a/Help/helps/dictnext.htm b/Help/helps/dictnext.htm deleted file mode 100644 index a0da524..0000000 --- a/Help/helps/dictnext.htm +++ /dev/null @@ -1,153 +0,0 @@ - - - - - - - - - - -dictnext - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dictnext

- -

Purpose

- -

Fetch the word after the given word in the KB dictionary hierarchy.

- -

Syntax

- -

wordConcept = dictnext(currwordConcept)

- -

wordConcept - type: con

- -

currwordConcept - type: con

- -

Returns

- -

The next word-concept in lexicographic traversal of the KB dictionary - hierarchy.

- -

Remarks

- -

The dictionary is a specialized part of the KB hierarchy used to store - and lookup words.  dictfirst - and dictnext provide a convenient way to traverse the dictionary without - worrying about the details of its implementation.  (I.e., - the presence of layers of dictionary indexing concepts.)

- -

Example

- -

@CODE

- -

L("con") = dictfirst();

- -

while (L("con"))

- -

  {

- -

  "output.txt" << conceptname(L("con") << "\n";

- -

  L("con") = dictnext(L("con"));

- -

  }

- -

Prints a list of the dictionary entries, one per line.

- -

See Also

- -

dictfirst, Knowledge - Base Functions

- - - - diff --git a/Help/helps/down.htm b/Help/helps/down.htm deleted file mode 100644 index ca4446e..0000000 --- a/Help/helps/down.htm +++ /dev/null @@ -1,138 +0,0 @@ - - - - - - - - - - -down - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

down

- -

Purpose

- -

Fetch first child of given concept.

- -

Syntax

- -

returnedConcept = down(currentConcept)

- -

returnedConcept - type: con
-currentConcept - type: con

- -

Returns

- -

First child concept.

- -

Remarks

- -

If currentConcept is a bad concept, a warning is written to output log. - If there is no child of currentConcept, a warning is written to the output - log.

- -

Example

- -

- -

Run the following code using the KB shown above:

- -

@CODE
-# Find 'mother' in the KB
-G("First") = findhierconcept("mother", findroot());
-# goto parent concept (female)
-G("Second") = up(G("First"));
-"output.txt" << conceptname(G("Second")) << "\n";
-# go back down to 'mother'
-G("First") = down(G("Second"));
-"output.txt" << conceptname(G("First")) << "\n";
-# find mother's first sibling, which should be 'daughter'
-G("Second") = next(G("First"));
-"output.txt" << conceptname(G("Second")) << "\n";
-# find daughter's previous sibling, which should be mother
-G("First") = prev(G("Second"));
-"output.txt" << conceptname(G("First")) << "\n";

- -

The output should look like:

- -

female
-mother
-daughter
-mother

- -

See Also

- -

up, prev, next, - Knowledge Base Functions

- -

 

- - - - diff --git a/Help/helps/eHelp.xml b/Help/helps/eHelp.xml deleted file mode 100644 index 53310af..0000000 --- a/Help/helps/eHelp.xml +++ /dev/null @@ -1,10 +0,0 @@ - - - - - - - - WebSearch - - \ No newline at end of file diff --git a/Help/helps/ehlpdhtm.js b/Help/helps/ehlpdhtm.js deleted file mode 100644 index cb52aac..0000000 --- a/Help/helps/ehlpdhtm.js +++ /dev/null @@ -1,4033 +0,0 @@ -// eHelp Corporation Dynamic HTML JavaScript -// Copyright 1998-2001 eHelp Corporation.All rights reserved. -// Version=4.57 - -// Warning:Do not modify this file.It is generated by RoboHELP and changes will be overwritten. - -//// Segment Begin -- (JavaScript 1.0) - -/// Section Begin - General and relative topics(JavaScript 1.0) - -//{{HH_SYMBOL_SECTION -var HH_ChmFilename = "D:\\dev\\Help\\Help.chm"; -var HH_WindowName = "NewWindow||VisualText"; -var HH_GlossaryFont = ""; -var HH_Glossary = "0,1"; -var HH_Avenue = "0,0"; -var HH_ActiveX = true; -//}}HH_SYMBOL_SECTION - -//Begin to support previous generic parameters -//Get the information about the browser. -var gstrBsAgent = navigator.userAgent.toLowerCase(); -var gnBsVer = parseInt(navigator.appVersion); - -var gbBsIE = (gstrBsAgent.indexOf('msie') != -1); -var gbBsNS = (gstrBsAgent.indexOf('mozilla') != -1) && ((gstrBsAgent.indexOf('spoofer') == -1) && (gstrBsAgent.indexOf('compatible') == -1)); -var gbBsOpera = (gstrBsAgent.indexOf('opera') != -1); - -var gbBsIE3Before = ((gbBsIE) && (gnBsVer <= 2)); -var gbBsNS3Before = ((gbBsNS) && (gnBsVer <= 3)); - -var gbBsNS2 = ((gbBsNS) && (gnBsVer <= 2)); -var gbBsNS3 = ((gbBsNS) && (gnBsVer == 3)); -var gbBsIE300301 = ((gbBsIE) && (gnBsVer == 2) && ((gstrBsAgent.indexOf("3.00") != -1)||(gstrBsAgent.indexOf("3.0a") != -1)||(gstrBsAgent.indexOf("3.0b")!=-1)||(gstrBsAgent.indexOf("3.01")!=-1))); -var gbBsIE302 = ((gbBsIE) && (gnBsVer == 2) && (gstrBsAgent.indexOf("3.02") != -1)); - -var gbBsNS4 = ((gbBsNS) && (gnBsVer >= 4)); -var gbBsNS6 = ((gbBsNS) && (gnBsVer >= 5)); - -var gbBsIE4 = ((gbBsIE) && (gnBsVer >= 4)); -var gbBsIE5 = false; -var gbBsIE55 = false; - -gbBsIE = (navigator.appName.indexOf("Microsoft") != -1); - -if (parseInt(navigator.appVersion) >= 4) { - gbBsIE4 = (navigator.appName.indexOf("Microsoft") != -1); - - if (gbBsIE4) { - var nPos = gstrBsAgent.indexOf("msie"); - var strIEversion = gstrBsAgent.substring(nPos + 5); - var nVersion = parseFloat(strIEversion); - if (nVersion >= 5) - gbBsIE5 = true; - if (nVersion >= 5.5) - gbBsIE55 = true; - } -} - -var gbBsMac = (gstrBsAgent.indexOf('mac') != -1); -var gbBsWindows = ((gstrBsAgent.indexOf('win') != -1) || (gstrBsAgent.indexOf('16bit') != -1)); -var gbBsOp3 = (gstrBsAgent.indexOf('opera') != -1); - -// Utilities functions. -function BsscHasExtJs() -{ - if( gbBsIE3Before || gbBsNS3Before) - return false; - return true; -} - -// Register event handler -var gBsOnLoads = new Array(); // An array holds all the onload event handler. -var gBsOnClicks = new Array(); // An array holds all the onClick event handler. -var gBsOnUnLoads = new Array(); // An array holds all the OnUnLoad event handler. -var gBsOnMouseOvers = new Array(); // An array holds all the OnMouseOver event handler. -var gBsOnMouseOuts = new Array(); // An array holds all the OnMouseOut event handler. - -var gbOrignalOnMouseDown = null; - -function BsscRegisterOnLoad(funcHandler) -{ - var nLength = gBsOnLoads.length; - gBsOnLoads[nLength] = funcHandler; -} - -function BsscRegisterOnClick(funcHandler) -{ - var nLength = gBsOnClicks.length; - gBsOnClicks[nLength] = funcHandler; -} - -function BsscRegisterOnUnLoad(funcHandler) -{ - var nLength = gBsOnUnLoads.length; - gBsOnUnLoads[nLength] = funcHandler; -} - -function BsscRegisterOnMouseOver(funcHandler) -{ - var nLength = gBsOnMouseOvers.length; - gBsOnMouseOvers[nLength] = funcHandler; -} - -function BsscRegisterOnMouseOut(funcHandler) -{ - var nLength = gBsOnMouseOuts.length; - gBsOnMouseOuts[nLength] = funcHandler; -} - - -function BsGeneralOnLoad() -{ - if (!gbBsIE4 && !gbBsNS4) - return; - - // Make everything visible in navigator - if (gbBsNS4 && !gbBsNS6) { - // Make some special effects items visible - for (var iLayer = 0; iLayer < document.layers.length; iLayer++) { - document.layers[iLayer].visibility = "show"; - document.layers[iLayer].left = 0; - } - } -} - -// If resize the netscape browser, need to reload it. -function BsReDo() -{ - if (innerWidth != origWidth || innerHeight != origHeight) - location.reload(); -} -// End of the local functions. - -// The following functions are used by the html files. -function BSSCOnLoad() -{ - if( !BsscHasExtJs() ) - return; - for (var nElement = gBsOnLoads.length - 1; nElement >= 0; nElement--) - gBsOnLoads[nElement](); -} - -function BSSCOnClick() -{ - if (!BsscHasExtJs()) return; - - for (var nElement = gBsOnClicks.length - 1; nElement >= 0; nElement--) - gBsOnClicks[nElement](); -} - -function BSSCOnUnload() -{ - if (!BsscHasExtJs()) return; - for (var nElement = gBsOnUnLoads.length - 1; nElement >= 0; nElement--) - { - gBsOnUnLoads[nElement](); - } -} - -function BSSCOnMouseOver() -{ - if (!BsscHasExtJs()) return; - for (var nElement = gBsOnMouseOvers.length - 1; nElement >= 0; nElement--) - { - gBsOnMouseOvers[nElement](); - } -} - -function BSSCOnMouseOut() -{ - if (!BsscHasExtJs()) return; - for (var nElement = gBsOnMouseOuts.length - 1; nElement >= 0; nElement--) - { - gBsOnMouseOuts[nElement](); - } -} -// End of invocation of the event handle functions. - - -// Add the GereralOnLoad to the onload array. -if (typeof(BsscRegisterOnLoad) != "undefined") -{ - BsscRegisterOnLoad(BsGeneralOnLoad); -} -if (gbBsNS4&&!gbBsNS6) { - origWidth = innerWidth; - origHeight = innerHeight; - onresize = BsReDo; -} -//End to support previous generic parameters - -//Begin to support previous HHActiveX invoking -function BsHHActivateComponents() -{ - if( HH_ActiveX && (HH_ChmFilename != "") && ((self == top) || (self == top.frames[0]))) - { - var objBody = document.all.tags("BODY")[0]; - if( typeof(objBody) == "object" ) - { - objBody.insertAdjacentHTML("beforeEnd", ''); - if (HHComponentActivator.object) - HHComponentActivator.Activate(HH_ChmFilename, HH_WindowName, HH_GlossaryFont, HH_Glossary, HH_Avenue); - } - } -} - -function BsHHActivXOnLoad() -{ - if( gbBsIE4 ) - BsHHActivateComponents(); -} - -if( typeof(BsscRegisterOnLoad) != "undefined" ) -{ - BsscRegisterOnLoad(BsHHActivXOnLoad); -} -//End to support previous HHActiveX invoking - -//Begin to support previous relative topics -//If webHelp needs Related Topics DHTMLcode, it's supposed to add it here -var gbPopupMenuTimeoutExpired = false; -var gbInPopupMenu = false; -var gbPopupMenuTopicList = null; -var gOlddocumentClick = null; - -////////////////////////////////////////////////////////////////////////////////////////// -// -// Popup Menu code -// -////////////////////////////////////////////////////////////////////////////////////////// - -var g_bIsPopupMenuInit = false; -function _WritePopupMenuLayer() -{ - if (!g_bIsPopupMenuInit) - { - if (gbBsNS4) { -//Do not try to write ininle styles for NS! NS can not handle it and will not stop downloading the html page... - if (gbBsNS6) - document.write(""); - else - document.write("
"); - } else{ - document.write(""); - if (gbBsIE4) { - document.write(""); - } - } - g_bIsPopupMenuInit = true; - } -} - -//Define variable arguments as: strTitle, strUrl -function MTE() -{ - this.strTitle = MTE.arguments[0]; - if (MTE.arguments.length > 1) - this.strURL = MTE.arguments[1]; - else - this.strURL = this.strTitle; -} - -// If the topic list is set, it is an array of TopicEntry objects (defined in WebHelp3.js) -function PopupMenu_SetTopicList(aPopupTopicArray) -{ - gbPopupMenuTopicList = aPopupTopicArray; -} - - -//Seek for the bsscright frame -function _SeekFrameByName( cRoot, strName ) -{ - if( cRoot == null ) return null; - if( cRoot.frames == null ) return null; - if( cRoot.frames[strName] != null ) return cRoot.frames[strName]; - for (var i=0; i"); - wndPopupLinks.document.write(""); - var strParaLine = ""; - for (var i = 0; i < (argLen - 2) / 2; i++) { - strParaLine = ""; - strParaLine += ""); - wndPopupLinks.document.close(); - window.gbInPopupMenu = true; - if (!gbBsIE) { - wndPopupLinks.focus(); - } - - return false; - } - - - if (((argLen < 5) && ((isNaN(fn_arguments[2])) || (gbPopupMenuTopicList == null))) || - ((argLen < 4) && ((!isNaN(fn_arguments[2])) && (gbPopupMenuTopicList != null)))) { - // Get the place that we will be putting the topic into - var strURL = ""; - if (isNaN(fn_arguments[2]) || (gbPopupMenuTopicList == null)) { - strURL = fn_arguments[3]; - } - else { - strURL = gbPopupMenuTopicList[fn_arguments[2]].strURL; - } - - if (targetDoc != null) { - targetDoc.location.href = strURL; - } - else { - if (fn_arguments[1] != null && typeof(fn_arguments[1]) != "undefined") - window.open(strURL, fn_arguments[1]); - else - window.open(strURL); - } - window.gbInPopupMenu = true; - return false; - } - - var strMenu = ""; - if (gbBsNS4) { - strMenu = ''; - } else { - strMenu = '
'; - } else { - strMenu += '' + gbPopupMenuTopicList[fn_arguments[i]].strTitle + ''; - } - strMenu += ''; - - if (isNaN(fn_arguments[i]) || (gbPopupMenuTopicList == null)) { - i += 2; - } else { - i += 1; - } - } - strMenu += "
"; - - if (gbBsMac) { - // totally hack. because ie5 in mac need something. is one of them. mac is mad. - strMenu +="
"; - } - - var layerPopup = null; - var stylePopup = null; - var nEventX = 0; - var nEventY = 0; - var nWindowWidth = 0; - if (gbBsIE4) { - - layerPopup = document.all["PopupMenu"]; - layerPopup.innerHTML = strMenu; - stylePopup = layerPopup.style; - - _BSPSGetClientSize(); - - // Get the position of the item causing the event (relative to its parent) - nEventX = window.event.clientX; - nEventY = window.event.clientY; - - if (nEventY + layerPopup.scrollHeight + 10 < gBsClientHeight) { - nEventY += document.body.scrollTop + 10; - } else { - nEventY = (document.body.scrollTop + gBsClientHeight) - layerPopup.scrollHeight - 20; - } - stylePopup.top = nEventY; - - var nPopupWidth = layerPopup.scrollWidth; - if (gbBsMac) { - nPopupWidth = 80; // we have no idea how to get the dynamic width of the popup. - } - if (nEventX + nPopupWidth + 20 > gBsClientWidth) { - if (gBsClientWidth - nPopupWidth < 5) { - stylePopup.left = 5; - } else { - stylePopup.left = gBsClientWidth - nPopupWidth - 5; - } - } else { - stylePopup.left = nEventX + document.body.scrollLeft + 20; - } - - stylePopup.visibility = "visible"; - if (!gOlddocumentClick && document.onclick) - gOlddocumentClick = document.onclick; - document.onclick = PopupMenu_HandleClick; - - } else if (gbBsNS6) { - layerPopup = document.getElementById("PopupMenu"); - layerPopup.style.visibility = "hidden"; - - var e = fn_arguments[0]; - nEventX = e.pageX; - nEventY = e.pageY; - _BSPSGetClientSize(); - layerPopup.innerHTML = strMenu; - - if (nEventY + layerPopup.offsetHeight + 20 < window.pageYOffset + gBsClientHeight) { - nEventY += 20; - } else { - nEventY = gBsClientHeight + window.pageYOffset - layerPopup.offsetHeight - 20; - } - - if (nEventX + layerPopup.offsetWidth + 20 > gBsClientWidth + window.pageXOffset) { - if (gBsClientWidth + window.pageXOffset - layerPopup.offsetWidth < 20) { - nEventX = 5; - } else { - nEventX = gBsClientWidth + window.pageXOffset - layerPopup.offsetWidth - 20; - } - } else { - nEventX += 20; - } - layerPopup.style.top = nEventY; - layerPopup.style.left = nEventX; - // set again to avoid the stupid frash in netscape 6. - layerPopup.innerHTML = strMenu; - layerPopup.style.visibility = "visible"; - //window.captureEvents(Event.MOUSEDOWN); - if (!gOlddocumentClick && document.onclick) - gOlddocumentClick = document.onclick; - window.onclick = PopupMenu_HandleClick; - } - else if (gbBsNS4) { - layerPopup = document.layers.PopupMenu; - layerPopup.visibility = "hide"; - stylePopup = layerPopup.document; - stylePopup.write(strMenu); - stylePopup.close(); - var e = fn_arguments[0]; - nEventX = e.pageX; - nEventY = e.pageY; - _BSPSGetClientSize(); - if (nEventY + layerPopup.clip.height + 20 < window.pageYOffset + gBsClientHeight) { - nEventY += 20; - } else { - nEventY = gBsClientHeight + window.pageYOffset- layerPopup.clip.height - 20; - } - layerPopup.top = nEventY; - - if (nEventX + layerPopup.clip.width + 20 > gBsClientWidth + window.pageXOffset) { - if (gBsClientWidth + window.pageXOffset - layerPopup.clip.width < 20) { - nEventX = 5; - } else { - nEventX = gBsClientWidth + window.pageXOffset - layerPopup.clip.width - 20; - } - } else { - nEventX += 20; - } - - layerPopup.left = nEventX; - - layerPopup.visibility = "show"; - - window.captureEvents(Event.MOUSEDOWN); - if (!gOlddocumentClick && document.onmousedown) - gOlddocumentClick = document.onmousedown; - window.onmousedown = PopupMenu_HandleClick; - } - - window.gbInPopupMenu = true; - window.gbPopupMenuTimeoutExpired = false; - setTimeout("PopupMenu_Timeout();", 100); - return false; -} - - -function PopupMenu_Timeout() -{ - window.gbPopupMenuTimeoutExpired = true; -} - -function PopupMenu_Over(e) -{ - if (gbBsIE4) { - e.srcElement.className = "PopupOver"; - } else if (gbBsNS4) { -// this.bgColor = "red"; -// e.target.document.className = "PopupOver"; - } - return; -} - -function PopupMenu_Out(e) -{ - if (gbBsIE4) { - e.srcElement.className = "PopupNotOver"; - } else if (gbBsNS4) { - this.bgColor = "#f0f0f0"; - } - return; -} - - -function PopupMenu_HandleClick(e) -{ - if (window.gbPopupMenuTimeoutExpired) { - - window.gbInPopupMenu = false; - - if (gbBsNS4 && !gbBsNS6) { - window.releaseEvents(Event.MOUSEDOWN); - } - - var layerPopup = null; - var stylePopup = null; - if (gbBsIE4) { - layerPopup = document.all["PopupMenu"]; - stylePopup = layerPopup.style; - stylePopup.visibility = "hidden"; - } else if (gbBsNS6) { - layerPopup = document.getElementById("PopupMenu"); - layerPopup.style.visibility = "hidden"; - } else if (gbBsNS4) { - layerPopup = document.layers.PopupMenu; - layerPopup.visibility = "hide"; - } - - if (gOlddocumentClick) - { - if (gbBsNS4 && !gbBsNS6) - document.onmousedown = gOlddocumentClick; - else - document.onclick = gOlddocumentClick; - } - } - return; -} - -// This function should be deleted when all old projects are cleaned up -function BSPSWritePopupFrameForIE4() -{ - return false; -} - -///////////////////////////////////////////////////////////////////// -function BSSCPopup_ClickMac() -{ - if ((!DHTMLPopupSupport()) && (gbBsIE4)) - { - var bClickOnAnchor = false; - var el; - if ((window.event != null) && - (window.event.srcElement != null)) - { - el = window.event.srcElement; - while (el != null) - { - if ((el.tagName == "A") || (el.tagName == "AREA")) { - bClickOnAnchor = true; - break; - } - if (el.tagName == "BODY") { - break; - } - el = el.parentElement; - } - } - if (BSSCPopup_IsPopup()) - { - if (!bClickOnAnchor) { - parent.window.gPopupWindow = null; - self.close(); - } - } - else - { - bClosePopupWindow = true; - if ((bClickOnAnchor) && - (el.href) && - ((el.href.indexOf("javascript:BSSCPopup") != -1) || (el.href.indexOf("javascript:null") != -1) || (el.href.indexOf("javascript:void(0)") != -1))) - { - bClosePopupWindow = false; - } - if (bClosePopupWindow) - { - if (window.gPopupWindow != null && !window.gPopupWindow.closed ) - { - window.gPopupWindow.close(); - } - } - } - } - } - -////////////////////////////////////////////////////////////////////// - -function BsPopupOnClick() -{ - if (!gbBsIE4) - return; - - BSSCPopup_ClickMac(); -} - -function _BSSCOnError(message) -{ - if(-1 != message.indexOf("denied") - || -1 != message.indexOf("Object required")) - return true; -} - -//End to support previous relative topics - - -/// Section End - General and relative topics (JavaScript 1.0) - -/// Section Begin - Popup (JavaScript 1.0) - -//Begin to support previous popup functions -////////////////////////////////////////////////////////////////////////////////////////////// -// -// Begin DHTML Popup Functions -// -////////////////////////////////////////////////////////////////////////////////////////////// -//variables used to isolate the browser type -var gBsDoc = null; -var gBsSty = null; -var gBsHtm = null; -var gBsStyVisShow = null; -var gBsStyVisHide = null; -var gBsClientWidth = 640; -var gBsClientHeight = 480; -var gBsBrowser = null; - -// here is the varible for judge popup windows size. these parameter is for IE5.0, it may need adjust for others. -var gBRateH_W = 0.618; // 1.618 Golden cut. -var gBMaxXOfParent = 0.8; -var gBMaxYOfParent = 0.8; -var gBscrollHeight = 16; -var gBscrollWidth = 16; -var gBpermitXDelta = 3; -var gBpermitYDelta = 3; - - -var arrayPopupURL = new Array(); -var arrayAbsPopupURL = new Array(); - -var arrayDirty = new Array(); - -function setAbsPopupURL(nIndex, strURL) -{ - arrayAbsPopupURL[nIndex] = strURL; -} - -function getAbsPopupURL(nIndex) -{ - if (nIndex == -1 || arrayAbsPopupURL.length <= nIndex) return null; - else - return arrayAbsPopupURL[nIndex]; -} - -function getPopupURL(nIndex) -{ - if (nIndex == -1 || arrayPopupURL.length <= nIndex) return null; - else - return arrayPopupURL[nIndex]; -} - -function getPopupID(nIndex) -{ - return gstrPopupID + nIndex; -} - -function getPopupShadowID(nIndex) -{ - return gstrPopupShadowID + nIndex; -} - -function getPopupTopicID(nIndex) -{ - return gstrPopupTopicID + nIndex; -} - -function getPopupIFrameID(nIndex) -{ - return gstrPopupIFrameID + nIndex; -} - -function getPopupIFrameName(nIndex) -{ - return gstrPopupIFrameName + nIndex; -} - - -function getPopupTopicStyle(nIndex) -{ - return eval("document.all['" + getPopupTopicID(nIndex) + "']").style; -} - -function getPopupShadowStyle(nIndex) -{ - return eval("document.all['" + getPopupShadowID(nIndex) + "']").style; -} - -function getPopupIFrame(nIndex) -{ - - return eval("document.frames['" + getPopupIFrameName(nIndex) + "']"); -} - -function getPopupDivStyle(nIndex) -{ - return eval("document.all['" + getPopupID(nIndex) + "']").style; -} - -function getPopupIFrameStyle(nIndex) -{ - return eval("document.all['" + getPopupIFrameName(nIndex) + "'].style"); -} - - -function findDiv(strURL) -{ - var i = 0; - for (i = 0; i < arrayPopupURL.length; i ++ ) { - if (arrayPopupURL[i] == strURL) { - return i; - } - } - return -1; -} - -var gnToken = -1; -function takeToken() -{ - gnToken ++; - if (gnToken > 10000) gnToken = 0; - return gnToken; -} - -function IsValidToken(nToken) -{ - return (gnToken == nToken); -} - -function addDiv(strURL) -{ - var i = 0; - for (i = 0; i < arrayPopupURL.length; i ++) { - if (arrayPopupURL[i] == null) { - arrayPopupURL[i] = strURL; - return i; - } - } - arrayPopupURL[i] = strURL; - arrayDirty[i] = true; - return i; -} - -function setDirty() -{ - var i = 0; - for (i = 0; i < arrayPopupURL.length; i ++ ) - arrayDirty[i] = true; -} - -function IsDirty(nIndex) -{ - if (nIndex == -1) - return true; - else - if (arrayDirty.length > nIndex) - return arrayDirty[nIndex]; - else - return true; -} - -function hideAll() -{ - var i = 0; - for (i = 0; i < arrayPopupURL.length; i ++ ) - getPopupDivStyle(i).visibility = gBsStyVisHide; -} - -function getCurrentPopupIFrame() -{ - var i = 0; - for (i = 0; i < arrayPopupURL.length; i ++) - if (getPopupDivStyle(i).visibility == gBsStyVisShow) - return getPopupIFrame(i); - return null; -} - -function setClear(nIndex) -{ - if (nIndex != -1) - arrayDirty[nIndex] = false; -} - -function _BSSCCreatePopupDiv(strURL) -{ - var nIndex = findDiv(strURL); - if (nIndex == -1 ) { - nIndex = addDiv(strURL); - BsPopup_CreateDiv(nIndex); - } - else { - if (IsDirty(nIndex)) { - if("object" == typeof(getPopupIFrame(nIndex).document)) - getPopupIFrame(nIndex).document.location.href = strURL; - } - } - return nIndex; - -} - -//the browser information itself -function _BSPSBrowserItself() -{ - var agent = navigator.userAgent.toLowerCase(); - this.major = parseInt(navigator.appVersion); - this.minor = parseFloat(navigator.appVersion); - this.ns = ((agent.indexOf('mozilla') != -1) && ((agent.indexOf('spoofer') == -1) && (agent.indexOf('compatible') == -1))); - this.ns2 = ((this.ns) && (this.major == 2)); - this.ns3 = ((this.ns) && (this.major == 3)); - this.ns4 = ((this.ns) && (this.major >= 4)); - this.ns6 = ((this.ns) && (this.major >= 5)); - this.ie = (agent.indexOf("msie") != -1); - this.ie3 = ((this.ie) && (this.major == 2)); - this.ie4 = ((this.ie) && (this.major >= 4)); - this.op3 = (agent.indexOf("opera") != -1); - - if (this.ns4) - { - gBsDoc = "document"; - gBsSty = ""; - gBsHtm = ".document"; - gBsStyVisShow = "show"; - gBsStyVisHide = "hide"; - - } - else if (this.ie4) - { - gBsDoc = "document.all"; - gBsSty = ".style"; - gBsHtm = ""; - gBsStyVisShow = "visible"; - gBsStyVisHide = "hidden"; - } -} - -//Here is the browser type -function _BSPSGetBrowserInfo() -{ - gBsBrowser = new _BSPSBrowserItself(); -} - -_BSPSGetBrowserInfo(); - -//Get client size info -function _BSPSGetClientSize() -{ - if (gBsBrowser.ns4) - { - gBsClientWidth = innerWidth; - gBsClientHeight = innerHeight; - - } - else if (gBsBrowser.ie4) - { - gBsClientWidth = document.body.clientWidth; - gBsClientHeight = document.body.clientHeight; - } -} - - -var gstrPopupID = 'BSSCPopup'; -var gstrPopupShadowID = 'BSSCPopupShadow'; -var gstrPopupTopicID = 'BSSCPopupTopic'; -var gstrPopupIFrameID = 'BSSCPopupIFrame'; -var gstrPopupIFrameName = 'BSSCPopupIFrameName'; - -var gstrPopupSecondWindowName = 'BSSCPopup'; - -var gPopupWindow = null; -var gnPopupClickX = 0; -var gnPopupClickY = 0; - -var gnPopupScreenClickX = 0; -var gnPopupScreenClickY = 0; - -var gbPopupTimeoutExpired = false; - - -function DHTMLPopupSupport() -{ - if ((gbBsIE4) && (!gbBsMac)) { - return true; - } - return false; -} - - - -function BSSCPopup_IsPopup() -{ - if (DHTMLPopupSupport() && (this.name.indexOf(gstrPopupIFrameName) != -1)) { - return true; - } else if ((gbBsNS4 || gbBsIE4) && (this.name.indexOf(gstrPopupID) != -1)) { - return true; - } else { - return false; - } -} - - -// If there is a hyperlink in a popup window, display the hyperlink in -// the original window. -if (BSSCPopup_IsPopup() && !gbBsIE4) { - document.write(""); -} - -// Local functions. -function BsPopup_CreateDiv(nIndex) -{ - if(!DHTMLPopupSupport()) - return; - // DO NOT SET Width and height for the div, otherwize it will make IE4 popup do not work when view the topic alone. - var strPopupDiv = ""; - - var objBody = document.all.tags("BODY")[0]; - if( typeof(objBody) != "object" ) - return; - - objBody.insertAdjacentHTML("beforeEnd", strPopupDiv); -} - -function BSSCPopup_Timeout(nIndex, nToken) -{ - if (!IsValidToken(nToken)) return; - - if ((getPopupIFrame(nIndex).document.readyState == "complete") && - (getPopupIFrame(nIndex).document.body != null)) { - window.getPopupDivStyle(nIndex).visibility = gBsStyVisShow; - setClear(nIndex); - window.gbPopupTimeoutExpired = true; - - BSSCPopup_ChangeTargettoParent(getPopupIFrame(nIndex).document); - getPopupIFrame(nIndex).document.body.onclick = BSSCPopupClicked; - - if (!gbOrignalOnMouseDown && document.onmousedown) - gbOrignalOnMouseDown = document.onmousedown; - - document.onmousedown = BSSCPopupParentClicked; - - } else { - setTimeout("BSSCPopup_Timeout(" + nIndex + "," + nToken + ")", 100); - } -} - - -// VH 08/10/00 -// do not change target to parent if the href is using javascript -function BSSCPopup_ChangeTargettoParent(tagsObject) -{ - var collA = tagsObject.all.tags("A"); - BSSCPopup_ChangeTargettoParent2(collA); - - var collIMG = tagsObject.all.tags("IMG"); - BSSCPopup_ChangeTargettoParent2(collIMG); -} - -function BSSCPopup_ChangeTargettoParent2(colls) -{ - var j = 0; - if (colls != null) { - for (j = 0; j < colls.length; j ++ ) - { - var strtemp = colls[j].href; - strtemp = strtemp.toLowerCase(); - if (strtemp.indexOf("javascript:") == -1) - if (colls[j].target == "") - colls[j].target = "_parent"; - } - - } -} - -function BSPSPopupTopicWinHelp(strURL) -{ - _BSSCPopup(strURL); - return; -} - -function _BSSCPopup(strURL, width, height) -{ - var cuswidth = 0; - var cusheight = 0; - if ("undefined" != typeof(width) && "undefined" != typeof(height)) { - cuswidth = width; - cusheight= height; - } - - if (DHTMLPopupSupport()) { - // If we are already in a popup, replace the contents - // if (BSSCPopup_IsPopup()) { - // parent._BSSCPopup(strURL, cuswidth, cusheight); - // } else { - var nToken = takeToken(); // take token first. - var nIndex = _BSSCCreatePopupDiv(strURL); - window.gbPopupTimeoutExpired = false; - var ntWidth = gBsClientWidth; - var ntHeight = gBsClientHeight; - _BSPSGetClientSize(); - if (ntWidth != gBsClientWidth || ntHeight != gBsClientHeight) { - setDirty(); - } - - - if (IsDirty(nIndex)) { - if (gbBsMac) { - setTimeout("BSSCPopup_AfterLoad(" + nIndex + "," + nToken + "," + cuswidth + "," + cusheight +")", 400); - } else { - setTimeout("BSSCPopup_AfterLoad(" + nIndex + "," + nToken + "," + cuswidth + "," + cusheight + ")", 100); - } - } - else { - MoveDivAndShow(nIndex ,nToken, cuswidth, cusheight); - } - // } - - } else { - _BSSCPopup2(strURL, cuswidth, cusheight); - } - return; -} - -if (gbBsIE55) -{ - var ehlpdhtm_fOldBefureUnload = window.onbeforeunload; - window.onbeforeunload = window_BUnload; -} - -function window_BUnload() -{ - for (var i = 0; i < arrayPopupURL.length; i ++) - document.all(getPopupID(i)).outerHTML = ""; - arrayPopupURL.length = 0; - if (ehlpdhtm_fOldBefureUnload) - ehlpdhtm_fOldBefureUnload(); -} - -function _BSSCPopup2(strURL, width, height) -{ - if (window.name == gstrPopupSecondWindowName) { - window.location = strURL; - } else { - if (!gbBsMac || !gBsBrowser.ns4) { - BSSCHidePopupWindow(); - } - var nX = 0; - var nY = 0; - var nHeight = 300; - var nWidth = 400; - if (width > 0 && height > 0) { - nHeight = height; - nWidth = width; - } - _BSPSGetClientSize(); - - nX = window.gnPopupScreenClickX; - nY = window.gnPopupScreenClickY; - - if (nY + nHeight + 40 > screen.availHeight) { - nY = screen.availHeight - nHeight - 40; - } - if (nX + nWidth + 40 > screen.availWidth) { - nX = screen.availWidth - nWidth - 40; - } - // Launch a separate window - var strParam = "titlebar=no,toolbar=no,status=no,location=no,menubar=no,resizable=yes,scrollbars=yes"; - if (gBsBrowser.ns) { - if (gBsBrowser.ns6) { - strParam += ",Height=" + nHeight + ",Width=" + nWidth; - strParam += ",screenX=" + nX + ",screenY=" + nY; - } - else { - strParam += ",OuterHeight=" + nHeight + ",OuterWidth=" + nWidth; - strParam += ",screenX=" + nX + ",screenY=" + nY; - strParam += ",dependent=yes"; - } - } - else { - strParam += ",height=" + nHeight + ",width=" + nWidth; - strParam += ",left=" + nX + ",top=" + nY; - } - window.gPopupWindow = window.open(strURL, gstrPopupSecondWindowName, strParam); - - if (gBsBrowser.ns4) { - window.gPopupWindow.captureEvents(Event.CLICK | Event.BLUE); - window.gPopupWindow.onclick = NonIEPopup_HandleClick; - window.gPopupWindow.onblur = NonIEPopup_HandleBlur; - } - else if (gBsBrowser.ie4) - { - setTimeout("setPopupFocus();", 100); - } - } - return; -} - -function setPopupFocus() -{ - window.gPopupWindow.focus(); -} - -function NonIEPopup_HandleBlur(e) -{ - window.gPopupWindow.focus(); -} - -function NonIEPopup_HandleClick(e) -{ - // Because navigator will give the event to the handler before the hyperlink, let's - // first route the event to see if we are clicking on a Popup menu in a popup. - document.routeEvent(e); - - // If a popup menu is active then don't do anything with the click - if (window.gPopupWindow.gbInPopupMenu) { - window.gPopupWindow.captureEvents(Event.CLICK); - window.gPopupWindow.onclick = NonIEPopup_HandleClick; - return false; - } - - // Close the popup window - if (e.target.href) { - if (e.target.target == "") - window.location.href = e.target.href; - else - window.open(e.target.href, e.target.target); - - if (e.target.href.indexOf("javascript:void(0)") == -1 && e.target.href.indexOf("javascript:null") == -1 && e.target.href.indexOf("BsscPopup") == -1) { - this.close(); - } - } else { - this.close(); - } - return false; - -} - -function BSSCPopup_AfterLoad(nIndex, nToken, cuswidth, cusheight) -{ - if (typeof(window.getPopupIFrame(nIndex).document) == "unknown") { - _BSSCPopup2(getPopupURL(nIndex), cuswidth, cusheight); - return; - } - if (!IsValidToken(nToken)) return; - - if ((window.getPopupIFrame(nIndex).document.readyState == "complete") && - (window.getPopupIFrame(nIndex).document.body != null)) { - if (window.getPopupIFrame(nIndex).document.location.href.indexOf("about:blank") != -1) { // add this check. IE will use about:blank" as the default vaule for Iframe. - window.getPopupIFrame(nIndex).document.location = getPopupURL(nIndex); - setTimeout("BSSCPopup_AfterLoad(" + nIndex + "," + nToken + "," + cuswidth + "," + cusheight + ")", 200); - } - else - { - setAbsPopupURL(nIndex, window.getPopupIFrame(nIndex).document.location.href); // change URL to abs url. - BSSCPopup_ResizeAfterLoad(nIndex, nToken, cuswidth, cusheight); - } - } else { - setTimeout("BSSCPopup_AfterLoad(" + nIndex + "," + nToken + "," + cuswidth + "," + cusheight + ")", 200); - } -} - - -function BSSCPopup_ResizeAfterLoad(nIndex, nToken, cuswidth, cusheight) -{ - if (window.gbPopupTimeoutExpired) return; - - if (!IsValidToken(nToken)) return; - - getPopupDivStyle(nIndex).visibility = gBsStyVisHide; - - // Determine the width and height for the window - //var size = new BSSCSize(0, 0); - //BSSCGetContentSize(window.getPopupIFrame(nIndex), size); - //var nWidth = size.x; - //var nHeight = size.y; - - _BSPSGetClientSize(); - - var size = new BSSCSize(0, 0); - - if (cuswidth <= 0 || cusheight <= 0) - BSSCGetContentSize(window.getPopupIFrame(nIndex), size); - else { - size.x = cuswidth; - size.y = cusheight; - } - - // Determine the width and height for the window - var nWidth = size.x; - var nHeight = size.y; - - // for small popup size, we should allow any size. - // The popup size should be ok if bigger than 0 - if (nWidth < 0 || nHeight < 0) return; // there must be something terribly wrong. - // if (nWidth < 40 || nHeight < 40) return; // there must be something terribly wrong. - - window.getPopupDivStyle(nIndex).pixelWidth = nWidth; - window.getPopupDivStyle(nIndex).pixelHeight = nHeight; - - window.getPopupShadowStyle(nIndex).pixelWidth = nWidth; - window.getPopupShadowStyle(nIndex).pixelHeight = nHeight; - window.getPopupTopicStyle(nIndex).pixelWidth = nWidth; - window.getPopupTopicStyle(nIndex).pixelHeight = nHeight; - if (gbBsIE55) - { - window.getPopupShadowStyle(nIndex).pixelWidth = nWidth + 2; - window.getPopupShadowStyle(nIndex).pixelHeight = nHeight + 2; - window.getPopupTopicStyle(nIndex).pixelWidth = nWidth + 2; - window.getPopupTopicStyle(nIndex).pixelHeight = nHeight + 2; - } - - window.getPopupIFrameStyle(nIndex).pixelWidth = nWidth; - window.getPopupIFrameStyle(nIndex).pixelHeight = nHeight; - if (gbBsIE55) - { - window.getPopupIFrameStyle(nIndex).top = 0; - window.getPopupIFrameStyle(nIndex).left = 0; - } - - var strURL = getPopupURL(nIndex); - if (strURL.indexOf("#") != -1) - getPopupIFrame(nIndex).location.href = strURL; // reload again, this will fix the bookmark misunderstand in IE5. - - MoveDivAndShow(nIndex, nToken, cuswidth, cusheight); -} - -function MoveDivAndShow(nIndex, nToken, cuswidth, cusheight) -{ - if (window.getPopupIFrame(nIndex).document.location.href != getAbsPopupURL(nIndex)) { // if redirect, reload again. - window.getPopupIFrame(nIndex).document.location = getPopupURL(nIndex); - setTimeout("BSSCPopup_AfterLoad(" + nIndex + "," + nToken + "," + cuswidth + "," + cusheight + ")", 200); - return; - } - - // Determine the position of the window - var nClickX = window.gnPopupClickX; - var nClickY = window.gnPopupClickY; - var nTop = 0; - var nLeft = 0; - - var nWidth = window.getPopupDivStyle(nIndex).pixelWidth; - var nHeight = window.getPopupDivStyle(nIndex).pixelHeight; - - if (nClickY + nHeight + 20 < gBsClientHeight + document.body.scrollTop) { - nTop = nClickY + 10; - } else { - nTop = (document.body.scrollTop + gBsClientHeight) - nHeight - 20; - } - if (nClickX + nWidth < gBsClientWidth + document.body.scrollLeft) { - nLeft = nClickX; - } else { - nLeft = (document.body.scrollLeft + gBsClientWidth) - nWidth - 8; - } - - if (nTop < document.body.scrollTop ) nTop = document.body.scrollTop + 1; - if (nLeft< document.body.scrollLeft) nLeft = document.body.scrollLeft + 1; - - - window.getPopupDivStyle(nIndex).left = nLeft; - window.getPopupDivStyle(nIndex).top = nTop; - - // Set the location of the background blocks - window.getPopupShadowStyle(nIndex).left = 6; - window.getPopupShadowStyle(nIndex).top = 6; - if (gbBsIE55) - { - window.getPopupShadowStyle(nIndex).left = 4; - window.getPopupShadowStyle(nIndex).top = 4; - } - - if (gbBsMac) { - // Total hack on the iMac to get the IFrame to position properly - window.getPopupIFrameStyle(nIndex).pixelLeft = 100; - window.getPopupIFrameStyle(nIndex).pixelLeft = 0; - // Explicitly call BSSCOnLoad because the Mac doesn't seem to do it - window.getPopupIFrame(nIndex).window.BSSCOnLoad(); - } - - BSSCPopup_Timeout(nIndex , nToken ); - - return; -} - -function BSSCSize(x, y) -{ - this.x = x; - this.y = y; -} - -function BSSCGetContentSize(thisWindow, size) -{ - if (!((gBsBrowser.ie4) || (gBsBrowser.ns4))) - return; - - if (gbBsMac) { - size.x = 300; - size.y = 300; - return; - } - - // Resize the width until it is wide enough to handle the content - // The trick is to start wide and determine when the scrollHeight changes - // because then we know a scrollbar is necessary. We can then go back - // to the next widest size (for no scrollbar) - - var ClientRate = gBsClientHeight / gBsClientWidth; - - var GoldenSize = new BSSCSize(0,0); - GoldenSize.x = gBsClientWidth * gBMaxXOfParent; - GoldenSize.y = gBsClientHeight *gBMaxYOfParent ; - - if (ClientRate > gBRateH_W) { - GoldenSize.y = GoldenSize.x * gBRateH_W; - } - else { - GoldenSize.x = GoldenSize.y / gBRateH_W; - } - - // Try to using parent specified max x. - var x = 0; - var maxgoldx = GoldenSize.x; - var maxx = gBsClientWidth * gBMaxXOfParent; - - // This double resize causes the document to re-render (and we need it to) - if (!gbBsIE5) - thisWindow.moveTo(10000,10000); // this is used to fix the flash on IE4. - thisWindow.resizeTo(1, 1); - thisWindow.resizeTo(1, 1); - thisWindow.resizeTo(maxgoldx, thisWindow.document.body.scrollHeight + gBscrollHeight); - thisWindow.resizeTo(maxgoldx, thisWindow.document.body.scrollHeight + gBscrollHeight); - - var miny = thisWindow.document.body.scrollHeight + gBscrollHeight; - - if (miny > GoldenSize.y) // the popup does not fix in the parent wanted golden area. so try to expand itself as large as it can - { - thisWindow.resizeTo(maxx , thisWindow.document.body.scrollHeight + gBscrollHeight); - thisWindow.resizeTo(maxx , thisWindow.document.body.scrollHeight + gBscrollHeight); - - miny = thisWindow.document.body.scrollHeight + gBscrollHeight; - maxy = gBsClientHeight * gBMaxYOfParent; - - if (miny > maxy) { // the popup must have a scroll, OK let it be. - miny = maxy; - size.x = maxx; - size.y = maxy; - thisWindow.document.body.scroll = 'yes'; // At this time we do want to show scroll any more. so it will looks better a little. - } - else { // popup still can fit in the parent area by someway. now we choose the same h/w rate as parent. - size.y = miny; - - // downsize from maxx , now I try to using binary divide. - x = maxx; - deltax = -maxx/2; - //j = 0; - while (true) { - x = x + deltax; - thisWindow.resizeTo(x, miny); - thisWindow.resizeTo(x, miny); - diffy = thisWindow.document.body.scrollHeight + gBscrollHeight - x * ClientRate; - if (diffy > gBpermitYDelta ) // it is higher than wanted, so x need to be wide a little bitter - deltax = Math.abs(deltax) /2; - else if (diffy < -gBpermitYDelta) // it is shorter than wanted, so x need to be narrow a little bitter - deltax = -Math.abs(deltax) /2; - else - // the y is close enough to wanted. - break; - if (Math.abs(deltax) < gBpermitXDelta) // the next change is too slight and it can be ignore. - break; - } - size.x = thisWindow.document.body.scrollWidth; //+ gBscrollWidth; - size.y = thisWindow.document.body.scrollHeight;// + gBscrollHeight; - thisWindow.document.body.scroll = 'no'; - - // Handle absurd cases just in case IE flakes - // if (size.y < 100) { - // size.y = 100; - // } - } - } - else { - if (thisWindow.document.body.scrollWidth > maxgoldx) { - size.x = maxx; - size.y = miny; - thisWindow.document.body.scroll = 'yes'; - } - else { - // downsize from maxgoldx , now I try to using binary divide. - x = maxgoldx; - deltax = -maxgoldx/2; - //i = 0; - while (true) { - x = x + deltax; - thisWindow.resizeTo(x, miny); - thisWindow.resizeTo(x, miny); - diffy = thisWindow.document.body.scrollHeight + gBscrollHeight - x * gBRateH_W; - if (diffy > gBpermitYDelta ) // it is higher than wanted, so x need to be wide a little bitter - deltax = Math.abs(deltax) /2; - else if (diffy < -gBpermitYDelta) // it is shorter than wanted, so x need to be narrow a little bitter - deltax = -Math.abs(deltax) /2; - else - // the y is close enough to wanted. - break; - if (Math.abs(deltax) < gBpermitXDelta) // the next change is too slight and it can be ignore. - break; - //i ++; - - } - //size.x = x - gBscrollWidth; - size.x = thisWindow.document.body.scrollWidth; //+ gBscrollWidth; - size.y = thisWindow.document.body.scrollHeight ; - thisWindow.document.body.scroll = 'no'; // At this time we do not want to show scroll any more. so it will looks better a little. - thisWindow.resizeTo(size.x, size.y); - if (thisWindow.document.body.scrollWidth > size.x) - { - size.x = thisWindow.document.body.scrollWidth; - } - if (thisWindow.document.body.scrollHeight > size.y) - { - size.y = thisWindow.document.body.scrollHeight; - } - } - } - - // no reload no scrollbar. - //size.x = size.x + 16; //reserve a width for scrollbar (IE 4.0 only) - - thisWindow.resizeTo(size.x, size.y); - thisWindow.resizeTo(size.x, size.y); - return; -} - - - -function BSSCPopupParentClicked() -{ - if (!window.gbPopupTimeoutExpired) { - return false; - } - - document.onmousedown = gbOrignalOnMouseDown; - - // Simply hide the popup - hideAll(); - - window.gbPopupTimeoutExpired = false; - - return true; -} - - -function BSSCPopupClicked() -{ - if (!window.gbPopupTimeoutExpired) { - return false; - } - - - var popupIFrame = getCurrentPopupIFrame(); - if (popupIFrame == null) { - return true; - } - -/* - if ("undefined" != typeof(popupIFrame.gbInPopupMenu) && - popupIFrame.gbInPopupMenu) { - return true; - }*/ - - if (!((popupIFrame.window.event != null) && - (popupIFrame.window.event.srcElement != null) && - ((popupIFrame.window.event.srcElement.tagName == "A") || - (popupIFrame.window.event.srcElement.tagName == "IMG")))) { - document.onmousedown = gbOrignalOnMouseDown; - - // Simply hide the popup - hideAll(); - - window.gbPopupTimeoutExpired = false; - - return true; - } -} - - -//trace the mouse over's position for hotspot -function BSPSPopupOnMouseOver(event) -{ - if (gBsBrowser.ie4) { - window.gnPopupClickX = event.clientX + document.body.scrollLeft; - window.gnPopupClickY = event.clientY + document.body.scrollTop; - window.gnPopupScreenClickX = event.screenX; - window.gnPopupScreenClickY = event.screenY; - } else if (gBsBrowser.ns4) { - window.gnPopupClickX = event.pageX - window.pageXOffset; - window.gnPopupClickY = event.pageY - window.pageYOffset; - window.gnPopupScreenClickX = event.screenX - window.pageXOffset; - window.gnPopupScreenClickY = event.screenY - window.pageYOffset; - } -} - - -function BSSCHidePopupWindow() -{ - if (window.gPopupWindow != null) { - if (gBsBrowser.ns4) { - if ((typeof window.gPopupWindow != "undefined") && (!window.gPopupWindow.closed)) { - window.gPopupWindow.close(); - window.gPopupWindow = null; - } - } - } - - return; -} - -// Add the PopupOnClick to the onclick array. -if (typeof(BsscRegisterOnClick) != "undefined") -{ - BsscRegisterOnClick(BsPopupOnClick); -} -//End to support previous popup functions - -/// Section End - Popup (JavaScript 1.0) - -/// Section Begin - Embedded Stub (JavaScript 1.0) - -var s_strAgent = navigator.userAgent.toLowerCase(); -var s_nVer = parseInt(navigator.appVersion); - -var s_bIE = (s_strAgent.indexOf('msie') != -1); -var s_bNS = (s_strAgent.indexOf('mozilla') != -1) && ((s_strAgent.indexOf('spoofer') == -1) && (s_strAgent.indexOf('compatible') == -1)); -var s_bOpera = (s_strAgent.indexOf('opera') != -1); - -var s_bIE3Before = ((s_bIE) && (s_nVer <= 2)); -var s_bNS3Before = ((s_bNS) && (s_nVer <= 3)); - -var s_bNS2 = ((s_bNS) && (s_nVer <= 2)); -var s_bNS3 = ((s_bNS) && (s_nVer == 3)); -var s_bIE300301 = ((s_bIE) && (s_nVer == 2) && ((s_strAgent.indexOf("3.00") != -1)||(s_strAgent.indexOf("3.0a") != -1)||(s_strAgent.indexOf("3.0b")!=-1)||(s_strAgent.indexOf("3.01")!=-1))); -var s_bIE302 = ((s_bIE) && (s_nVer == 2) && (s_strAgent.indexOf("3.02") != -1)); - - -function HasExtJs() -{ - if (s_bIE3Before) { return false;} - if (s_bNS3Before) { return false;} - if (typeof (BsGeneralOnLoad) == "undefined"){ return false; } - return true; -} - - -function BSSCCreatePopupDiv() -{ - return; -} - - -function WritePopupMenuLayer() -{ - if (HasExtJs()) {_WritePopupMenuLayer();} -} - -function BSSCPopup(strURL, width, height) -{ - if (HasExtJs()) { - _BSSCPopup(strURL, width, height); - }else{ - //Create a temporary window first to ensure the real popup comes up on top - var wndTemp = null; - if (!s_bNS3) { - wndTemp = window.open("", "temp", "titlebar=no,toolbar=no,status=no,location=no,menubar=no,resizable=yes,scrollbars=yes,height=3,width=4"); - } - // Create the real popup window - var wndPopup = window.open(strURL, "BSSCPopup", "titlebar=no,toolbar=no,status=no,location=no,menubar=no,resizable=yes,scrollbars=yes,height=300,width=400"); - // Close the temporary - if (!s_bNS3) { - wndTemp.close(); - } else { - wndPopup.focus(); - } - } -} - -var gbWndTemp = null, gbWndPopupLinks = null; -var gbstrParaTotal = ""; - -function PopupMenu_Invoke() -{ - if (HasExtJs()) { - return _PopupMenu_Invoke(PopupMenu_Invoke.arguments); - } - if (s_bNS3Before || s_bIE3Before ) { - var argLen = PopupMenu_Invoke.arguments.length; - if (argLen < 5) { - window.document.location.href = PopupMenu_Invoke.arguments[3]; - return false; - } - gbWndTemp = null; - gbWndPopupLinks = null; - gbstrParaTotal = ""; - for (var i = 0; i < (argLen - 2) / 2; i++) { - var strParaLine = ""; - if (s_bNS2 || s_bOpera){ - strParaLine += "" - strParaLine += PopupMenu_Invoke.arguments[2 * i + 2]; - strParaLine += ""; - } else { - strParaLine += ""); - if (s_bNS2 || s_bOpera) { - gbWndPopupLinks.document.write(""); - } else { - //YJ: IE301,302 and NS3.x works fine - gbWndPopupLinks.document.write("<"); - gbWndPopupLinks.document.write("script>"); - gbWndPopupLinks.document.write("function gotoUrl(aUrl) {opener.window.location=aUrl; close();}"); - gbWndPopupLinks.document.write("<"); - gbWndPopupLinks.document.write("/script>"); - } - gbWndPopupLinks.document.write(""); - gbWndPopupLinks.document.write(gbstrParaTotal); - gbWndPopupLinks.document.write(""); - gbWndPopupLinks.document.close(); - - // Close the temporary - if (!s_bNS3 && gbWndTemp != null) { - gbWndTemp.close(); - }else { - gbWndPopupLinks.focus(); - } - - return true; - } - return false; -} - -/// Section End - Embedded Stub (JavaScript 1.0) - -//// Segment End -- (JavaScript 1.0) - -//// Segment Begin -- (JavaScript 1.2) -/// Section Begin - kadov DHTM (JavaScript 1.2) - -//Begin to support extended and dropdown text effects. -function kadovIsParagraph(el) -{ - return( el.tagName == "P" || el.tagName.indexOf("H") == 0 ) ? true : false; -} - -function kadovInitEachChild(el) -{ - for(var i=0; i "") ) - child.style.setAttribute( "x-on-pageload", "" ); - - var href = child.getAttribute("href") - if( href != null && href > "" && href.indexOf( "BSSCPopup" ) >= 0 ) - kadovFilePopupInit(child.id); // Init for Popup - else if( child.className == "dropspot" || child.className == "expandspot" || - child.className == "glossterm" ) - kadovTextPopupInit(child.id);// Init for Expanding/Glossary or DropDown text - else if( child.className == "trigger") - kadovInitTrigger(child.id);// Init for Trigger - else - { - kadovInitEffects(child.id);// Init for DHTML effects - CEngine.SetOneTargetInitialState( child.id ); - } - } - - if( (child.tagName == "IMG") && (child.getAttribute("dynsrc") > "") ) - child.start = "mouseover";// to start a AVI file. fileopen doesn't work - - kadovInitEachChild(child); - } -} - -function kadovRetrieveTextInner(el) -{ - var x = ""; - if( (!el) || (el.tagName == "!") || (el.tagName == "SCRIPT" )) - return x; - - if( kadovIsParagraph(el) ) - { - var strNewID = " "; - if( el.id != "" ) - strNewID += "id=" + el.id + "_NewSpan "; - x = "" + el.innerHTML + ""; - } - else - { - for(var i=0; i 0 ) - if( (nTagClose - nTagOpen) != nDistance ) - return strRawHTML; - - var strCleanOnce = strRawHTML.substring(0, nTagOpen) + strRawHTML.substr(nTagClose + strTagClose.length) ; - return kadovRetrieveCleanHTML( strCleanOnce, strTagOpen, strTagClose ); -} - -function kadovAdjustObjectTag(strRawHTML, nStartPos) -{// adjust object tag for related topics HTML control, because innerHTML misses out the item settings - - //Is there any DTC? - var strDTCTagOpen = ''; - var nDTCTagOpen = strRawHTML.indexOf( strDTCTagOpen, nStartPos ); - if( nDTCTagOpen < 0 ) - return strRawHTML; - var nDTCTagClose = strRawHTML.indexOf( strDTCTagClose, nDTCTagOpen ); - if( nDTCTagClose < nDTCTagOpen) - return strRawHTML; // no Design Time Controls; - - //Is the DTC HTML Help Control? - var strRTObjTagOpen = 'classid=clsid:ADB880A6-D8FF-11CF-9377-00AA003B7A11'; - var strRTObjTagClose = ''; - var nRTObjTagOpen = strRawHTML.indexOf( strRTObjTagOpen, nDTCTagOpen ); - if( nRTObjTagOpen < nDTCTagOpen ) - return strRawHTML; - var nRTObjTagClose = strRawHTML.indexOf( strRTObjTagClose, nRTObjTagOpen ); - if( nRTObjTagClose < nRTObjTagOpen ) - return strRawHTML; // is not a HTML help control - - // Is it a related Topics html help control? - var strRTObjLabel = ''; - } - - // to insert the reconstructed item params into runtime object tag - var strAdjustedHTML = strRawHTML.substring(0,nRTObjTagClose) + strRunTimeItemParam + strRawHTML.substring(nRTObjTagClose, strRawHTML.length); - return kadovAdjustObjectTag(strAdjustedHTML, nDTCTagClose + strDTCTagClose.length); -} - -function kadovTextPopupOnLoad( el ) -{ - if( !CCSSP.bIsWinOS && !CCSSP.bIsSunOS) - return 0; - - if( typeof(el) == "string" ) - el = document.all.item(el); - - var src = el.getAttribute( "x-use-popup" ); - if(!src) - return 0; - - var name = src; - if( src.substr(0,1) == "#" ) - name = src.substr(1, src.length-1); - var srcDiv = document.all.item(name); - if( !srcDiv ) - return 1; - - var type = el.getAttribute( "x-popup-type" ); - var setup = el.getAttribute( "x-tmp-setup" ); - var newId = name; - if( newId.indexOf( "_tmp") <= 0 ) - newId += "_tmp"; - - if( !setup ) - { - el.setAttribute( "x-tmp-setup", 1 ); - - if( type == "pulldown" ) - { - var strAdjust = kadovAdjustObjectTag(srcDiv.innerHTML,0); - var strCleanHTML = kadovRetrieveCleanHTML(strAdjust, ""); - strCleanHTML = kadovRetrieveCleanHTML(strCleanHTML, ""); - - //work around the bug in HH.exe that highlight the phrases when use Search tab - //this approach is just removing the tag inserted by Microsoft in the runtime - strCleanHTML = kadovRetrieveCleanHTML(strCleanHTML, "", 52); - - var strStyle = " style='display:none; position:relative;"; - var newDiv = "
" + strCleanHTML + "
"; - - srcDiv.outerHTML = ""; // empty the original DIV tag - var elParentPra = kadovFindParentParagraph(el); - if( elParentPra ) - elParentPra.insertAdjacentHTML( "AfterEnd", newDiv ); - } - else if( type == "expanding" ) - { - var inner = kadovRetrieveTextInner(srcDiv); - if( inner == "" ) - inner = srcDiv.innerHTML; - var strAdjust = kadovAdjustObjectTag(inner,0); - var strCleanHTML = kadovRetrieveCleanHTML(strAdjust, ""); - strCleanHTML = kadovRetrieveCleanHTML(strCleanHTML, ""); - var strClassName = (el.className == "glossterm") ? "glosstext" : "expandtext"; - var newSpan = ""; - srcDiv.outerHTML = ""; // empty the original DIV tag - el.insertAdjacentHTML( "AfterEnd", newSpan ); - } - } - return 0; -} - -function kadovTextPopup( el ) -{ - if( (!CCSSP.bIsWinOS && !CCSSP.bIsSunOS) || (window.event == null) ) - return; - window.event.cancelBubble = true; - - if( typeof(el) == "string" ) - el = document.all.item(el); - - var src = el.getAttribute( "x-use-popup" ); - if(!src) - return; - - var name = src; - if( src.substr(0,1) == "#" ) - name = src.substr(1, src.length-1) + "_tmp"; - var srcDiv = document.all.item(name); - if( !srcDiv ) - return; - - var type = el.getAttribute( "x-popup-type" ); - var setup = el.getAttribute( "x-tmp-setup" ); - if( srcDiv ) - { - if( srcDiv.style.display == "" ) - srcDiv.style.display = "none"; - else - { - srcDiv.style.display = ""; - if( typeof(srcDiv.bInitialized) == "undefined" ) - { - srcDiv.bInitialized = true; - kadovInitEffects(name); - kadovInitEachChild(srcDiv); - } - } - } - return; -} - -function kadovFindParentParagraph( el ) -{ - if( typeof(el) == "string" ) - el = document.all.item(el); - if( (!el) || el.tagName == "BODY" ) - return null; - if( kadovIsParagraph(el.parentElement) ) - return el.parentElement; - else - return kadovFindParentParagraph( el.parentElement ); -} - -//Begin HTML code invoked function -function kadovRegisterEventHandler( obj, strEvent, strEventHandler ) -{ - if( !gbBsIE4 ) - return; - CCSSP.RegisterEventHandler( obj, strEvent, strEventHandler ); -} - -function kadovTextPopupInit( el ) -{ - if( (!gbBsIE4) || (!CCSSP.bIsWinOS && !CCSSP.bIsSunOS)) - return; - - if( typeof(el) == "string" ) - el = document.all.item(el); - if( el != null ) - { - CCSSP.RegisterEventHandler( el, "onclick", "kadovTextPopup(" + el.id +");" ); - CCSSP.RegisterEventHandler( window, "onload", "kadovTextPopupOnLoad(" + el.id +");" ); - } -} -//End HTML code invoked function - -//End to support extended and dropdown text effects. - -//Begin to convert iWrite format to RoboEditor Format for DHTML effects -function kadovInitTriggersInHead( ) -{ - if( Object.xDelayedInitElements ) - { - var x = Object.xDelayedInitElements; - for(i=0; i= 0 && nNext < values.length ) - { - functions[nIdx] = values.substr( nStart, nNext-nStart+1); - nStart = nNext + 1; - nIdx++; - nNext = values.indexOf( "\)", nStart); - } - - for( var i=0; i= 0 ) - { - nPageClick = arrForClickCount[j].indexOf("="); - if( nPageClick > 0 ) - { - nClickTimes = arrForClickCount[j].substring( nPageClick + 1, arrForClickCount[j].length) * 1; - break; - } - } - } - var args = srcargs; - if( j < arrForClickCount.length ) - {// to strip out the "clicks=99" from the arguments string - args = ""; - for( var k = 0; k < arrForClickCount.length; k ++ ) - { - if( k != j ) - { - args += arrForClickCount[k]; - if( k < arrForClickCount.length - 1 ) - args += ","; - } - } - } - bsscFXInit( null, id, translatedProp, fnname, args, nClickTimes ); - } -} - -function kadovTranslateProp( prop ) -{ - switch( prop ) - { - case "x-on-hover" : return "bsschover"; - case "x-on-pageclick" : return "bsscpageclick"; - case "x-on-pageload" : return "bsscpageload"; - case "x-on-trigger-1" : return "bssctrigger1"; - case "x-on-trigger-2" : return "bssctrigger2"; - } - return null; -} -//End to convert iWrite format to RoboEditor Format for DHTML effects - -//Begin the definition of one entry to DHTML effects -function bsscFXInit( trigger_ID, target_ID, event_type, - action_type, action_setting, event_addional ) -{ - if( (!CCSSP.bIsWinOS && !CCSSP.bIsSunOS ) || typeof(target_ID) != "string" )//MUST have a target_ID - return; // we don't support Navigator yet - - if( typeof(event_type) == "string" ) - event_type = event_type.toLowerCase(); - if( typeof(action_type) == "string" ) - action_type = action_type.toLowerCase(); - if( typeof(action_setting) == "string" ) - action_setting = action_setting.toLowerCase(); - - // to get the target element then add it to the target list - var eleTarget = CCSSP.GetObject( target_ID ); - if( (eleTarget != null) && (event_type != null) && (action_type != null) ) - { - CEngine.AddOneTarget( target_ID, eleTarget ); - CEngine.BuildTargetObject(target_ID, event_type, action_type, action_setting, event_addional); - } - - // to validate the trigger_ID parameter - if( typeof(trigger_ID) == "string" && trigger_ID != "" ) - CEngine.BuildTriggerObject( trigger_ID, target_ID ); -} -//End the definition of one entry to DHTML effects - - -/// Section End - kadov DHTM (JavaScript 1.2) - - -/// Section Begin - CCSSP DHTM (JavaScript 1.2) - -//Begin JavaScript libary for cross-platform positioning object. -function CCSSP(){} // constructor of CCSSP class - -CCSSP.strAgent = navigator.userAgent.toLowerCase(); -CCSSP.nAppVersion = parseInt(navigator.appVersion); - -CCSSP.bIsWinOS = ((CCSSP.strAgent.indexOf("win") >= 0) || (CCSSP.strAgent.indexOf("16bit") >= 0)); -CCSSP.bIsMacOS = (CCSSP.strAgent.indexOf("mac") >= 0); -CCSSP.bIsSunOS = (CCSSP.strAgent.indexOf("sunos") != -1); - -CCSSP.bIsIE = (navigator.appName.indexOf("Microsoft") >= 0); -CCSSP.bIsIE4 = (CCSSP.bIsIE && (CCSSP.nAppVersion >= 4)); -CCSSP.bIsIE5 = (CCSSP.bIsIE4 && (CCSSP.strAgent.indexOf("msie 5") != -1) ) - -CCSSP.bIsNav = (navigator.appName.indexOf("Netscape") >= 0); -CCSSP.bIsNav4 = (CCSSP.bIsNav && (CCSSP.nAppVersion >= 4)); -CCSSP.bIsNav6 = (CCSSP.bIsNav && (CCSSP.nAppVersion >= 5)); - -CCSSP.GetObject = function( obj ) -{//convert object name string or reference into a valid object reference - if( typeof(obj) == "object" ) - return obj; - else if( typeof(obj) == "string" && obj != "") - { - if( CCSSP.bIsNav4 ) - return eval("document." + obj); - else - return eval("document.all." + obj); - } - else - return null; -} - -CCSSP.MoveObjectTo = function(obj, x, y) -{//positioning an object at a specific pixel coordinate - if( CCSSP.bIsNav4 ) - obj.moveTo(x,y); - else - { - obj.style.pixelLeft = x; - obj.style.pixelTop = y; - } -} - -CCSSP.MoveObjectBy = function(obj, dx, dy) -{//moveing a object by x and/or y pixel - if( CCSSP.bIsNav4 ) - obj.moveBy(dx,dy); - else - { - obj.style.pixelLeft += dx; - obj.style.pixelTop += dy; - } -} - -CCSSP.SetObjectBGColor = function(obj, color) -{//set the background color of an object - if( CCSSP.bIsNav4 ) - obj.bgColor = color; - else - obj.style.backgroundColor = color; -} - -CCSSP.ShowObject = function(obj, bShow) -{// set the object to be visible or invisible - if( CCSSP.bIsNav4 ) - obj.visibility = (bShow == true) ? 'show' : 'hide'; - else - obj.style.visibility = (bShow == true) ? 'visible' : 'hidden';// when hidden, it still occupy some space. -} - -CCSSP.GetObjectLeft = function(obj) -{// retrieve the x coordinate of a posionable object - if( CCSSP.bIsNav4 ) - return obj.left; - else - return obj.style.pixelLeft; -} - -CCSSP.GetObjectTop = function(obj) -{// retrieve the y coordinate of a posionable object - if( CCSSP.bIsNav4 ) - return obj.top; - else - return obj.style.pixelTop; -} - -CCSSP.GetObjectContainLeft = function(obj) -{// retrieve the x coordinate of a posionable object relative to it's parent element - if( CCSSP.bIsNav4 ) - return obj.pageX; - else - { - if( obj == document.body ) - return obj.clientLeft; - else - return obj.offsetLeft; - } -} - -CCSSP.GetObjectWindowLeft = function(obj) -{// retrieve the x coordinate of a posionable object relative to browser window - if( CCSSP.bIsNav4 ) - return obj.pageX; - else - { - var nOffsetWindowLeft = 0; - for(var element = obj; element; element = element.offsetParent) - nOffsetWindowLeft += CCSSP.GetObjectContainLeft(element); - return nOffsetWindowLeft; - } -} - -CCSSP.GetObjectContainTop = function(obj) -{// retrieve the y coordinate of a posionable object relative to it's parent element - if( CCSSP.bIsNav4 ) - return obj.pageY; - else - { - if( obj == document.body ) - return obj.clientTop; - else - return obj.offsetTop; - } -} - -CCSSP.GetObjectWindowTop = function(obj) -{// retrieve the y coordinate of a posionable object relative to browser window - if( CCSSP.bIsNav4 ) - return obj.pageY; - else - { - var nOffsetWindowTop = 0; - for(var element = obj; element; element = element.offsetParent) - nOffsetWindowTop += CCSSP.GetObjectContainTop(element); - return nOffsetWindowTop; - } -} - -CCSSP.GetObjectHeight = function(obj) -{// retrieve the height of a posionable object - if( CCSSP.bIsNav4 ) - return obj.clip.height; - else - return obj.offsetHeight; -} - -CCSSP.GetObjectWidth = function(obj) -{// retrieve the width of a posionable object - if( CCSSP.bIsNav4 ) - return obj.clip.width; - else - return obj.offsetWidth; -} - -CCSSP.RegisterEventHandler = function( srcObj, rawEventName, funcHandler ) -{ // to add the "funcHandler" as the "rawEventName" 's handler to the "srcObj" object,the original event handler will be combined - if (CCSSP.bIsNav4 && !CCSSP.bIsNav6) - return ; - var oldHandler = ""; - if (CCSSP.bIsMacOS && CCSSP.bIsIE4) - { - if (typeof(srcObj[rawEventName.toLowerCase()])=="unknown") - { //search for