@ansonlai/docx-redline-js 0.2.0 → 0.4.0

package/scripts/test-multilevel-bullet-visual.mjs

@@ -0,0 +1,187 @@
+import '../tests/setup-xml-provider.mjs';
+
+import fs from 'fs';
+import path from 'path';
+import { spawnSync } from 'child_process';
+import { fileURLToPath } from 'url';
+import { buildMinimalDocx } from './lib/minimal-zip.mjs';
+import { applyOperationsToDocumentXml } from '../services/standalone-operation-runner.js';
+import {
+    acceptTrackedChangesInOoxml,
+    rejectTrackedChangesInOoxml
+} from '../index.js';
+import { elementsByLocalName, parseXml } from '../tests/helpers/ooxml-assertions.mjs';
+
+const __dirname = path.dirname(fileURLToPath(import.meta.url));
+const rootDir = path.resolve(__dirname, '..');
+const outputDir = path.join(rootDir, 'tmp', 'multilevel-bullet-visual');
+const pythonExe = path.join(rootDir, 'tmp', 'visual-qa-venv', 'Scripts', 'python.exe');
+
+const NS_W = 'http://schemas.openxmlformats.org/wordprocessingml/2006/main';
+
+const NUMBERING_XML = `<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<w:numbering xmlns:w="${NS_W}" xmlns:w15="http://schemas.microsoft.com/office/word/2012/wordml">
+  <w:abstractNum w:abstractNumId="0" w15:restartNumberingAfterBreak="0">
+    <w:nsid w:val="71A1B101"/><w:multiLevelType w:val="hybridMultilevel"/><w:tmpl w:val="71A1B101"/>
+    <w:lvl w:ilvl="0"><w:start w:val="1"/><w:numFmt w:val="upperRoman"/><w:lvlText w:val="%1."/><w:lvlJc w:val="left"/><w:pPr><w:tabs><w:tab w:val="num" w:pos="720"/></w:tabs><w:ind w:left="720" w:hanging="360"/></w:pPr><w:rPr><w:rFonts w:ascii="Calibri" w:hAnsi="Calibri"/></w:rPr></w:lvl>
+    <w:lvl w:ilvl="1"><w:start w:val="1"/><w:numFmt w:val="lowerLetter"/><w:lvlText w:val="(%2)"/><w:lvlJc w:val="left"/><w:pPr><w:tabs><w:tab w:val="num" w:pos="1440"/></w:tabs><w:ind w:left="1440" w:hanging="360"/></w:pPr><w:rPr><w:rFonts w:ascii="Calibri" w:hAnsi="Calibri"/></w:rPr></w:lvl>
+    <w:lvl w:ilvl="2"><w:start w:val="1"/><w:numFmt w:val="bullet"/><w:lvlText w:val="•"/><w:lvlJc w:val="left"/><w:pPr><w:tabs><w:tab w:val="num" w:pos="2160"/></w:tabs><w:ind w:left="2160" w:hanging="360"/></w:pPr><w:rPr><w:rFonts w:ascii="Calibri" w:hAnsi="Calibri"/></w:rPr></w:lvl>
+  </w:abstractNum>
+  <w:num w:numId="10"><w:abstractNumId w:val="0"/></w:num>
+</w:numbering>`;
+
+const SOURCE_DOCUMENT_XML = `<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<w:document xmlns:w="${NS_W}">
+  <w:body>
+    <w:p><w:pPr><w:pStyle w:val="Heading1"/></w:pPr><w:r><w:rPr><w:b/><w:sz w:val="32"/></w:rPr><w:t>Corporate Governance Charter</w:t></w:r></w:p>
+    <w:p><w:pPr><w:numPr><w:ilvl w:val="0"/><w:numId w:val="10"/></w:numPr></w:pPr><w:r><w:t>Corporate Governance Framework.</w:t></w:r></w:p>
+    <w:p><w:pPr><w:numPr><w:ilvl w:val="1"/><w:numId w:val="10"/></w:numPr></w:pPr><w:r><w:t>Audit and Risk Oversight.</w:t></w:r></w:p>
+    <w:p><w:pPr><w:numPr><w:ilvl w:val="2"/><w:numId w:val="10"/></w:numPr></w:pPr><w:r><w:t>Review internal accounting controls.</w:t></w:r></w:p>
+    <w:p><w:pPr><w:numPr><w:ilvl w:val="2"/><w:numId w:val="10"/></w:numPr></w:pPr><w:r><w:t>Deprecated manual transaction logging.</w:t></w:r></w:p>
+    <w:p><w:pPr><w:numPr><w:ilvl w:val="1"/><w:numId w:val="10"/></w:numPr></w:pPr><w:r><w:t>Executive Compensation.</w:t></w:r></w:p>
+    <w:p><w:pPr><w:numPr><w:ilvl w:val="2"/><w:numId w:val="10"/></w:numPr></w:pPr><w:r><w:t>Review executive incentive structures.</w:t></w:r></w:p>
+    <w:p><w:pPr><w:numPr><w:ilvl w:val="0"/><w:numId w:val="10"/></w:numPr></w:pPr><w:r><w:t>Regulatory Compliance Protocol.</w:t></w:r></w:p>
+    <w:sectPr/>
+  </w:body>
+</w:document>`;
+
+async function main() {
+    fs.mkdirSync(outputDir, { recursive: true });
+
+    console.log('=== Multi-Level Bullet Visual Test ===');
+    console.log('Building 3-level list document and executing concurrent multi-level operations...');
+
+    // Multi-level operations:
+    // 1. Level 0 edit: Corporate Governance Framework -> Corporate Governance and Enterprise Risk Framework
+    // 2. Level 2 insert: under Audit and Risk Oversight, insert "Conduct annual cybersecurity vulnerability audit."
+    // 3. Level 2 edit: Review internal accounting controls -> Review and audit internal accounting controls
+    // 4. Level 2 consolidation/edit: Deprecated manual transaction logging -> Maintain automated electronic audit trails
+    // 5. Level 1 edit: Executive Compensation -> Executive Compensation and Performance Oversight
+    const operations = [
+        {
+            type: 'redline',
+            target: 'Corporate Governance Framework.',
+            modified: 'Corporate Governance and Enterprise Risk Framework.'
+        },
+        {
+            type: 'list-change',
+            target: 'Audit and Risk Oversight.',
+            modified: '1. Audit and Risk Oversight.\n  - Conduct annual cybersecurity vulnerability audit.'
+        },
+        {
+            type: 'redline',
+            target: 'Review internal accounting controls.',
+            modified: 'Review and audit internal accounting controls.'
+        },
+        {
+            type: 'redline',
+            target: 'Deprecated manual transaction logging.',
+            modified: 'Maintain automated electronic audit trails.'
+        },
+        {
+            type: 'redline',
+            target: 'Executive Compensation.',
+            modified: 'Executive Compensation and Performance Oversight.'
+        }
+    ];
+
+    const result = await applyOperationsToDocumentXml(SOURCE_DOCUMENT_XML, operations, 'VisualTester');
+    if (!result.hasChanges) {
+        throw new Error('Operations did not report changes');
+    }
+
+    // Generate the 3 OOXML states
+    const allMarkupXml = result.documentXml;
+    const acceptAllXml = acceptTrackedChangesInOoxml(allMarkupXml, { author: 'VisualTester' }).oxml;
+    const _rejectAllXml = rejectTrackedChangesInOoxml(allMarkupXml, { author: 'VisualTester' }).oxml;
+
+    // Verify structural invariants in OOXML
+    const acceptedDoc = parseXml(acceptAllXml);
+    const acceptedParagraphs = elementsByLocalName(acceptedDoc, 'p');
+    // Heading + 8 list items = 9 paragraphs
+    console.log(`Accepted paragraphs: ${acceptedParagraphs.length}`);
+
+    const parts = { numberingXml: NUMBERING_XML };
+    const docxPath = path.join(outputDir, 'multilevel-bullet-edits.docx');
+    fs.writeFileSync(docxPath, buildMinimalDocx(allMarkupXml, parts));
+    console.log(`Generated DOCX fixture: ${docxPath}`);
+
+    // Render 3 views using Word COM via PowerShell
+    console.log('Rendering 3 views in Microsoft Word COM (ExportAsFixedFormat)...');
+    const psScriptPath = path.join(__dirname, 'render-multilevel-bullet-visual.ps1');
+    const psResult = spawnSync('powershell', [
+        '-NoProfile',
+        '-ExecutionPolicy',
+        'Bypass',
+        '-File',
+        psScriptPath,
+        '-DocxPath',
+        docxPath,
+        '-OutputDir',
+        outputDir
+    ], {
+        encoding: 'utf8',
+        cwd: rootDir
+    });
+
+    if (psResult.status !== 0) {
+        console.error('Word COM rendering failed:', psResult.stderr || psResult.stdout);
+        throw new Error('Word COM rendering failed');
+    }
+    console.log(psResult.stdout.trim());
+
+    // Render high-res PNGs and contact sheet using PyMuPDF
+    console.log('Converting PDFs to 150 DPI page images using PyMuPDF...');
+    const pyScript = `
+import os
+import fitz
+from PIL import Image
+
+output_dir = r"${outputDir}"
+views = ["allMarkup", "acceptAll", "rejectAll"]
+rendered_images = []
+
+for view in views:
+    pdf_path = os.path.join(output_dir, f"multilevel-bullet--{view}.pdf")
+    doc = fitz.open(pdf_path)
+    page = doc.load_page(0)
+    pix = page.get_pixmap(dpi=150)
+    img_path = os.path.join(output_dir, f"multilevel-bullet--{view}--page1.png")
+    pix.save(img_path)
+    rendered_images.append(img_path)
+    print(f"Rendered {img_path} ({pix.width}x{pix.height})")
+    doc.close()
+
+# Create side-by-side contact sheet
+images = [Image.open(p) for p in rendered_images]
+total_w = sum(img.width for img in images) + 40
+max_h = max(img.height for img in images) + 20
+sheet = Image.new("RGB", (total_w, max_h), (240, 240, 240))
+x = 10
+for img in images:
+    sheet.paste(img, (x, 10))
+    x += img.width + 10
+
+sheet_path = os.path.join(output_dir, "multilevel-bullet--contact-sheet.png")
+sheet.save(sheet_path)
+print(f"Created contact sheet: {sheet_path} ({sheet.width}x{sheet.height})")
+`;
+
+    const pyResult = spawnSync(pythonExe, ['-c', pyScript], {
+        encoding: 'utf8',
+        cwd: rootDir
+    });
+
+    if (pyResult.status !== 0) {
+        console.error('PyMuPDF conversion failed:', pyResult.stderr || pyResult.stdout);
+        throw new Error('PyMuPDF conversion failed');
+    }
+    console.log(pyResult.stdout.trim());
+
+    console.log('\n✅ Visual test execution complete! Images ready for multimodal inspection in:', outputDir);
+}
+
+main().catch(err => {
+    console.error('Fatal error:', err);
+    process.exit(1);
+});

package/scripts/word-com-corpus-suite.ps1

@@ -0,0 +1,43 @@
+param(
+    [string]$FixturesDir = "tmp/superdoc-word-fixtures",
+    [string]$SourcesDir = "tmp/superdoc-corpus"
+)
+
+$ErrorActionPreference = 'Stop'
+
+$repoRoot = Split-Path -Parent $PSScriptRoot
+$fixturesPath = if ([System.IO.Path]::IsPathRooted($FixturesDir)) {
+    [System.IO.Path]::GetFullPath($FixturesDir)
+} else {
+    [System.IO.Path]::GetFullPath((Join-Path $repoRoot $FixturesDir))
+}
+$sourcesPath = if ([System.IO.Path]::IsPathRooted($SourcesDir)) {
+    [System.IO.Path]::GetFullPath($SourcesDir)
+} else {
+    [System.IO.Path]::GetFullPath((Join-Path $repoRoot $SourcesDir))
+}
+$manifestPath = Join-Path $repoRoot 'tests\corpus\superdoc-english-legal-administrative.json'
+$manifest = Get-Content -LiteralPath $manifestPath -Raw -Encoding UTF8 | ConvertFrom-Json
+$fetchArgs = @((Join-Path $repoRoot 'scripts\fetch-superdoc-corpus.mjs'))
+foreach ($document in $manifest.documents) {
+    $fetchArgs += '--id'
+    $fetchArgs += [string]$document.id
+}
+
+& node @fetchArgs
+if ($LASTEXITCODE -ne 0) { exit $LASTEXITCODE }
+
+& node (Join-Path $repoRoot 'scripts\prepare-superdoc-word-corpus.mjs') `
+    --input-dir $sourcesPath `
+    --output-dir $fixturesPath
+if ($LASTEXITCODE -ne 0) { exit $LASTEXITCODE }
+
+& (Join-Path $repoRoot 'scripts\package-superdoc-word-fixtures.ps1') `
+    -FixturesDir $fixturesPath `
+    -SourcesDir $sourcesPath
+if ($LASTEXITCODE -ne 0) { exit $LASTEXITCODE }
+
+& (Join-Path $repoRoot 'scripts\word-com-differential.ps1') `
+    -FixturesDir $fixturesPath `
+    -SourcesDir $sourcesPath
+exit $LASTEXITCODE

package/scripts/word-com-corpus-visual-suite.ps1

@@ -0,0 +1,116 @@
+param(
+    [string]$FixturesDir = "",
+    [string]$OutputDir = "tmp/superdoc-word-visual-review/rendered",
+    [string[]]$Case = @()
+)
+
+$ErrorActionPreference = 'Stop'
+$repoRoot = Split-Path -Parent $PSScriptRoot
+if (-not $FixturesDir) { $FixturesDir = "tmp/superdoc-word-visual-review/fixtures-$PID" }
+$fixturesPath = [System.IO.Path]::GetFullPath((Join-Path $repoRoot $FixturesDir))
+$outputPath = [System.IO.Path]::GetFullPath((Join-Path $repoRoot $OutputDir))
+$manifestPath = Join-Path $outputPath 'manifest.json'
+New-Item -ItemType Directory -Force -Path $fixturesPath, $outputPath | Out-Null
+
+& (Join-Path $PSScriptRoot 'word-com-corpus-suite.ps1') -FixturesDir $fixturesPath
+if ($LASTEXITCODE -ne 0) { exit $LASTEXITCODE }
+
+$prepareArgs = @((Join-Path $PSScriptRoot 'prepare-corpus-word-visual-review.mjs'), "--output=$manifestPath")
+foreach ($caseName in $Case) { $prepareArgs += "--case=$caseName" }
+& node @prepareArgs
+if ($LASTEXITCODE -ne 0) { exit $LASTEXITCODE }
+
+$manifest = Get-Content -LiteralPath $manifestPath -Raw -Encoding UTF8 | ConvertFrom-Json
+$word = $null
+$rpcDisconnected = -2147417848 # 0x80010108 RPC_E_DISCONNECTED
+
+function Test-RpcDisconnected($exception) {
+    return $exception -is [Runtime.InteropServices.COMException] -and $exception.HResult -eq $rpcDisconnected
+}
+
+function Start-WordApplication {
+    $application = New-Object -ComObject Word.Application
+    $application.Visible = $false
+    $application.DisplayAlerts = 0
+    return $application
+}
+
+function Stop-WordApplicationSafely($application) {
+    if ($null -eq $application) { return }
+    try { $application.Quit() | Out-Null }
+    catch {
+        if (-not (Test-RpcDisconnected $_.Exception)) { throw }
+        Write-Warning 'Word disconnected while quitting; the COM process was already gone.'
+    }
+}
+
+function Close-WordDocumentSafely($document) {
+    if ($null -eq $document) { return }
+    try { $document.Close(0) | Out-Null }
+    catch {
+        if (-not (Test-RpcDisconnected $_.Exception)) { throw }
+        # ExportAsFixedFormat occasionally tears down Word's document proxy
+        # after successfully writing a PDF. Do not let cleanup hide the valid
+        # render; the next operation will restart Word if needed.
+        Write-Warning 'Word disconnected after export while closing the document.'
+    }
+}
+
+function Export-View($word, [string]$docxPath, [string]$pdfPath, [string]$viewName) {
+    $document = $null
+    try {
+        $document = $word.Documents.OpenNoRepairDialog($docxPath)
+        if ($viewName -eq 'acceptAll') { $document.AcceptAllRevisions(); $exportItem = 0 }
+        elseif ($viewName -eq 'rejectAll') { $document.RejectAllRevisions(); $exportItem = 0 }
+        else { $document.ShowRevisions = $true; $document.PrintRevisions = $true; $exportItem = 7 }
+        $pages = $document.ComputeStatistics(2)
+        $document.ExportAsFixedFormat($pdfPath, 17, $false, 0, 0, 1, 1, $exportItem, $true, $true, 0, $true, $true, $false)
+        $pdf = Get-Item -LiteralPath $pdfPath -ErrorAction Stop
+        if ($pdf.Length -lt 1000 -or $pages -lt 1) { throw "Word produced invalid visual evidence" }
+        return @{ pages = [int]$pages; bytes = [long]$pdf.Length }
+    }
+    finally { Close-WordDocumentSafely $document }
+}
+
+try {
+    $word = Start-WordApplication
+    $manifest.word.version = [string]$word.Version
+    try { $manifest.word.build = [string]$word.Build } catch { $manifest.word.build = $null }
+    foreach ($entry in $manifest.cases) {
+        $docxPath = Join-Path $fixturesPath "$($entry.name).docx"
+        foreach ($viewProperty in $entry.views.PSObject.Properties) {
+            $pdfPath = Join-Path $outputPath ([string]$viewProperty.Value.pdf)
+            $render = $null
+            for ($attempt = 1; $attempt -le 2; $attempt++) {
+                try {
+                    $render = Export-View $word $docxPath $pdfPath ([string]$viewProperty.Name)
+                    break
+                }
+                catch {
+                    if (-not (Test-RpcDisconnected $_.Exception) -or $attempt -eq 2) { throw }
+                    Write-Warning "Word disconnected while rendering $($entry.identity) $($viewProperty.Name); restarting and retrying once."
+                    Stop-WordApplicationSafely $word
+                    $word = Start-WordApplication
+                }
+            }
+            $viewProperty.Value.status = 'rendered'
+            $viewProperty.Value.pages = $render.pages
+            $viewProperty.Value.bytes = $render.bytes
+        }
+        $entry.renderStatus = 'rendered'
+        Write-Output "RENDERED $($entry.identity) (markup, accept-all, reject-all)"
+    }
+}
+finally {
+    Stop-WordApplicationSafely $word
+    $manifest | ConvertTo-Json -Depth 12 | Set-Content -LiteralPath $manifestPath -Encoding UTF8
+}
+
+Write-Output "Corpus visual evidence: $($manifest.cases.Count) cases, $($manifest.cases.Count * 3) PDFs."
+Write-Output "Manifest: $manifestPath"
+
+# Keep the self-contained comparison dashboard synchronized with the exact
+# process-isolated DOCX fixtures that Word just validated and rendered.
+& node (Join-Path $PSScriptRoot 'build-test-dashboard.mjs') --corpus-fixtures-dir $fixturesPath
+if ($LASTEXITCODE -ne 0) { exit $LASTEXITCODE }
+Write-Output "Dashboard refreshed from: $fixturesPath"

package/scripts/word-com-differential.ps1

@@ -1,5 +1,6 @@
 param(
-    [string]$FixturesDir = "tmp/validation-docx"
+    [string]$FixturesDir = "tmp/validation-docx",
+    [string]$SourcesDir = ""
 )
 
 # Differential validation against desktop Microsoft Word (the authoritative
@@ -15,10 +16,68 @@ param(
 
 $ErrorActionPreference = 'Stop'
 
-function Get-NormalizedText([string]$text) {
+function Get-ComparableText([string]$text, [string]$fidelity = 'exact') {
     if ($null -eq $text) { return '' }
-    $text = $text -replace [string][char]7, ' '   # table cell markers
-    return ($text -replace '\s+', ' ').Trim()
+    $text = $text -replace [string][char]7, ''   # table cell markers
+    $text = $text -replace [string][char]2, ''   # footnote/endnote reference markers
+    $text = $text -replace "`r`n", "`n"
+    $text = $text -replace "`r", "`n"
+    $text = $text -replace "`n+$", ''           # Word's terminal paragraph mark
+    if ($fidelity -eq 'normalized') {
+        return ($text -replace '\s+', ' ').Trim()
+    }
+    return $text
+}
+
+function Test-ContainsAssertions([string]$text, $required, $absent, [string]$phase, [string]$name) {
+    $passed = $true
+    foreach ($value in @($required)) {
+        if (-not $text.Contains([string]$value)) {
+            Write-Output "FAIL ${name}: ${phase} is missing expected text: $value"
+            $passed = $false
+        }
+    }
+    foreach ($value in @($absent)) {
+        if ($text.Contains([string]$value)) {
+            Write-Output "FAIL ${name}: ${phase} still contains superseded text: $value"
+            $passed = $false
+        }
+    }
+    return $passed
+}
+
+function Test-UntouchedPartHashes([string]$docxPath, $expectedHashes, [string]$name) {
+    $properties = @($expectedHashes.PSObject.Properties)
+    if ($properties.Count -eq 0) { return $true }
+
+    Add-Type -AssemblyName System.IO.Compression.FileSystem
+    $archive = [System.IO.Compression.ZipFile]::OpenRead($docxPath)
+    $sha256 = [System.Security.Cryptography.SHA256]::Create()
+    try {
+        foreach ($property in $properties) {
+            $entry = $archive.GetEntry([string]$property.Name)
+            if ($null -eq $entry) {
+                Write-Output "FAIL ${name}: package is missing untouched part $($property.Name)"
+                return $false
+            }
+            $stream = $entry.Open()
+            try {
+                $actual = ([System.BitConverter]::ToString($sha256.ComputeHash($stream))).Replace('-', '').ToLowerInvariant()
+            }
+            finally {
+                $stream.Dispose()
+            }
+            if ($actual -ne [string]$property.Value) {
+                Write-Output "FAIL ${name}: untouched part hash changed for $($property.Name)"
+                return $false
+            }
+        }
+    }
+    finally {
+        $sha256.Dispose()
+        $archive.Dispose()
+    }
+    return $true
 }
 
 $resolvedDir = Resolve-Path -LiteralPath $FixturesDir -ErrorAction SilentlyContinue
@@ -27,7 +86,17 @@ if (-not $resolvedDir) {
     exit 1
 }
 
-$expectations = Get-ChildItem -LiteralPath $resolvedDir -Filter '*.expected.json' | Sort-Object Name
+$suitePath = Join-Path $resolvedDir 'suite.json'
+if (Test-Path -LiteralPath $suitePath) {
+    $suite = Get-Content -LiteralPath $suitePath -Raw -Encoding UTF8 | ConvertFrom-Json
+    $expectations = @($suite.cases | ForEach-Object {
+        $caseName = if ($_ -is [string]) { [string]$_ } else { [string]$_.name }
+        Get-Item -LiteralPath (Join-Path $resolvedDir "$caseName.expected.json") -ErrorAction Stop
+    })
+}
+else {
+    $expectations = @(Get-ChildItem -LiteralPath $resolvedDir -Filter '*.expected.json' | Sort-Object Name)
+}
 if ($expectations.Count -eq 0) {
     Write-Error "No *.expected.json fixtures in '$resolvedDir'. Run: node scripts/export-validation-fixtures.mjs"
     exit 1
@@ -38,10 +107,38 @@ $failures = 0
 $results = @()
 
 function Open-FixtureDocument($word, [string]$path) {
-    # Single-argument Open: Windows PowerShell 5.1 COM binding rejects the
-    # long optional-parameter signature. Defaults leave the document
-    # writable, which accept/reject requires; nothing is ever saved.
-    return $word.Documents.Open($path)
+    # OpenNoRepairDialog prevents malformed real-world packages from blocking
+    # a headless run behind an invisible repair prompt. It still leaves the
+    # document writable for accept/reject; nothing is ever saved.
+    return $word.Documents.OpenNoRepairDialog($path)
+}
+
+function Get-ScopeText($document, [string]$scope, [string]$fidelity) {
+    if ($scope -ne 'headers') {
+        return Get-ComparableText $document.Content.Text $fidelity
+    }
+    $parts = @()
+    foreach ($section in @($document.Sections)) {
+        foreach ($headerType in 1..3) {
+            $header = $section.Headers.Item($headerType)
+            if ($header.Exists) {
+                $parts += Get-ComparableText $header.Range.Text $fidelity
+            }
+        }
+    }
+    return ($parts -join "`n")
+}
+
+function Get-ScopeRevisionCount($document, [string]$scope) {
+    if ($scope -ne 'headers') { return $document.Revisions.Count }
+    $count = 0
+    foreach ($section in @($document.Sections)) {
+        foreach ($headerType in 1..3) {
+            $header = $section.Headers.Item($headerType)
+            if ($header.Exists) { $count += $header.Range.Revisions.Count }
+        }
+    }
+    return $count
 }
 
 try {
@@ -60,23 +157,63 @@ try {
         # -Encoding UTF8 is required: Windows PowerShell 5.1 otherwise reads
         # BOM-less UTF-8 sidecars as ANSI and garbles non-ASCII expectations.
         $expected = Get-Content -LiteralPath $expectationFile.FullName -Raw -Encoding UTF8 | ConvertFrom-Json
-        $expectedAccepted = Get-NormalizedText $expected.expectedAcceptedText
-        $expectedRejected = Get-NormalizedText $expected.expectedRejectedText
+        $fidelity = if ($expected.textFidelity) { [string]$expected.textFidelity } else { 'exact' }
+        $assertionMode = if ($expected.assertionMode) { [string]$expected.assertionMode } else { 'exact' }
+        $assertionScope = if ($expected.assertionScope) { [string]$expected.assertionScope } else { 'document' }
+        if ($assertionMode -eq 'exact') {
+            $expectedAccepted = Get-ComparableText $expected.expectedAcceptedText $fidelity
+            $expectedRejected = Get-ComparableText $expected.expectedRejectedText $fidelity
+        }
+        elseif ($assertionMode -eq 'word-source-exact') {
+            if (-not $SourcesDir) { throw "${name}: word-source-exact requires -SourcesDir" }
+            $sourcePath = Join-Path (Resolve-Path -LiteralPath $SourcesDir) "$($expected.sourceId).docx"
+            $sourceDocument = $null
+            try {
+                $sourceDocument = Open-FixtureDocument $word ([string]$sourcePath)
+                $expectedRejected = Get-ScopeText $sourceDocument $assertionScope $fidelity
+            }
+            finally {
+                if ($null -ne $sourceDocument) { $sourceDocument.Close(0) | Out-Null }
+            }
+            $replacements = if ($expected.replacements) { @($expected.replacements) } else { @(@{
+                originalTarget = $expected.originalTarget
+                modifiedTarget = $expected.modifiedTarget
+            }) }
+            $expectedAccepted = $expectedRejected
+            foreach ($replacement in $replacements) {
+                $targetCount = [regex]::Matches($expectedAccepted, [regex]::Escape([string]$replacement.originalTarget)).Count
+                if ($targetCount -ne 1) {
+                    throw "${name}: expected target '$($replacement.originalTarget)' occurs $targetCount times in Word's source $assertionScope text"
+                }
+                $expectedAccepted = $expectedAccepted.Replace([string]$replacement.originalTarget, [string]$replacement.modifiedTarget)
+            }
+        }
         $caseFailed = $false
         $document = $null
 
+        if ($expected.untouchedPartSha256 -and -not (Test-UntouchedPartHashes $docxPath $expected.untouchedPartSha256 $name)) {
+            $failures++
+            $results += "FAIL $name"
+            continue
+        }
+
         # Phase 1: open cleanly, revisions present, accept-all matches intent.
         try {
             $document = Open-FixtureDocument $word ([string]$docxPath)
-            $revisionCount = $document.Revisions.Count
+            $revisionCount = Get-ScopeRevisionCount $document $assertionScope
             if ($revisionCount -lt 1) {
                 Write-Output "FAIL ${name}: Word sees no tracked revisions"
                 $caseFailed = $true
             }
             else {
                 $document.AcceptAllRevisions()
-                $acceptedText = Get-NormalizedText $document.Content.Text
-                if ($acceptedText -ne $expectedAccepted) {
+                $acceptedText = Get-ScopeText $document $assertionScope $fidelity
+                if ($assertionMode -eq 'contains') {
+                    if (-not (Test-ContainsAssertions $acceptedText $expected.expectedAcceptedContains $expected.expectedAcceptedAbsent 'accept-all' $name)) {
+                        $caseFailed = $true
+                    }
+                }
+                elseif ($acceptedText -ne $expectedAccepted) {
                     Write-Output "FAIL ${name}: accept-all mismatch"
                     Write-Output "  expected: $expectedAccepted"
                     Write-Output "  actual:   $acceptedText"
@@ -97,8 +234,13 @@ try {
             try {
                 $document = Open-FixtureDocument $word ([string]$docxPath)
                 $document.RejectAllRevisions()
-                $rejectedText = Get-NormalizedText $document.Content.Text
-                if ($rejectedText -ne $expectedRejected) {
+                $rejectedText = Get-ScopeText $document $assertionScope $fidelity
+                if ($assertionMode -eq 'contains') {
+                    if (-not (Test-ContainsAssertions $rejectedText $expected.expectedRejectedContains $expected.expectedRejectedAbsent 'reject-all' $name)) {
+                        $caseFailed = $true
+                    }
+                }
+                elseif ($rejectedText -ne $expectedRejected) {
                     Write-Output "FAIL ${name}: reject-all mismatch"
                     Write-Output "  expected: $expectedRejected"
                     Write-Output "  actual:   $rejectedText"

package/scripts/word-com-suite.ps1

@@ -0,0 +1,19 @@
+param(
+    [string]$OutputDir = "tmp/word-validation"
+)
+
+$ErrorActionPreference = 'Stop'
+$repoRoot = Split-Path -Parent $PSScriptRoot
+$exportScript = Join-Path $PSScriptRoot 'export-validation-fixtures.mjs'
+$differentialScript = Join-Path $PSScriptRoot 'word-com-differential.ps1'
+Push-Location $repoRoot
+try {
+    & node $exportScript --output-dir $OutputDir
+    if ($LASTEXITCODE -ne 0) { exit $LASTEXITCODE }
+
+    & powershell -NoProfile -ExecutionPolicy Bypass -File $differentialScript -FixturesDir $OutputDir
+    exit $LASTEXITCODE
+}
+finally {
+    Pop-Location
+}