docrev 0.11.1 → 0.11.3

package/lib/word.ts

@@ -1,526 +1,526 @@
-/**
- * Word document extraction utilities
- * Handle reading text, comments, and anchors from .docx files
- */
-
-import * as fs from 'fs';
-import * as path from 'path';
-import AdmZip from 'adm-zip';
-import type { WordComment, CommentAnchor, WordMetadata, TrackChangesResult } from './types.js';
-import {
-  openDocx,
-  readPartText,
-  buildCommentAnchorModel,
-  extractComments,
-  walkBody,
-  type FlowItem,
-} from './ooxml.js';
-
-// =============================================================================
-// Constants
-// =============================================================================
-
-/** Characters of context to extract around comment anchors */
-const ANCHOR_CONTEXT_SIZE = 100;
-
-/** Characters of context before comment range start */
-const CONTEXT_BEFORE_SIZE = 500;
-
-// =============================================================================
-// Public API
-// =============================================================================
-
-/**
- * Extract comments from Word document's comments.xml
- * @param docxPath - Path to .docx file
- * @returns Array of extracted comments
- * @throws {TypeError} If docxPath is not a string
- * @throws {Error} If file not found or invalid docx
- */
-export async function extractWordComments(docxPath: string): Promise<WordComment[]> {
-  if (typeof docxPath !== 'string') {
-    throw new TypeError(`docxPath must be a string, got ${typeof docxPath}`);
-  }
-  if (!fs.existsSync(docxPath)) {
-    throw new Error(`File not found: ${docxPath}`);
-  }
-
-  const zip = openDocx(docxPath);
-  return extractComments(zip)
-    .filter((c) => c.id && c.text)
-    .map((c) => ({ id: c.id, author: c.author, date: c.date, text: c.text }));
-}
-
-/**
- * Extract comment anchors (where comments are attached) from document.xml
- * Returns mapping of comment ID to the text they're anchored to
- * @param docxPath - Path to .docx file
- * @returns Map of comment ID to anchor info
- * @throws {TypeError} If docxPath is not a string
- * @throws {Error} If invalid docx structure
- */
-export async function extractCommentAnchors(docxPath: string): Promise<Map<string, CommentAnchor>> {
-  if (typeof docxPath !== 'string') {
-    throw new TypeError(`docxPath must be a string, got ${typeof docxPath}`);
-  }
-
-  const zip = openDocx(docxPath);
-  if (!zip.getEntry('word/document.xml')) {
-    throw new Error('Invalid docx: no document.xml');
-  }
-
-  const { fullDocText, comments } = buildCommentAnchorModel(zip);
-  const anchors = new Map<string, CommentAnchor>();
-
-  for (const range of comments) {
-    anchors.set(range.id, {
-      text: range.anchor,
-      context: fullDocText.slice(Math.max(0, range.start - CONTEXT_BEFORE_SIZE), range.start).slice(-ANCHOR_CONTEXT_SIZE),
-    });
-  }
-
-  return anchors;
-}
-
-/**
- * Extract plain text from Word document (strips track change markup)
- * @param docxPath - Path to .docx file
- * @returns Extracted plain text (accepted changes applied)
- * @throws {TypeError} If docxPath is not a string
- * @throws {Error} If file not found
- */
-export async function extractTextFromWord(docxPath: string): Promise<string> {
-  if (typeof docxPath !== 'string') {
-    throw new TypeError(`docxPath must be a string, got ${typeof docxPath}`);
-  }
-  const result = await extractPlainTextWithTrackChanges(docxPath);
-  // Strip CriticMarkup: accept insertions, remove deletions, apply substitutions
-  let text = result.text;
-  text = text.replace(/\{~~[^~]*~>([^~]*)~~\}/g, '$1');  // substitutions → new
-  text = text.replace(/\{\+\+([^+]*)\+\+\}/g, '$1');      // insertions → keep
-  text = text.replace(/\{--[^}]*--\}/g, '');               // deletions → remove
-  return text;
-}
-
-/**
- * Get document metadata from Word file
- * @param docxPath - Path to .docx file
- * @returns Document metadata
- * @throws {TypeError} If docxPath is not a string
- */
-export async function getWordMetadata(docxPath: string): Promise<WordMetadata> {
-  if (typeof docxPath !== 'string') {
-    throw new TypeError(`docxPath must be a string, got ${typeof docxPath}`);
-  }
-
-  const zip = new AdmZip(docxPath);
-  const coreEntry = zip.getEntry('docProps/core.xml');
-
-  if (!coreEntry) {
-    return {};
-  }
-
-  const coreXml = zip.readAsText(coreEntry);
-  const metadata: WordMetadata = {};
-
-  // Extract common metadata fields
-  const patterns: Record<string, RegExp> = {
-    title: /<dc:title>([^<]*)<\/dc:title>/,
-    author: /<dc:creator>([^<]*)<\/dc:creator>/,
-    created: /<dcterms:created[^>]*>([^<]*)<\/dcterms:created>/,
-    modified: /<dcterms:modified[^>]*>([^<]*)<\/dcterms:modified>/,
-  };
-
-  for (const [key, pattern] of Object.entries(patterns)) {
-    const match = coreXml.match(pattern);
-    if (match) {
-      (metadata as any)[key] = match[1];
-    }
-  }
-
-  return metadata;
-}
-
-/**
- * Check if file is a valid Word document
- * @param filePath - Path to file to check
- * @returns True if valid .docx file
- */
-export function isWordDocument(filePath: string): boolean {
-  if (typeof filePath !== 'string') return false;
-  if (!fs.existsSync(filePath)) return false;
-  if (!filePath.toLowerCase().endsWith('.docx')) return false;
-
-  try {
-    const zip = new AdmZip(filePath);
-    return zip.getEntry('word/document.xml') !== null;
-  } catch {
-    return false;
-  }
-}
-
-/**
- * Extract text content from XML element, handling nested elements
- * @param xml - XML string
- * @returns Plain text content
- */
-function extractTextFromXml(xml: string): string {
-  let text = '';
-  // Match w:t elements (regular text)
-  const textPattern = /<w:t[^>]*>([^<]*)<\/w:t>/g;
-  let match: RegExpExecArray | null;
-  while ((match = textPattern.exec(xml)) !== null) {
-    text += match[1];
-  }
-  // Also match w:delText (deleted text)
-  const delTextPattern = /<w:delText[^>]*>([^<]*)<\/w:delText>/g;
-  while ((match = delTextPattern.exec(xml)) !== null) {
-    text += match[1];
-  }
-  return text;
-}
-
-/**
- * Extract track changes (insertions and deletions) from Word document
- * Converts Word's w:ins and w:del elements to CriticMarkup format
- *
- * @param docxPath - Path to Word document
- * @returns Track changes result with content and stats
- */
-export async function extractTrackChanges(docxPath: string): Promise<TrackChangesResult> {
-  if (!fs.existsSync(docxPath)) {
-    throw new Error(`File not found: ${docxPath}`);
-  }
-
-  const zip = new AdmZip(docxPath);
-  const documentEntry = zip.getEntry('word/document.xml');
-
-  if (!documentEntry) {
-    throw new Error('Invalid docx: no document.xml');
-  }
-
-  let xml = zip.readAsText(documentEntry);
-  let insertions = 0;
-  let deletions = 0;
-
-  // Check if there are any track changes
-  const hasInsertions = xml.includes('<w:ins ');
-  const hasDeletions = xml.includes('<w:del ');
-
-  if (!hasInsertions && !hasDeletions) {
-    return { hasTrackChanges: false, content: null, stats: { insertions: 0, deletions: 0 } };
-  }
-
-  // Process insertions: <w:ins ...>...</w:ins> -> {++...++}
-  // Match the full w:ins element including nested content
-  xml = xml.replace(/<w:ins\b[^>]*>([\s\S]*?)<\/w:ins>/g, (match, content) => {
-    const text = extractTextFromXml(content);
-    if (text.trim()) {
-      insertions++;
-      return `{++${text}++}`;
-    }
-    return text;
-  });
-
-  // Process deletions: <w:del ...>...</w:del> -> {--...--}
-  xml = xml.replace(/<w:del\b[^>]*>([\s\S]*?)<\/w:del>/g, (match, content) => {
-    const text = extractTextFromXml(content);
-    if (text.trim()) {
-      deletions++;
-      return `{--${text}--}`;
-    }
-    return '';
-  });
-
-  return {
-    hasTrackChanges: true,
-    content: xml,
-    stats: { insertions, deletions },
-  };
-}
-
-/**
- * Extract a single marker's content starting at position i.
- * Returns { content, end } where end is the position after the closing marker,
- * or null if no valid closing marker found.
- */
-function extractMarker(text: string, i: number, open: string, close: string): { content: string; end: number } | null {
-  if (!text.startsWith(open, i)) return null;
-  const start = i + open.length;
-  const closeIdx = text.indexOf(close, start);
-  if (closeIdx === -1) return null;
-  return { content: text.slice(start, closeIdx), end: closeIdx + close.length };
-}
-
-/**
- * Greedily collect consecutive markers of the same type.
- * E.g. {++a++}{++b++}{++c++} → "abc", advancing past all three.
- */
-function collectConsecutive(text: string, i: number, open: string, close: string): { content: string; end: number } | null {
-  const first = extractMarker(text, i, open, close);
-  if (!first) return null;
-
-  let content = first.content;
-  let end = first.end;
-
-  while (end < text.length) {
-    const next = extractMarker(text, end, open, close);
-    if (!next) break;
-    content += next.content;
-    end = next.end;
-  }
-
-  return { content, end };
-}
-
-/**
- * Scan text for adjacent CriticMarkup markers and:
- * 1. Merge consecutive same-type markers: {++a++}{++b++} → {++ab++}
- * 2. Merge adjacent del+ins or ins+del into substitutions: {--old--}{++new++} → {~~old~>new~~}
- *
- * Uses a linear scanner — no regex backtracking, no ambiguity.
- */
-function mergeAdjacentMarkers(text: string): string {
-  let result = '';
-  let i = 0;
-
-  while (i < text.length) {
-    // --- Deletion block ---
-    if (text.startsWith('{--', i)) {
-      const del = collectConsecutive(text, i, '{--', '--}');
-      if (!del) { result += text[i]; i++; continue; }
-
-      // Skip spaces, then check for adjacent insertion
-      let j = del.end;
-      while (j < text.length && text[j] === ' ') j++;
-
-      const ins = collectConsecutive(text, j, '{++', '++}');
-      if (ins) {
-        // Merge into substitution
-        const trailing = del.content.endsWith(' ') || ins.content.endsWith(' ');
-        result += `{~~${del.content.trimEnd()}~>${ins.content.trimEnd()}~~}${trailing ? ' ' : ''}`;
-        i = ins.end;
-      } else {
-        // Emit merged deletion
-        result += `{--${del.content}--}`;
-        i = del.end;
-      }
-      continue;
-    }
-
-    // --- Insertion block ---
-    if (text.startsWith('{++', i)) {
-      const ins = collectConsecutive(text, i, '{++', '++}');
-      if (!ins) { result += text[i]; i++; continue; }
-
-      // Skip spaces, then check for adjacent deletion
-      let j = ins.end;
-      while (j < text.length && text[j] === ' ') j++;
-
-      const del = collectConsecutive(text, j, '{--', '--}');
-      if (del) {
-        // Merge into substitution (del → ins order in output)
-        const trailing = del.content.endsWith(' ') || ins.content.endsWith(' ');
-        result += `{~~${del.content.trimEnd()}~>${ins.content.trimEnd()}~~}${trailing ? ' ' : ''}`;
-        i = del.end;
-      } else {
-        // Emit merged insertion
-        result += `{++${ins.content}++}`;
-        i = ins.end;
-      }
-      continue;
-    }
-
-    result += text[i];
-    i++;
-  }
-
-  return result;
-}
-
-/**
- * Extract plain text from Word XML with track changes preserved as CriticMarkup.
- * This is a pandoc-free fallback that reads document.xml directly.
- *
- * Converts:
- *   <w:ins> content </w:ins>  →  {++text++}
- *   <w:del> content </w:del>  →  {--text--}
- *
- * Also detects headings (w:pStyle Heading1-6) and outputs markdown # syntax.
- *
- * @param docxPath - Path to Word document
- * @returns Plain text with CriticMarkup and stats
- */
-export async function extractPlainTextWithTrackChanges(docxPath: string): Promise<{
-  text: string;
-  hasTrackChanges: boolean;
-  stats: { insertions: number; deletions: number };
-}> {
-  if (!fs.existsSync(docxPath)) {
-    throw new Error(`File not found: ${docxPath}`);
-  }
-
-  const zip = openDocx(docxPath);
-  const docXml = readPartText(zip, 'word/document.xml');
-  if (docXml === null) {
-    throw new Error('Invalid docx: no document.xml');
-  }
-
-  let insertions = 0;
-  let deletions = 0;
-  const paragraphs: string[] = [];
-
-  // One ordered walk drives everything: paragraph and heading boundaries,
-  // run text (entities already decoded, tabs/breaks rendered), and the
-  // track-change spans that become CriticMarkup. Field codes (w:instrText)
-  // never reach the text because the walker only reads w:t / w:delText.
-  let paraOut = '';
-  let headingLevel = 0;
-  let mode: 'normal' | 'ins' | 'del' = 'normal';
-  let buffer = '';
-
-  const flushSpan = (open: string, close: string, isIns: boolean) => {
-    if (buffer.trim()) {
-      if (isIns) insertions++;
-      else deletions++;
-      paraOut += `${open}${buffer}${close}`;
-    } else if (buffer.length > 0) {
-      // Whitespace-only edits are kept as plain text to preserve spacing.
-      paraOut += buffer;
-    }
-    buffer = '';
-  };
-
-  const endParagraph = () => {
-    let text = mergeAdjacentMarkers(paraOut);
-    text = text.replace(/ {2,}/g, ' ');
-    if (text.trim()) {
-      paragraphs.push(
-        headingLevel >= 1 && headingLevel <= 6 ? '#'.repeat(headingLevel) + ' ' + text.trim() : text,
-      );
-    }
-    paraOut = '';
-    headingLevel = 0;
-    mode = 'normal';
-    buffer = '';
-  };
-
-  for (const item of walkBody(docXml) as FlowItem[]) {
-    switch (item.kind) {
-      case 'paraStart':
-        paraOut = '';
-        headingLevel = item.level;
-        mode = 'normal';
-        buffer = '';
-        break;
-      case 'paraEnd':
-        endParagraph();
-        break;
-      case 'text':
-        if (mode === 'normal') paraOut += item.text;
-        else buffer += item.text;
-        break;
-      case 'insStart':
-        mode = 'ins';
-        buffer = '';
-        break;
-      case 'insEnd':
-        flushSpan('{++', '++}', true);
-        mode = 'normal';
-        break;
-      case 'delStart':
-        mode = 'del';
-        buffer = '';
-        break;
-      case 'delEnd':
-        flushSpan('{--', '--}', false);
-        mode = 'normal';
-        break;
-      default:
-        break;
-    }
-  }
-
-  return {
-    text: paragraphs.join('\n\n'),
-    hasTrackChanges: insertions > 0 || deletions > 0,
-    stats: { insertions, deletions },
-  };
-}
-
-interface ExtractWithTrackChangesOptions {
-  mediaDir?: string;
-}
-
-/**
- * Extract Word document content with track changes preserved as CriticMarkup
- * Uses pandoc with track-changes=all option to preserve insertions/deletions
- *
- * @param docxPath - Path to Word document
- * @param options - Options
- * @returns Track changes result with text and stats
- */
-export async function extractWithTrackChanges(
-  docxPath: string,
-  options: ExtractWithTrackChangesOptions = {}
-): Promise<{ text: string; hasTrackChanges: boolean; stats: { insertions: number; deletions: number } }> {
-  const { mediaDir } = options;
-
-  if (!fs.existsSync(docxPath)) {
-    throw new Error(`File not found: ${docxPath}`);
-  }
-
-  const { execSync } = await import('child_process');
-
-  // Use pandoc with --track-changes=all to preserve track changes
-  // This outputs insertions as [insertion]{.insertion} and deletions as [deletion]{.deletion}
-  let pandocArgs = `"${docxPath}" -t markdown --wrap=none --track-changes=all`;
-  if (mediaDir) {
-    pandocArgs += ` --extract-media="${mediaDir}"`;
-  }
-
-  let text: string;
-  try {
-    text = execSync(`pandoc ${pandocArgs}`, {
-      encoding: 'utf-8',
-      maxBuffer: 50 * 1024 * 1024,
-    });
-  } catch (err: any) {
-    throw new Error(`Pandoc extraction failed: ${err.message}`);
-  }
-
-  // Count track changes from pandoc output
-  let insertions = 0;
-  let deletions = 0;
-
-  // Pandoc outputs track changes as:
-  // [inserted text]{.insertion author="..."}
-  // [deleted text]{.deletion author="..."}
-
-  // Convert pandoc's track change format to CriticMarkup
-  // Insertions: [text]{.insertion ...} -> {++text++}
-  text = text.replace(/\[([^\]]*)\]\{\.insertion[^}]*\}/g, (match, content) => {
-    if (content.trim()) {
-      insertions++;
-      return `{++${content}++}`;
-    }
-    return '';
-  });
-
-  // Deletions: [text]{.deletion ...} -> {--text--}
-  text = text.replace(/\[([^\]]*)\]\{\.deletion[^}]*\}/g, (match, content) => {
-    if (content.trim()) {
-      deletions++;
-      return `{--${content}--}`;
-    }
-    return '';
-  });
-
-  const hasTrackChanges = insertions > 0 || deletions > 0;
-
-  return {
-    text,
-    hasTrackChanges,
-    stats: { insertions, deletions },
-  };
-}
+/**
+ * Word document extraction utilities
+ * Handle reading text, comments, and anchors from .docx files
+ */
+
+import * as fs from 'fs';
+import * as path from 'path';
+import AdmZip from 'adm-zip';
+import type { WordComment, CommentAnchor, WordMetadata, TrackChangesResult } from './types.js';
+import {
+  openDocx,
+  readPartText,
+  buildCommentAnchorModel,
+  extractComments,
+  walkBody,
+  type FlowItem,
+} from './ooxml.js';
+
+// =============================================================================
+// Constants
+// =============================================================================
+
+/** Characters of context to extract around comment anchors */
+const ANCHOR_CONTEXT_SIZE = 100;
+
+/** Characters of context before comment range start */
+const CONTEXT_BEFORE_SIZE = 500;
+
+// =============================================================================
+// Public API
+// =============================================================================
+
+/**
+ * Extract comments from Word document's comments.xml
+ * @param docxPath - Path to .docx file
+ * @returns Array of extracted comments
+ * @throws {TypeError} If docxPath is not a string
+ * @throws {Error} If file not found or invalid docx
+ */
+export async function extractWordComments(docxPath: string): Promise<WordComment[]> {
+  if (typeof docxPath !== 'string') {
+    throw new TypeError(`docxPath must be a string, got ${typeof docxPath}`);
+  }
+  if (!fs.existsSync(docxPath)) {
+    throw new Error(`File not found: ${docxPath}`);
+  }
+
+  const zip = openDocx(docxPath);
+  return extractComments(zip)
+    .filter((c) => c.id && c.text)
+    .map((c) => ({ id: c.id, author: c.author, date: c.date, text: c.text }));
+}
+
+/**
+ * Extract comment anchors (where comments are attached) from document.xml
+ * Returns mapping of comment ID to the text they're anchored to
+ * @param docxPath - Path to .docx file
+ * @returns Map of comment ID to anchor info
+ * @throws {TypeError} If docxPath is not a string
+ * @throws {Error} If invalid docx structure
+ */
+export async function extractCommentAnchors(docxPath: string): Promise<Map<string, CommentAnchor>> {
+  if (typeof docxPath !== 'string') {
+    throw new TypeError(`docxPath must be a string, got ${typeof docxPath}`);
+  }
+
+  const zip = openDocx(docxPath);
+  if (!zip.getEntry('word/document.xml')) {
+    throw new Error('Invalid docx: no document.xml');
+  }
+
+  const { fullDocText, comments } = buildCommentAnchorModel(zip);
+  const anchors = new Map<string, CommentAnchor>();
+
+  for (const range of comments) {
+    anchors.set(range.id, {
+      text: range.anchor,
+      context: fullDocText.slice(Math.max(0, range.start - CONTEXT_BEFORE_SIZE), range.start).slice(-ANCHOR_CONTEXT_SIZE),
+    });
+  }
+
+  return anchors;
+}
+
+/**
+ * Extract plain text from Word document (strips track change markup)
+ * @param docxPath - Path to .docx file
+ * @returns Extracted plain text (accepted changes applied)
+ * @throws {TypeError} If docxPath is not a string
+ * @throws {Error} If file not found
+ */
+export async function extractTextFromWord(docxPath: string): Promise<string> {
+  if (typeof docxPath !== 'string') {
+    throw new TypeError(`docxPath must be a string, got ${typeof docxPath}`);
+  }
+  const result = await extractPlainTextWithTrackChanges(docxPath);
+  // Strip CriticMarkup: accept insertions, remove deletions, apply substitutions
+  let text = result.text;
+  text = text.replace(/\{~~[^~]*~>([^~]*)~~\}/g, '$1');  // substitutions → new
+  text = text.replace(/\{\+\+([^+]*)\+\+\}/g, '$1');      // insertions → keep
+  text = text.replace(/\{--[^}]*--\}/g, '');               // deletions → remove
+  return text;
+}
+
+/**
+ * Get document metadata from Word file
+ * @param docxPath - Path to .docx file
+ * @returns Document metadata
+ * @throws {TypeError} If docxPath is not a string
+ */
+export async function getWordMetadata(docxPath: string): Promise<WordMetadata> {
+  if (typeof docxPath !== 'string') {
+    throw new TypeError(`docxPath must be a string, got ${typeof docxPath}`);
+  }
+
+  const zip = new AdmZip(docxPath);
+  const coreEntry = zip.getEntry('docProps/core.xml');
+
+  if (!coreEntry) {
+    return {};
+  }
+
+  const coreXml = zip.readAsText(coreEntry);
+  const metadata: WordMetadata = {};
+
+  // Extract common metadata fields
+  const patterns: Record<string, RegExp> = {
+    title: /<dc:title>([^<]*)<\/dc:title>/,
+    author: /<dc:creator>([^<]*)<\/dc:creator>/,
+    created: /<dcterms:created[^>]*>([^<]*)<\/dcterms:created>/,
+    modified: /<dcterms:modified[^>]*>([^<]*)<\/dcterms:modified>/,
+  };
+
+  for (const [key, pattern] of Object.entries(patterns)) {
+    const match = coreXml.match(pattern);
+    if (match) {
+      (metadata as any)[key] = match[1];
+    }
+  }
+
+  return metadata;
+}
+
+/**
+ * Check if file is a valid Word document
+ * @param filePath - Path to file to check
+ * @returns True if valid .docx file
+ */
+export function isWordDocument(filePath: string): boolean {
+  if (typeof filePath !== 'string') return false;
+  if (!fs.existsSync(filePath)) return false;
+  if (!filePath.toLowerCase().endsWith('.docx')) return false;
+
+  try {
+    const zip = new AdmZip(filePath);
+    return zip.getEntry('word/document.xml') !== null;
+  } catch {
+    return false;
+  }
+}
+
+/**
+ * Extract text content from XML element, handling nested elements
+ * @param xml - XML string
+ * @returns Plain text content
+ */
+function extractTextFromXml(xml: string): string {
+  let text = '';
+  // Match w:t elements (regular text)
+  const textPattern = /<w:t[^>]*>([^<]*)<\/w:t>/g;
+  let match: RegExpExecArray | null;
+  while ((match = textPattern.exec(xml)) !== null) {
+    text += match[1];
+  }
+  // Also match w:delText (deleted text)
+  const delTextPattern = /<w:delText[^>]*>([^<]*)<\/w:delText>/g;
+  while ((match = delTextPattern.exec(xml)) !== null) {
+    text += match[1];
+  }
+  return text;
+}
+
+/**
+ * Extract track changes (insertions and deletions) from Word document
+ * Converts Word's w:ins and w:del elements to CriticMarkup format
+ *
+ * @param docxPath - Path to Word document
+ * @returns Track changes result with content and stats
+ */
+export async function extractTrackChanges(docxPath: string): Promise<TrackChangesResult> {
+  if (!fs.existsSync(docxPath)) {
+    throw new Error(`File not found: ${docxPath}`);
+  }
+
+  const zip = new AdmZip(docxPath);
+  const documentEntry = zip.getEntry('word/document.xml');
+
+  if (!documentEntry) {
+    throw new Error('Invalid docx: no document.xml');
+  }
+
+  let xml = zip.readAsText(documentEntry);
+  let insertions = 0;
+  let deletions = 0;
+
+  // Check if there are any track changes
+  const hasInsertions = xml.includes('<w:ins ');
+  const hasDeletions = xml.includes('<w:del ');
+
+  if (!hasInsertions && !hasDeletions) {
+    return { hasTrackChanges: false, content: null, stats: { insertions: 0, deletions: 0 } };
+  }
+
+  // Process insertions: <w:ins ...>...</w:ins> -> {++...++}
+  // Match the full w:ins element including nested content
+  xml = xml.replace(/<w:ins\b[^>]*>([\s\S]*?)<\/w:ins>/g, (match, content) => {
+    const text = extractTextFromXml(content);
+    if (text.trim()) {
+      insertions++;
+      return `{++${text}++}`;
+    }
+    return text;
+  });
+
+  // Process deletions: <w:del ...>...</w:del> -> {--...--}
+  xml = xml.replace(/<w:del\b[^>]*>([\s\S]*?)<\/w:del>/g, (match, content) => {
+    const text = extractTextFromXml(content);
+    if (text.trim()) {
+      deletions++;
+      return `{--${text}--}`;
+    }
+    return '';
+  });
+
+  return {
+    hasTrackChanges: true,
+    content: xml,
+    stats: { insertions, deletions },
+  };
+}
+
+/**
+ * Extract a single marker's content starting at position i.
+ * Returns { content, end } where end is the position after the closing marker,
+ * or null if no valid closing marker found.
+ */
+function extractMarker(text: string, i: number, open: string, close: string): { content: string; end: number } | null {
+  if (!text.startsWith(open, i)) return null;
+  const start = i + open.length;
+  const closeIdx = text.indexOf(close, start);
+  if (closeIdx === -1) return null;
+  return { content: text.slice(start, closeIdx), end: closeIdx + close.length };
+}
+
+/**
+ * Greedily collect consecutive markers of the same type.
+ * E.g. {++a++}{++b++}{++c++} → "abc", advancing past all three.
+ */
+function collectConsecutive(text: string, i: number, open: string, close: string): { content: string; end: number } | null {
+  const first = extractMarker(text, i, open, close);
+  if (!first) return null;
+
+  let content = first.content;
+  let end = first.end;
+
+  while (end < text.length) {
+    const next = extractMarker(text, end, open, close);
+    if (!next) break;
+    content += next.content;
+    end = next.end;
+  }
+
+  return { content, end };
+}
+
+/**
+ * Scan text for adjacent CriticMarkup markers and:
+ * 1. Merge consecutive same-type markers: {++a++}{++b++} → {++ab++}
+ * 2. Merge adjacent del+ins or ins+del into substitutions: {--old--}{++new++} → {~~old~>new~~}
+ *
+ * Uses a linear scanner — no regex backtracking, no ambiguity.
+ */
+function mergeAdjacentMarkers(text: string): string {
+  let result = '';
+  let i = 0;
+
+  while (i < text.length) {
+    // --- Deletion block ---
+    if (text.startsWith('{--', i)) {
+      const del = collectConsecutive(text, i, '{--', '--}');
+      if (!del) { result += text[i]; i++; continue; }
+
+      // Skip spaces, then check for adjacent insertion
+      let j = del.end;
+      while (j < text.length && text[j] === ' ') j++;
+
+      const ins = collectConsecutive(text, j, '{++', '++}');
+      if (ins) {
+        // Merge into substitution
+        const trailing = del.content.endsWith(' ') || ins.content.endsWith(' ');
+        result += `{~~${del.content.trimEnd()}~>${ins.content.trimEnd()}~~}${trailing ? ' ' : ''}`;
+        i = ins.end;
+      } else {
+        // Emit merged deletion
+        result += `{--${del.content}--}`;
+        i = del.end;
+      }
+      continue;
+    }
+
+    // --- Insertion block ---
+    if (text.startsWith('{++', i)) {
+      const ins = collectConsecutive(text, i, '{++', '++}');
+      if (!ins) { result += text[i]; i++; continue; }
+
+      // Skip spaces, then check for adjacent deletion
+      let j = ins.end;
+      while (j < text.length && text[j] === ' ') j++;
+
+      const del = collectConsecutive(text, j, '{--', '--}');
+      if (del) {
+        // Merge into substitution (del → ins order in output)
+        const trailing = del.content.endsWith(' ') || ins.content.endsWith(' ');
+        result += `{~~${del.content.trimEnd()}~>${ins.content.trimEnd()}~~}${trailing ? ' ' : ''}`;
+        i = del.end;
+      } else {
+        // Emit merged insertion
+        result += `{++${ins.content}++}`;
+        i = ins.end;
+      }
+      continue;
+    }
+
+    result += text[i];
+    i++;
+  }
+
+  return result;
+}
+
+/**
+ * Extract plain text from Word XML with track changes preserved as CriticMarkup.
+ * This is a pandoc-free fallback that reads document.xml directly.
+ *
+ * Converts:
+ *   <w:ins> content </w:ins>  →  {++text++}
+ *   <w:del> content </w:del>  →  {--text--}
+ *
+ * Also detects headings (w:pStyle Heading1-6) and outputs markdown # syntax.
+ *
+ * @param docxPath - Path to Word document
+ * @returns Plain text with CriticMarkup and stats
+ */
+export async function extractPlainTextWithTrackChanges(docxPath: string): Promise<{
+  text: string;
+  hasTrackChanges: boolean;
+  stats: { insertions: number; deletions: number };
+}> {
+  if (!fs.existsSync(docxPath)) {
+    throw new Error(`File not found: ${docxPath}`);
+  }
+
+  const zip = openDocx(docxPath);
+  const docXml = readPartText(zip, 'word/document.xml');
+  if (docXml === null) {
+    throw new Error('Invalid docx: no document.xml');
+  }
+
+  let insertions = 0;
+  let deletions = 0;
+  const paragraphs: string[] = [];
+
+  // One ordered walk drives everything: paragraph and heading boundaries,
+  // run text (entities already decoded, tabs/breaks rendered), and the
+  // track-change spans that become CriticMarkup. Field codes (w:instrText)
+  // never reach the text because the walker only reads w:t / w:delText.
+  let paraOut = '';
+  let headingLevel = 0;
+  let mode: 'normal' | 'ins' | 'del' = 'normal';
+  let buffer = '';
+
+  const flushSpan = (open: string, close: string, isIns: boolean) => {
+    if (buffer.trim()) {
+      if (isIns) insertions++;
+      else deletions++;
+      paraOut += `${open}${buffer}${close}`;
+    } else if (buffer.length > 0) {
+      // Whitespace-only edits are kept as plain text to preserve spacing.
+      paraOut += buffer;
+    }
+    buffer = '';
+  };
+
+  const endParagraph = () => {
+    let text = mergeAdjacentMarkers(paraOut);
+    text = text.replace(/ {2,}/g, ' ');
+    if (text.trim()) {
+      paragraphs.push(
+        headingLevel >= 1 && headingLevel <= 6 ? '#'.repeat(headingLevel) + ' ' + text.trim() : text,
+      );
+    }
+    paraOut = '';
+    headingLevel = 0;
+    mode = 'normal';
+    buffer = '';
+  };
+
+  for (const item of walkBody(docXml) as FlowItem[]) {
+    switch (item.kind) {
+      case 'paraStart':
+        paraOut = '';
+        headingLevel = item.level;
+        mode = 'normal';
+        buffer = '';
+        break;
+      case 'paraEnd':
+        endParagraph();
+        break;
+      case 'text':
+        if (mode === 'normal') paraOut += item.text;
+        else buffer += item.text;
+        break;
+      case 'insStart':
+        mode = 'ins';
+        buffer = '';
+        break;
+      case 'insEnd':
+        flushSpan('{++', '++}', true);
+        mode = 'normal';
+        break;
+      case 'delStart':
+        mode = 'del';
+        buffer = '';
+        break;
+      case 'delEnd':
+        flushSpan('{--', '--}', false);
+        mode = 'normal';
+        break;
+      default:
+        break;
+    }
+  }
+
+  return {
+    text: paragraphs.join('\n\n'),
+    hasTrackChanges: insertions > 0 || deletions > 0,
+    stats: { insertions, deletions },
+  };
+}
+
+interface ExtractWithTrackChangesOptions {
+  mediaDir?: string;
+}
+
+/**
+ * Extract Word document content with track changes preserved as CriticMarkup
+ * Uses pandoc with track-changes=all option to preserve insertions/deletions
+ *
+ * @param docxPath - Path to Word document
+ * @param options - Options
+ * @returns Track changes result with text and stats
+ */
+export async function extractWithTrackChanges(
+  docxPath: string,
+  options: ExtractWithTrackChangesOptions = {}
+): Promise<{ text: string; hasTrackChanges: boolean; stats: { insertions: number; deletions: number } }> {
+  const { mediaDir } = options;
+
+  if (!fs.existsSync(docxPath)) {
+    throw new Error(`File not found: ${docxPath}`);
+  }
+
+  const { execSync } = await import('child_process');
+
+  // Use pandoc with --track-changes=all to preserve track changes
+  // This outputs insertions as [insertion]{.insertion} and deletions as [deletion]{.deletion}
+  let pandocArgs = `"${docxPath}" -t markdown --wrap=none --track-changes=all`;
+  if (mediaDir) {
+    pandocArgs += ` --extract-media="${mediaDir}"`;
+  }
+
+  let text: string;
+  try {
+    text = execSync(`pandoc ${pandocArgs}`, {
+      encoding: 'utf-8',
+      maxBuffer: 50 * 1024 * 1024,
+    });
+  } catch (err: any) {
+    throw new Error(`Pandoc extraction failed: ${err.message}`);
+  }
+
+  // Count track changes from pandoc output
+  let insertions = 0;
+  let deletions = 0;
+
+  // Pandoc outputs track changes as:
+  // [inserted text]{.insertion author="..."}
+  // [deleted text]{.deletion author="..."}
+
+  // Convert pandoc's track change format to CriticMarkup
+  // Insertions: [text]{.insertion ...} -> {++text++}
+  text = text.replace(/\[([^\]]*)\]\{\.insertion[^}]*\}/g, (match, content) => {
+    if (content.trim()) {
+      insertions++;
+      return `{++${content}++}`;
+    }
+    return '';
+  });
+
+  // Deletions: [text]{.deletion ...} -> {--text--}
+  text = text.replace(/\[([^\]]*)\]\{\.deletion[^}]*\}/g, (match, content) => {
+    if (content.trim()) {
+      deletions++;
+      return `{--${content}--}`;
+    }
+    return '';
+  });
+
+  const hasTrackChanges = insertions > 0 || deletions > 0;
+
+  return {
+    text,
+    hasTrackChanges,
+    stats: { insertions, deletions },
+  };
+}