officeparser 7.0.3 → 7.2.0

package/dist/parsers/WordParser.js

@@ -86,13 +86,21 @@ const zipUtils_js_1 = require("../utils/zipUtils.js");
  * @returns A promise resolving to the parsed AST
  */
 const parseWord = async (buffer, config) => {
+    // Honour cancellation requests immediately — before opening the ZIP archive, loading XML
+    // files, or kicking off any OCR work.  DOCX files can be large and the inflate + XML-parse
+    // steps are synchronous-heavy, so failing fast here avoids wasted CPU time.
+    (0, errorUtils_js_1.checkAbortSignal)(config.abortSignal);
     const documentFileRegex = /word\/document[\d+]?.xml/;
     const footnotesFileRegex = /word\/footnotes[\d+]?.xml/;
     const endnotesFileRegex = /word\/endnotes[\d+]?.xml/;
+    const commentsFileRegex = /word\/comments[\d+]?.xml/;
+    const headerFileRegex = /word\/header[\d+]?.xml/;
+    const footerFileRegex = /word\/footer[\d+]?.xml/;
     const numberingFileRegex = /word\/numbering[\d+]?.xml/;
     const mediaFileRegex = /(word\/)?media\/.*/;
     const corePropsFileRegex = /docProps\/core[\d+]?.xml/;
     const customPropsFileRegex = /docProps\/custom\.xml/;
+    const appPropsFileRegex = /docProps\/app[\d+]?.xml/;
     const relsFileRegex = /word\/_rels\/document[\d+]?.xml\.rels/;
     const stylesFileRegex = /word\/styles[\d+]?.xml/;
     // Helper to extract formatting from run properties XML string
@@ -232,6 +240,7 @@ const parseWord = async (buffer, config) => {
         !!x.match(numberingFileRegex) ||
         !!x.match(corePropsFileRegex) ||
         !!x.match(customPropsFileRegex) ||
+        !!x.match(appPropsFileRegex) ||
         !!x.match(relsFileRegex) ||
         !!x.match(stylesFileRegex) ||
         (!!config.extractAttachments && !!x.match(mediaFileRegex)));
@@ -244,9 +253,20 @@ const parseWord = async (buffer, config) => {
         if (Object.keys(customProperties).length > 0)
             metadata.customProperties = customProperties;
     }
+    const appPropsFile = files.find(f => f.path.match(appPropsFileRegex));
+    if (appPropsFile) {
+        const appProperties = (0, xmlUtils_js_1.parseOOXMLAppProperties)(appPropsFile.content.toString());
+        if (Object.keys(appProperties).length > 0) {
+            metadata.nativeProperties = appProperties;
+            if (appProperties['Pages'] && typeof appProperties['Pages'] === 'number') {
+                metadata.pages = appProperties['Pages'];
+            }
+        }
+    }
     const footnoteMap = new Map();
     const endnoteMap = new Map();
-    const collectedNotes = [];
+    const commentMap = new Map();
+    const commentMetadataMap = new Map();
     const attachments = [];
     const mediaFiles = files.filter(f => f.path.match(mediaFileRegex));
     // Extract relationships
@@ -449,6 +469,8 @@ const parseWord = async (buffer, config) => {
         // Extract text and children
         let text = '';
         const children = [];
+        const notes = [];
+        const comments = [];
         // Traverse children of paragraph (runs, hyperlinks, etc.)
         const processChildNode = (node) => {
             if ((0, xmlUtils_js_1.isElement)(node) && node.nodeName === 'w:r') {
@@ -616,11 +638,14 @@ const parseWord = async (buffer, config) => {
                                 children: noteNodes,
                                 metadata: { noteType: 'footnote', noteId: id }
                             };
-                            if (config.putNotesAtLast) {
-                                collectedNotes.push(noteNode);
+                            if (children.length > 0) {
+                                const target = children[children.length - 1];
+                                if (!target.notes)
+                                    target.notes = [];
+                                target.notes.push(noteNode);
                             }
                             else {
-                                children.push(noteNode);
+                                notes.push(noteNode);
                             }
                         }
                     }
@@ -635,11 +660,40 @@ const parseWord = async (buffer, config) => {
                                 children: noteNodes,
                                 metadata: { noteType: 'endnote', noteId: id }
                             };
-                            if (config.putNotesAtLast) {
-                                collectedNotes.push(noteNode);
+                            if (children.length > 0) {
+                                const target = children[children.length - 1];
+                                if (!target.notes)
+                                    target.notes = [];
+                                target.notes.push(noteNode);
+                            }
+                            else {
+                                notes.push(noteNode);
+                            }
+                        }
+                    }
+                }
+                // Comments inside runs
+                if (!config.ignoreComments) {
+                    const commentRef = (0, xmlUtils_js_1.getFirstElementByTagName)(runNode, "w:commentReference");
+                    if (commentRef) {
+                        const id = commentRef.getAttribute("w:id");
+                        if (id && commentMap.has(id)) {
+                            const commentNodes = commentMap.get(id);
+                            const commentInfo = commentMetadataMap.get(id);
+                            const commentNode = {
+                                type: 'comment',
+                                text: commentNodes.map((n) => n.text).join(' '),
+                                children: commentNodes,
+                                metadata: commentInfo || { commentId: id }
+                            };
+                            if (children.length > 0) {
+                                const target = children[children.length - 1];
+                                if (!target.comments)
+                                    target.comments = [];
+                                target.comments.push(commentNode);
                             }
                             else {
-                                children.push(noteNode);
+                                comments.push(commentNode);
                             }
                         }
                     }
@@ -744,6 +798,8 @@ const parseWord = async (buffer, config) => {
                 type: 'list',
                 text: text,
                 children: children,
+                ...(notes.length > 0 ? { notes } : {}),
+                ...(comments.length > 0 ? { comments } : {}),
                 metadata: {
                     listType,
                     indentation: ilvl,
@@ -765,6 +821,8 @@ const parseWord = async (buffer, config) => {
                 type: 'heading',
                 text: text,
                 children: children,
+                ...(notes.length > 0 ? { notes } : {}),
+                ...(comments.length > 0 ? { comments } : {}),
                 metadata: { level, alignment, paragraphIndentation: paraIndentation, style: pStyleVal ?? undefined, ...commonMetadata }
             };
             if (config.includeRawContent)
@@ -776,6 +834,8 @@ const parseWord = async (buffer, config) => {
                 type: 'paragraph',
                 text: text,
                 children: children,
+                ...(notes.length > 0 ? { notes } : {}),
+                ...(comments.length > 0 ? { comments } : {}),
                 metadata: { alignment, paragraphIndentation: paraIndentation, style: pStyleVal ?? undefined, ...commonMetadata }
             };
             if (config.includeRawContent)
@@ -842,6 +902,15 @@ const parseWord = async (buffer, config) => {
                 };
                 if (colSpan > 1)
                     cellNode.metadata.colSpan = colSpan;
+                if (tcPr) {
+                    const shd = (0, xmlUtils_js_1.getFirstElementByTagName)(tcPr, "w:shd");
+                    if (shd) {
+                        const fill = shd.getAttribute("w:fill");
+                        if (fill && fill !== "auto") {
+                            cellNode.metadata.backgroundColor = "#" + fill;
+                        }
+                    }
+                }
                 if (isVMerge) {
                     if (vMergeRestart) {
                         vMergeMap.set(visualCol, { node: cellNode, span: 1 });
@@ -911,6 +980,77 @@ const parseWord = async (buffer, config) => {
             }
         }
     }
+    // Pre-process comments
+    if (!config.ignoreComments) {
+        const commentsFile = files.find(f => f.path.match(commentsFileRegex));
+        if (commentsFile) {
+            const commentsDoc = (0, xmlUtils_js_1.parseXmlString)(commentsFile.content.toString());
+            const commentsXml = commentsFile.content.toString();
+            const commentNodes = (0, xmlUtils_js_1.getElementsByTagName)(commentsDoc, "w:comment");
+            for (const node of commentNodes) {
+                const id = node.getAttribute("w:id");
+                if (!id)
+                    continue;
+                const author = node.getAttribute("w:author") || undefined;
+                const date = node.getAttribute("w:date") || undefined;
+                const initials = node.getAttribute("w:initials") || undefined;
+                commentMetadataMap.set(id, { commentId: id, author, date, initials });
+                const pNodes = (0, xmlUtils_js_1.getElementsByTagName)(node, "w:p");
+                commentMap.set(id, pNodes.map(p => parseParagraph(p, commentsXml)));
+            }
+        }
+    }
+    // Pre-process headers and footers
+    const headers = [];
+    const footers = [];
+    if (!config.ignoreHeadersAndFooters) {
+        const headerFiles = files.filter(f => f.path.match(headerFileRegex));
+        for (const hFile of headerFiles) {
+            const hDoc = (0, xmlUtils_js_1.parseXmlString)(hFile.content.toString());
+            const hXml = hFile.content.toString();
+            const hNodes = Array.from(hDoc.documentElement.childNodes).filter(xmlUtils_js_1.isElement);
+            for (const child of hNodes) {
+                if (child.nodeName === 'w:p')
+                    headers.push(parseParagraph(child, hXml));
+                else if (child.nodeName === 'w:tbl')
+                    headers.push(parseTable(child, hXml));
+                else if (child.nodeName === 'w:sdt') {
+                    const contentNode = (0, xmlUtils_js_1.getFirstElementByTagName)(child, "w:sdtContent");
+                    if (contentNode) {
+                        for (const sdtChild of Array.from(contentNode.childNodes).filter(xmlUtils_js_1.isElement)) {
+                            if (sdtChild.nodeName === 'w:p')
+                                headers.push(parseParagraph(sdtChild, hXml));
+                            else if (sdtChild.nodeName === 'w:tbl')
+                                headers.push(parseTable(sdtChild, hXml));
+                        }
+                    }
+                }
+            }
+        }
+        const footerFiles = files.filter(f => f.path.match(footerFileRegex));
+        for (const fFile of footerFiles) {
+            const fDoc = (0, xmlUtils_js_1.parseXmlString)(fFile.content.toString());
+            const fXml = fFile.content.toString();
+            const fNodes = Array.from(fDoc.documentElement.childNodes).filter(xmlUtils_js_1.isElement);
+            for (const child of fNodes) {
+                if (child.nodeName === 'w:p')
+                    footers.push(parseParagraph(child, fXml));
+                else if (child.nodeName === 'w:tbl')
+                    footers.push(parseTable(child, fXml));
+                else if (child.nodeName === 'w:sdt') {
+                    const contentNode = (0, xmlUtils_js_1.getFirstElementByTagName)(child, "w:sdtContent");
+                    if (contentNode) {
+                        for (const sdtChild of Array.from(contentNode.childNodes).filter(xmlUtils_js_1.isElement)) {
+                            if (sdtChild.nodeName === 'w:p')
+                                footers.push(parseParagraph(sdtChild, fXml));
+                            else if (sdtChild.nodeName === 'w:tbl')
+                                footers.push(parseTable(sdtChild, fXml));
+                        }
+                    }
+                }
+            }
+        }
+    }
     for (const file of files) {
         if (file.path.match(mediaFileRegex))
             continue;
@@ -924,6 +1064,12 @@ const parseWord = async (buffer, config) => {
             continue;
         if (file.path.match(endnotesFileRegex))
             continue;
+        if (file.path.match(commentsFileRegex))
+            continue;
+        if (file.path.match(headerFileRegex))
+            continue;
+        if (file.path.match(footerFileRegex))
+            continue;
         const documentContent = file.content.toString();
         if (config.includeRawContent) {
             rawContents.push(documentContent);
@@ -989,9 +1135,6 @@ const parseWord = async (buffer, config) => {
             assignOcr(content);
         }
     }
-    if (config.putNotesAtLast && collectedNotes.length > 0) {
-        content.push(...collectedNotes);
-    }
     const toTextSync = () => content.map(c => {
         // Recursive text extraction
         const getText = (node) => {
@@ -1008,6 +1151,10 @@ const parseWord = async (buffer, config) => {
         };
         return getText(c);
     }).filter(t => t != '').join(config.newlineDelimiter);
-    return (0, astUtils_js_1.createAST)('docx', { ...metadata, formatting: docDefaults, styleMap: styleMap }, content, attachments, config, toTextSync);
+    const auxiliaryContent = (headers.length > 0 || footers.length > 0) ? {
+        ...(headers.length > 0 ? { headers } : {}),
+        ...(footers.length > 0 ? { footers } : {})
+    } : undefined;
+    return (0, astUtils_js_1.createAST)('docx', { ...metadata, formatting: docDefaults, styleMap: styleMap }, content, attachments, config, auxiliaryContent, toTextSync);
 };
 exports.parseWord = parseWord;