officeparser 7.1.0 → 7.2.0

package/dist/parsers/CsvParser.js

@@ -110,6 +110,6 @@ const parseCsv = async (buffer, config) => {
             .join(config.newlineDelimiter)
             .replace(/\n{3,}/g, '\n\n');
     };
-    return (0, astUtils_js_1.createAST)('csv', { title: 'Sheet1' }, [sheetNode], [], config, toTextSync);
+    return (0, astUtils_js_1.createAST)('csv', { title: 'Sheet1' }, [sheetNode], [], config, undefined, toTextSync);
 };
 exports.parseCsv = parseCsv;

package/dist/parsers/ExcelParser.js

@@ -51,18 +51,23 @@ const parseExcel = async (buffer, config) => {
     const mediaFileRegex = /xl\/media\/.*/;
     const corePropsFileRegex = /docProps\/core\.xml/;
     const customPropsFileRegex = /docProps\/custom\.xml/;
+    const appPropsFileRegex = /docProps\/app\.xml/;
     const relsRegex = /xl\/worksheets\/_rels\/sheet\d+\.xml\.rels/g;
     const drawingRelsRegex = /xl\/drawings\/_rels\/drawing\d+\.xml\.rels/g;
+    const commentsRegex = /xl\/comments\d+\.xml/g;
     const files = await (0, zipUtils_js_1.extractFiles)(buffer, (x) => !!x.match(sheetsRegex) ||
         !!x.match(drawingsRegex) ||
         !!x.match(chartsRegex) ||
+        (!config.ignoreComments && !!x.match(commentsRegex)) ||
         x === stringsFilePath ||
         x === 'xl/styles.xml' ||
         x === 'xl/workbook.xml' ||
         x === 'xl/_rels/workbook.xml.rels' ||
         !!x.match(corePropsFileRegex) ||
         !!x.match(customPropsFileRegex) ||
-        (!!config.extractAttachments && (!!x.match(mediaFileRegex) || !!x.match(relsRegex) || !!x.match(drawingRelsRegex))));
+        !!x.match(appPropsFileRegex) ||
+        (!!config.extractAttachments && (!!x.match(mediaFileRegex) || !!x.match(drawingRelsRegex))) ||
+        ((!!config.extractAttachments || !config.ignoreComments) && !!x.match(relsRegex)));
     const sharedStringsFile = files.find(f => f.path === stringsFilePath);
     // Updated to store structured content (rich text runs) or simple string
     const sharedStrings = [];
@@ -415,6 +420,52 @@ const parseExcel = async (buffer, config) => {
             if (config.includeRawContent) {
                 rawContents.push(file.content.toString());
             }
+            const sheetFilename = file.path.split('/').pop() || '';
+            const relsFilename = `xl/worksheets/_rels/${sheetFilename}.rels`;
+            const relsFile = files.find(f => f.path === relsFilename);
+            const drawingMap = {}; // rId -> drawingPath
+            const sheetCommentsMap = {};
+            if (relsFile) {
+                const relsXml = (0, xmlUtils_js_1.parseXmlString)(relsFile.content.toString());
+                const relationships = (0, xmlUtils_js_1.getElementsByTagName)(relsXml, "Relationship");
+                for (const rel of relationships) {
+                    const id = rel.getAttribute("Id");
+                    const target = rel.getAttribute("Target");
+                    const type = rel.getAttribute("Type");
+                    if (id && target && type) {
+                        if (config.extractAttachments && type.includes('drawing')) {
+                            drawingMap[id] = 'xl/drawings/' + target.replace('../drawings/', '');
+                        }
+                        else if (!config.ignoreComments && type.includes('comments')) {
+                            const commentsPath = 'xl/' + target.replace('../', '');
+                            const cFile = files.find(f => f.path === commentsPath);
+                            if (cFile) {
+                                const cXml = (0, xmlUtils_js_1.parseXmlString)(cFile.content.toString());
+                                const commentNodes = (0, xmlUtils_js_1.getElementsByTagName)(cXml, "comment");
+                                const authorsList = (0, xmlUtils_js_1.getElementsByTagName)(cXml, "author");
+                                const authors = authorsList.map(a => a.textContent || '');
+                                for (const cNode of commentNodes) {
+                                    const ref = cNode.getAttribute("ref");
+                                    const authorId = cNode.getAttribute("authorId");
+                                    const author = authorId !== null ? authors[parseInt(authorId)] : undefined;
+                                    const tNodes = (0, xmlUtils_js_1.getElementsByTagName)(cNode, "t");
+                                    const text = tNodes.map(t => t.textContent || '').join('');
+                                    if (ref && text) {
+                                        if (!sheetCommentsMap[ref])
+                                            sheetCommentsMap[ref] = [];
+                                        sheetCommentsMap[ref].push({
+                                            type: 'comment',
+                                            text: text,
+                                            children: [{ type: 'text', text: text, formatting: {} }],
+                                            metadata: author ? { author } : undefined
+                                        });
+                                    }
+                                }
+                            }
+                        }
+                    }
+                }
+            }
             const rows = [];
             const sheetXml = file.content.toString();
             // regex to match <row> elements, capturing:
@@ -485,7 +536,9 @@ const parseExcel = async (buffer, config) => {
                     // Parse cell coordinate
                     const coordMatch = cAttrs.match(/r="([A-Z]+)(\d+)"/);
                     let colIndex;
+                    let ref;
                     if (coordMatch) {
+                        ref = coordMatch[1] + coordMatch[2];
                         colIndex = colToNumber(coordMatch[1]);
                         // If row index is missing in cell coord (unlikely but possible), use rowIndex
                     }
@@ -525,10 +578,12 @@ const parseExcel = async (buffer, config) => {
                                 formatting: cellFormatting
                             });
                         }
+                        const commentsNodeList = (ref && sheetCommentsMap[ref]) ? sheetCommentsMap[ref] : undefined;
                         const cellNode = {
                             type: 'cell',
                             text: text,
                             children: cellNodes,
+                            comments: commentsNodeList,
                             metadata: { row: rowIndex, col: colIndex }
                         };
                         if (config.includeRawContent) {
@@ -551,23 +606,6 @@ const parseExcel = async (buffer, config) => {
             }
             // Handle Drawings in Sheet (images and charts)
             if (config.extractAttachments) {
-                // Parse Sheet Rels to map drawing rIds
-                const sheetFilename = file.path.split('/').pop() || '';
-                const relsFilename = `xl/worksheets/_rels/${sheetFilename}.rels`;
-                const relsFile = files.find(f => f.path === relsFilename);
-                const drawingMap = {}; // rId -> drawingPath
-                if (relsFile) {
-                    const relsXml = (0, xmlUtils_js_1.parseXmlString)(relsFile.content.toString());
-                    const relationships = (0, xmlUtils_js_1.getElementsByTagName)(relsXml, "Relationship");
-                    for (const rel of relationships) {
-                        const id = rel.getAttribute("Id");
-                        const target = rel.getAttribute("Target");
-                        const type = rel.getAttribute("Type");
-                        if (id && target && type && type.includes('drawing')) {
-                            drawingMap[id] = 'xl/drawings/' + target.replace('../drawings/', '');
-                        }
-                    }
-                }
                 const drawingMatches = file.content.toString().match(/<drawing r:id="(.*?)"/g);
                 if (drawingMatches) {
                     for (const match of drawingMatches) {
@@ -637,6 +675,12 @@ const parseExcel = 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;
+    }
     // Link OCR text and chart data to content nodes (like PPTX parser)
     const assignAttachmentData = (nodes) => {
         for (const node of nodes) {
@@ -681,6 +725,6 @@ const parseExcel = async (buffer, config) => {
         };
         return getText(c);
     }).filter(t => t != '').join(config.newlineDelimiter);
-    return (0, astUtils_js_1.createAST)('xlsx', metadata, content, attachments, config, toTextSync);
+    return (0, astUtils_js_1.createAST)('xlsx', metadata, content, attachments, config, undefined, toTextSync);
 };
 exports.parseExcel = parseExcel;

package/dist/parsers/HtmlParser.js

@@ -127,6 +127,15 @@ const parseHtml = async (buffer, config) => {
         if (titleNode && titleNode.children.length > 0 && titleNode.children[0].text) {
             metadata.title = titleNode.children[0].text;
         }
+        metadata.nativeProperties = {};
+        for (const child of head.children) {
+            if (child.tagName === 'meta') {
+                const name = child.attributes?.name || child.attributes?.property || child.attributes?.['http-equiv'];
+                if (name) {
+                    metadata.nativeProperties[name] = child.attributes?.content || '';
+                }
+            }
+        }
         const extractMeta = (name) => {
             for (const child of head.children) {
                 if (child.tagName === 'meta' && (child.attributes?.name === name || child.attributes?.property === name)) {
@@ -539,6 +548,6 @@ const parseHtml = async (buffer, config) => {
         return getText(n);
     }).join(config.newlineDelimiter)
         .replace(/\n{3,}/g, '\n\n'); // Normalize excessive whitespace
-    return (0, astUtils_js_1.createAST)('html', metadata, content, attachments, config, toTextSync);
+    return (0, astUtils_js_1.createAST)('html', metadata, content, attachments, config, undefined, toTextSync);
 };
 exports.parseHtml = parseHtml;

package/dist/parsers/MarkdownParser.js

@@ -21,11 +21,20 @@ const parseMarkdown = async (buffer, config) => {
             textStr = textStr.substring(endIdx + 5);
             const lines = frontMatter.split('\n');
             const customProps = {};
+            const nativeProps = {};
             for (const line of lines) {
                 const match = line.match(/^([^:]+):\s*(.*)$/);
                 if (match) {
                     const key = match[1].trim();
                     let val = match[2].trim().replace(/^"(.*)"$/, '$1');
+                    let parsedVal = val;
+                    if (val === 'true')
+                        parsedVal = true;
+                    else if (val === 'false')
+                        parsedVal = false;
+                    else if (!isNaN(Number(val)) && val !== '')
+                        parsedVal = Number(val);
+                    nativeProps[key] = parsedVal;
                     if (key === 'title')
                         metadata.title = val;
                     else if (key === 'author')
@@ -37,20 +46,14 @@ const parseMarkdown = async (buffer, config) => {
                     else if (key === 'description')
                         metadata.description = val;
                     else {
-                        // Try to infer type for custom props
-                        if (val === 'true')
-                            customProps[key] = true;
-                        else if (val === 'false')
-                            customProps[key] = false;
-                        else if (!isNaN(Number(val)) && val !== '')
-                            customProps[key] = Number(val);
-                        else
-                            customProps[key] = val;
+                        customProps[key] = parsedVal;
                     }
                 }
             }
             if (Object.keys(customProps).length > 0)
                 metadata.customProperties = customProps;
+            if (Object.keys(nativeProps).length > 0)
+                metadata.nativeProperties = nativeProps;
         }
     }
     // Extract code blocks first to protect their contents
@@ -360,6 +363,6 @@ const parseMarkdown = async (buffer, config) => {
         return getText(n);
     }).join(config.newlineDelimiter)
         .replace(/\n{3,}/g, '\n\n'); // Normalize excessive whitespace
-    return (0, astUtils_js_1.createAST)('md', metadata, content, attachments, config, toTextSync);
+    return (0, astUtils_js_1.createAST)('md', metadata, content, attachments, config, undefined, toTextSync);
 };
 exports.parseMarkdown = parseMarkdown;

package/dist/parsers/OpenOfficeParser.js

@@ -157,9 +157,9 @@ const parseOpenOffice = async (buffer, config) => {
                     if (textPosition.startsWith("super"))
                         formatting.superscript = true;
                 }
-                if (Object.keys(formatting).length > 0)
-                    styleMap[name] = formatting;
             }
+            if (Object.keys(formatting).length > 0)
+                styleMap[name] = formatting;
         }
     };
     if (stylesDom) {
@@ -310,11 +310,17 @@ const parseOpenOffice = async (buffer, config) => {
                                 noteId: noteId
                             }
                         };
-                        if (config.putNotesAtLast) {
-                            notes.push(noteNode);
+                        if (children.length > 0 && children[children.length - 1].type === 'text') {
+                            const precedingNode = children[children.length - 1];
+                            if (!precedingNode.notes) {
+                                precedingNode.notes = [];
+                            }
+                            precedingNode.notes.push(noteNode);
                         }
                         else {
-                            children.push(noteNode);
+                            const emptyTextNode = { type: 'text', text: '' };
+                            emptyTextNode.notes = [noteNode];
+                            children.push(emptyTextNode);
                         }
                     }
                 }
@@ -492,22 +498,33 @@ const parseOpenOffice = async (buffer, config) => {
                             const element = child;
                             if (element.tagName === "text:p" || element.tagName === "text:h") {
                                 const pContent = parseParagraphContent(element, paraStyleMap, styleMap, config, sourceXml);
-                                const pNode = {
-                                    type: element.tagName === "text:h" ? 'heading' : 'paragraph',
-                                    text: pContent.text,
-                                    children: pContent.children,
-                                    metadata: {
-                                        ...(pContent.alignment ? { alignment: pContent.alignment } : {}),
-                                        ...(pContent.style ? { style: pContent.style } : {})
-                                    }
-                                };
+                                let pNode;
+                                if (element.tagName === "text:h") {
+                                    pNode = {
+                                        type: 'heading',
+                                        text: pContent.text,
+                                        children: pContent.children,
+                                        metadata: {
+                                            level: parseInt(element.getAttribute("text:outline-level") || "1"),
+                                            ...(pContent.alignment ? { alignment: pContent.alignment } : {}),
+                                            ...(pContent.style ? { style: pContent.style } : {})
+                                        }
+                                    };
+                                }
+                                else {
+                                    pNode = {
+                                        type: 'paragraph',
+                                        text: pContent.text,
+                                        children: pContent.children,
+                                        metadata: {
+                                            ...(pContent.alignment ? { alignment: pContent.alignment } : {}),
+                                            ...(pContent.style ? { style: pContent.style } : {})
+                                        }
+                                    };
+                                }
                                 // Clean up metadata if empty
-                                if (Object.keys(pNode.metadata || {}).length === 0)
+                                if (pNode.type === 'paragraph' && Object.keys(pNode.metadata || {}).length === 0) {
                                     delete pNode.metadata;
-                                if (element.tagName === "text:h") {
-                                    if (!pNode.metadata)
-                                        pNode.metadata = {};
-                                    pNode.metadata.level = parseInt(element.getAttribute("text:outline-level") || "1");
                                 }
                                 if (config.includeRawContent) {
                                     pNode.rawContent = (0, xmlUtils_js_1.getRawContent)(element, sourceXml, config);
@@ -539,11 +556,18 @@ const parseOpenOffice = async (buffer, config) => {
                 }
                 // Add cell(s) for repeated columns
                 for (let k = 0; k < colsRepeated; k++) {
+                    // Apply cell background color if defined in styleMap
+                    const cellStyleName = cell.getAttribute("table:style-name");
+                    const cellBgColor = cellStyleName && styleMap[cellStyleName]?.backgroundColor;
                     const cellNode = {
                         type: 'cell',
                         text: cellText,
                         children: cellChildren.length > 0 ? (k === 0 ? cellChildren : JSON.parse(JSON.stringify(cellChildren))) : [],
-                        metadata: { row: rowIndex, col: colIndex }
+                        metadata: {
+                            row: rowIndex,
+                            col: colIndex,
+                            ...(cellBgColor ? { backgroundColor: cellBgColor } : {})
+                        }
                     };
                     const cellMetadata = cellNode.metadata;
                     if (colSpan > 1)
@@ -621,9 +645,15 @@ const parseOpenOffice = async (buffer, config) => {
                 if (Object.keys(styleInfo).length > 0) {
                     paragraphStyleMap[name] = styleInfo;
                 }
+                const cellProps = (0, xmlUtils_js_1.getFirstElementByTagName)(style, "style:table-cell-properties");
+                const formatting = {};
+                if (cellProps) {
+                    const bgColor = cellProps.getAttribute("fo:background-color");
+                    if (bgColor && bgColor !== 'transparent')
+                        formatting.backgroundColor = bgColor;
+                }
                 const textProps = (0, xmlUtils_js_1.getFirstElementByTagName)(style, "style:text-properties");
                 if (textProps) {
-                    const formatting = {};
                     if (textProps.getAttribute("fo:font-weight") === "bold" || textProps.getAttribute("style:font-weight-asian") === "bold")
                         formatting.bold = true;
                     if (textProps.getAttribute("fo:font-style") === "italic" || textProps.getAttribute("style:font-style-asian") === "italic")
@@ -654,9 +684,9 @@ const parseOpenOffice = async (buffer, config) => {
                         if (textPosition.startsWith("super"))
                             formatting.superscript = true;
                     }
-                    if (Object.keys(formatting).length > 0)
-                        styleMap[name] = formatting;
                 }
+                if (Object.keys(formatting).length > 0)
+                    styleMap[name] = formatting;
             }
         }
         // Start traversal
@@ -1271,12 +1301,9 @@ const parseOpenOffice = async (buffer, config) => {
                     }
                     content.push(slideNode);
                     if (noteNode && noteNode.children && noteNode.children.length > 0) {
-                        if (config.putNotesAtLast) {
-                            odpNotes.push(noteNode);
-                        }
-                        else {
-                            content.push(noteNode);
-                        }
+                        if (!slideNode.notes)
+                            slideNode.notes = [];
+                        slideNode.notes.push(noteNode);
                     }
                 }
                 if (odpNotes.length > 0) {
@@ -1499,10 +1526,6 @@ const parseOpenOffice = async (buffer, config) => {
             };
         }
     }
-    // Append notes to content if configured
-    if (config.putNotesAtLast && notes.length > 0) {
-        content.push(...notes);
-    }
     const toTextSync = () => content.map(c => {
         const getText = (node) => {
             let t = '';
@@ -1524,6 +1547,6 @@ const parseOpenOffice = async (buffer, config) => {
     return (0, astUtils_js_1.createAST)(fileType, {
         ...metadata,
         styleMap: combinedStyleMap
-    }, content, attachments, config, toTextSync);
+    }, content, attachments, config, undefined, toTextSync);
 };
 exports.parseOpenOffice = parseOpenOffice;

package/dist/parsers/PdfParser.js

@@ -351,6 +351,17 @@ const parsePdf = async (buffer, config) => {
         'IsCollectionPresent', 'IsSignaturesPresent', 'PDFFormatVersion'
     ]);
     if (info) {
+        metadata.nativeProperties = {};
+        for (const [key, val] of Object.entries(info)) {
+            if (key === 'Custom' && typeof val === 'object' && !Array.isArray(val) && !(val instanceof Date) && val !== null) {
+                for (const [customKey, customVal] of Object.entries(val)) {
+                    metadata.nativeProperties[customKey] = customVal;
+                }
+            }
+            else {
+                metadata.nativeProperties[key] = val;
+            }
+        }
         const customProperties = {};
         for (const key of Object.keys(info)) {
             if (standardPdfInfoKeys.has(key))
@@ -378,6 +389,17 @@ const parsePdf = async (buffer, config) => {
             metadata.customProperties = customProperties;
         }
     }
+    if (meta.metadata) {
+        if (!metadata.nativeProperties)
+            metadata.nativeProperties = {};
+        const xmp = meta.metadata;
+        if (typeof xmp.getAll === 'function') {
+            metadata.nativeProperties['XMP'] = xmp.getAll();
+        }
+        else {
+            metadata.nativeProperties['XMP'] = xmp;
+        }
+    }
     // --- Embedded File Attachment Extraction ---
     /**
      * PDF can contain embedded files (not images in content, but attached files).
@@ -793,7 +815,7 @@ const parsePdf = async (buffer, config) => {
         });
     }
     const toTextSync = () => content.map(c => c.text).join(config.newlineDelimiter);
-    return (0, astUtils_js_1.createAST)('pdf', metadata, content, attachments, config, toTextSync);
+    return (0, astUtils_js_1.createAST)('pdf', metadata, content, attachments, config, undefined, toTextSync);
 };
 exports.parsePdf = parsePdf;
 /**