officeparser 6.0.6 → 6.1.0

package/dist/parsers/PdfParser.d.ts

@@ -56,7 +56,7 @@
  * @see https://mozilla.github.io/pdf.js/ PDF.js documentation
  * @see https://www.adobe.com/devnet/pdf/pdf_reference.html PDF Reference
  */
-import { OfficeParserAST, OfficeParserConfig } from '../types';
+import { OfficeParserAST, OfficeParserConfig } from '../types.js';
 /**
  * Parses a PDF file and extracts content.
  *

package/dist/parsers/PdfParser.js

@@ -59,53 +59,15 @@
  */
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.parsePdf = void 0;
-const errorUtils_1 = require("../utils/errorUtils");
-const imageUtils_1 = require("../utils/imageUtils");
-const ocrUtils_1 = require("../utils/ocrUtils");
-const moduleLoader_1 = require("../utils/moduleLoader");
-/**
- * Parses PDF creation/modification date strings.
- * PDF dates are in format: D:YYYYMMDDHHmmSSOHH'mm'
- * Where O is timezone offset direction (+/-), HH is hours, mm is minutes.
- *
- * @param dateString - The PDF date string
- * @returns Parsed Date object or undefined if parsing fails
- */
-function parsePdfDate(dateString) {
-    if (!dateString)
-        return undefined;
-    try {
-        // Remove "D:" prefix if present
-        let str = dateString.startsWith('D:') ? dateString.slice(2) : dateString;
-        // Extract components: YYYYMMDDHHmmSS
-        const year = parseInt(str.slice(0, 4), 10);
-        const month = parseInt(str.slice(4, 6), 10) - 1; // 0-indexed
-        const day = parseInt(str.slice(6, 8), 10) || 1;
-        const hour = parseInt(str.slice(8, 10), 10) || 0;
-        const minute = parseInt(str.slice(10, 12), 10) || 0;
-        const second = parseInt(str.slice(12, 14), 10) || 0;
-        // Handle timezone if present
-        const tzMatch = str.slice(14).match(/([+-Z])(\d{2})'?(\d{2})?'?/);
-        if (tzMatch) {
-            const tzSign = tzMatch[1] === '-' ? -1 : 1;
-            const tzHours = parseInt(tzMatch[2], 10) || 0;
-            const tzMinutes = parseInt(tzMatch[3], 10) || 0;
-            const offset = tzSign * (tzHours * 60 + tzMinutes);
-            // Create date in UTC and adjust for timezone
-            const utc = Date.UTC(year, month, day, hour, minute, second);
-            return new Date(utc - offset * 60000);
-        }
-        return new Date(year, month, day, hour, minute, second);
-    }
-    catch {
-        // Fallback: try native Date parsing
-        try {
-            return new Date(dateString);
-        }
-        catch {
-            return undefined;
-        }
-    }
+const errorUtils_js_1 = require("../utils/errorUtils.js");
+const imageUtils_js_1 = require("../utils/imageUtils.js");
+const ocrUtils_js_1 = require("../utils/ocrUtils.js");
+const moduleLoader_js_1 = require("../utils/moduleLoader.js");
+const dateUtils_js_1 = require("../utils/dateUtils.js");
+const envUtils_js_1 = require("../utils/envUtils.js");
+/** Type guard for TextItem in PDF.js 5.x */
+function isTextItem(item) {
+    return item && typeof item.str === 'string' && Array.isArray(item.transform) && item.transform.length >= 6;
 }
 /**
  * Calculates statistics about font sizes in the document.
@@ -155,27 +117,21 @@ function detectHeadingLevel(fontSize, fontStats) {
         return 5; // 1.2x-1.35x = H5
     return 0; // Below threshold
 }
-/**
- * Checks if a text position falls within a link annotation rectangle.
- */
-/**
- * Checks if a text item falls within a link annotation rectangle.
- * Uses the center point of the text item to avoid false positives for text
- * that starts immediately after a link (which would share the same x coordinate boundary).
- */
 function findLinkForText(item, annotations) {
-    const centerX = item.x + (item.width / 2);
-    const centerY = item.y + (item.height / 2);
+    const itemMinX = item.x;
+    const itemMaxX = item.x + item.width;
+    const itemMinY = item.y;
+    const itemMaxY = item.y + item.height;
     for (const annot of annotations) {
-        // PDF annotation rects are [x1, y1, x2, y2] in page coordinates
         const [x1, y1, x2, y2] = annot.rect;
-        const minX = Math.min(x1, x2);
-        const maxX = Math.max(x1, x2);
-        const minY = Math.min(y1, y2);
-        const maxY = Math.max(y1, y2);
-        // Check if center point is within the rect (strict, no tolerance)
-        if (centerX >= minX && centerX <= maxX &&
-            centerY >= minY && centerY <= maxY) {
+        const annotMinX = Math.min(x1, x2);
+        const annotMaxX = Math.max(x1, x2);
+        const annotMinY = Math.min(y1, y2);
+        const annotMaxY = Math.max(y1, y2);
+        // Check for any intersection between boxes
+        const intersects = (itemMinX < annotMaxX && itemMaxX > annotMinX) &&
+            (itemMinY < annotMaxY && itemMaxY > annotMinY);
+        if (intersects) {
             return annot;
         }
     }
@@ -307,20 +263,20 @@ function convertToRgbaBuffer(data, width, height, kind) {
  * @returns Promise resolving to the parsed AST
  */
 const parsePdf = async (buffer, config) => {
-    const pdfjs = await (0, moduleLoader_1.loadPdfJs)();
+    const pdfjs = await (0, moduleLoader_js_1.loadPdfJs)();
     // Configure worker
     if (config.pdfWorkerSrc) {
         pdfjs.GlobalWorkerOptions.workerSrc = config.pdfWorkerSrc;
     }
     else {
         // Fallbacks when no workerSrc is provided
-        // @ts-ignore
-        if (typeof window !== 'undefined') {
+        if (envUtils_js_1.isBrowser) {
             // Browser: Default to CDN
             pdfjs.GlobalWorkerOptions.workerSrc = `https://unpkg.com/pdfjs-dist@${pdfjs.version}/build/pdf.worker.min.mjs`;
         }
         else {
             // Node.js: Try to auto-resolve local worker path to avoid remote fetch errors
+            (0, envUtils_js_1.assertNode)('pdf-worker-auto-resolution');
             try {
                 // We use require.resolve to find the exact path of the installed package.
                 // @ts-ignore - 'require' is available in Node.js/CommonJS environment
@@ -344,8 +300,9 @@ const parsePdf = async (buffer, config) => {
         pdfDocument = await loadingTask.promise;
     }
     catch (e) {
-        if (e.message?.includes('workerSrc') || e.message?.includes('No "GlobalWorkerOptions.workerSrc" specified')) {
-            throw (0, errorUtils_1.getOfficeError)(errorUtils_1.OfficeErrorType.PDF_WORKER_MISSING, config);
+        const message = e instanceof Error ? e.message : String(e);
+        if (message.includes('workerSrc') || message.includes('No "GlobalWorkerOptions.workerSrc" specified')) {
+            throw (0, errorUtils_js_1.getOfficeError)(errorUtils_js_1.OfficeErrorType.PDF_WORKER_MISSING, config);
         }
         throw e;
     }
@@ -364,15 +321,50 @@ const parsePdf = async (buffer, config) => {
     const info = meta.info;
     const metadata = {
         pages: numPages,
-        title: info?.Title || undefined,
-        author: info?.Author || undefined,
-        subject: info?.Subject || undefined,
-        description: info?.Keywords || undefined, // Map Keywords to description as closest match
-        created: parsePdfDate(info?.CreationDate),
-        modified: parsePdfDate(info?.ModDate),
+        title: info?.Title,
+        author: info?.Author,
+        subject: info?.Subject,
+        description: info?.Keywords, // Map Keywords to description as closest match
+        created: (0, dateUtils_js_1.parseOfficeDate)(info?.CreationDate),
+        modified: (0, dateUtils_js_1.parseOfficeDate)(info?.ModDate),
         // Note: lastModifiedBy is not available in PDF format - there's no concept of "last modifier"
         // The Author field only tracks original author.
     };
+    // Extract non-standard entries from the PDF Info dictionary as custom properties.
+    // The standard keys are defined by the PDF spec; anything else is user/tool-defined.
+    const standardPdfInfoKeys = new Set([
+        'Title', 'Author', 'Subject', 'Keywords', 'Creator', 'Producer',
+        'CreationDate', 'ModDate', 'Trapped', 'IsAcroFormPresent', 'IsXFAPresent',
+        'IsCollectionPresent', 'IsSignaturesPresent', 'PDFFormatVersion'
+    ]);
+    if (info) {
+        const customProperties = {};
+        for (const key of Object.keys(info)) {
+            if (standardPdfInfoKeys.has(key))
+                continue;
+            const val = info[key];
+            if (val === null || val === undefined)
+                continue;
+            // pdf.js groups document-level custom metadata under a 'Custom' object.
+            // Flatten its entries directly into customProperties.
+            if (key === 'Custom' && typeof val === 'object' && !Array.isArray(val) && !(val instanceof Date)) {
+                for (const [customKey, customVal] of Object.entries(val)) {
+                    if (customVal === null || customVal === undefined)
+                        continue;
+                    if (typeof customVal === 'string' || typeof customVal === 'number' || typeof customVal === 'boolean' || customVal instanceof Date) {
+                        customProperties[customKey] = customVal;
+                    }
+                }
+                continue;
+            }
+            if (typeof val === 'string' || typeof val === 'number' || typeof val === 'boolean' || val instanceof Date) {
+                customProperties[key] = val;
+            }
+        }
+        if (Object.keys(customProperties).length > 0) {
+            metadata.customProperties = customProperties;
+        }
+    }
     // --- Embedded File Attachment Extraction ---
     /**
      * PDF can contain embedded files (not images in content, but attached files).
@@ -384,7 +376,7 @@ const parsePdf = async (buffer, config) => {
             for (const name in embeddedFiles) {
                 const file = embeddedFiles[name];
                 const fileBuffer = Buffer.from(file.content);
-                const attachment = (0, imageUtils_1.createAttachment)(file.filename, fileBuffer);
+                const attachment = (0, imageUtils_js_1.createAttachment)(file.filename, fileBuffer);
                 attachments.push(attachment);
             }
         }
@@ -412,23 +404,30 @@ const parsePdf = async (buffer, config) => {
         }
         const commonObjs = page.commonObjs;
         for (const item of textContent.items) {
-            const transform = item.transform;
+            // PDF.js 5.x: textContent.items can contain TextMarkedContent which lack
+            // 'str' and 'transform'. Skip these to avoid crashes and page skipping.
+            if (!isTextItem(item)) {
+                continue;
+            }
+            // At this point we know the item is a TextItem
+            const textItem = item;
+            const transform = textItem.transform;
             const x = transform[4];
             const y = transform[5];
-            const width = item.width || 0;
-            const height = item.height || Math.abs(transform[3]) || 12;
+            const width = textItem.width || 0;
+            const height = textItem.height || Math.abs(transform[3]) || 12;
             // Extract formatting from font
             const formatting = {};
             let fontName;
-            if (item.fontName && commonObjs) {
+            if (textItem.fontName && commonObjs) {
                 try {
-                    if (commonObjs.has(item.fontName)) {
+                    if (commonObjs.has(textItem.fontName)) {
                         // Use callback-based get to ensure safe resolution
                         const fontData = await new Promise((resolve) => {
-                            // @ts-ignore
-                            commonObjs.get(item.fontName, (data) => resolve(data));
+                            // @ts-ignore - commonObjs.get is callback-based in legacy builds
+                            commonObjs.get(textItem.fontName, (data) => resolve(data));
                         });
-                        if (fontData?.name) {
+                        if (fontData?.name && typeof fontData.name === 'string') {
                             // Remove PDF subset prefix (6 uppercase letters + '+')
                             fontName = fontData.name.replace(/^[A-Z]{6}\+/, '');
                             formatting.font = fontName;
@@ -454,7 +453,7 @@ const parsePdf = async (buffer, config) => {
                 y,
                 width,
                 height,
-                text: item.str,
+                text: textItem.str,
                 fontName,
                 formatting
             });
@@ -503,11 +502,10 @@ const parsePdf = async (buffer, config) => {
                             }
                             if (hasObj) {
                                 // Use callback-based get to ensure safe resolution
-                                const rawObj = await new Promise((resolve) => {
-                                    // @ts-ignore
+                                const imgObj = await new Promise((resolve) => {
+                                    // @ts-ignore - targetObjs.get is callback-based
                                     targetObjs.get(imgName, (data) => resolve(data));
                                 });
-                                const imgObj = rawObj;
                                 // Browser-specific: Handle ImageBitmap if data is missing
                                 if (typeof window !== 'undefined' && !imgObj.data && imgObj.bitmap) {
                                     try {
@@ -580,13 +578,16 @@ const parsePdf = async (buffer, config) => {
         const pageContent = [];
         // Extract link annotations for this page
         const annotations = [];
+        const matchedAnnotations = new Set();
         try {
             const annots = await page.getAnnotations();
             for (const annot of annots) {
                 if (annot.subtype === 'Link' && annot.rect) {
+                    // PDF.js 5.x compatibility: url might be in 'url', 'unsafeUrl', or 'data.url'
+                    const url = annot.url || annot.unsafeUrl || annot.data?.url;
                     annotations.push({
                         rect: annot.rect,
-                        url: annot.url,
+                        url: url,
                         dest: annot.dest
                     });
                 }
@@ -598,7 +599,8 @@ const parsePdf = async (buffer, config) => {
         }
         // Sort items: Y descending (top to bottom), then X ascending (left to right)
         pageItems.sort((a, b) => {
-            if (Math.abs(b.y - a.y) > 5)
+            // Relax tolerance slightly (5 -> 7) for PDF.js 5.x coordinate precision
+            if (Math.abs(b.y - a.y) > 7)
                 return b.y - a.y;
             return a.x - b.x;
         });
@@ -672,6 +674,8 @@ const parsePdf = async (buffer, config) => {
                 currentNode.text += text;
                 // Check for link
                 const link = findLinkForText(item, annotations);
+                if (link)
+                    matchedAnnotations.add(link);
                 let textMetadata;
                 if (link) {
                     if (link.url) {
@@ -733,14 +737,14 @@ const parsePdf = async (buffer, config) => {
                         const imageBuffer = convertToRgbaBuffer(item.data, item.width, item.height, item.kind);
                         // Encode as BMP
                         const bmpBuffer = encodeBmp(item.width, item.height, new Uint8Array(imageBuffer));
-                        const attachment = (0, imageUtils_1.createAttachment)(attachmentName, bmpBuffer);
+                        const attachment = (0, imageUtils_js_1.createAttachment)(attachmentName, bmpBuffer);
                         attachment.mimeType = 'image/bmp';
                         // Perform OCR if enabled
                         if (config.ocr) {
                             try {
                                 // Skip OCR for very small images/artifacts (e.g. < 10px) to avoid Tesseract warnings
                                 if (item.width >= 10 && item.height >= 10) {
-                                    attachment.ocrText = (await (0, ocrUtils_1.performOcr)(bmpBuffer, config.ocrLanguage)).trim();
+                                    attachment.ocrText = (await (0, ocrUtils_js_1.performOcr)(bmpBuffer, { language: config.ocrLanguage, ...config.ocrConfig })).trim();
                                 }
                             }
                             catch (e) {
@@ -760,7 +764,7 @@ const parsePdf = async (buffer, config) => {
                         });
                     }
                     catch (e) {
-                        (0, errorUtils_1.logWarning)(`Failed to process image ${attachmentName}:`, config, e);
+                        (0, errorUtils_js_1.logWarning)(`Failed to process image ${attachmentName}:`, config, e);
                     }
                 }
             }

package/dist/parsers/PowerPointParser.d.ts

@@ -21,7 +21,7 @@
  * @module PowerPointParser
  * @see https://www.ecma-international.org/publications-and-standards/standards/ecma-376/
  */
-import { OfficeParserAST, OfficeParserConfig } from '../types';
+import { OfficeParserAST, OfficeParserConfig } from '../types.js';
 /**
  * Parses a PowerPoint presentation (.pptx) and extracts slides and notes.
  *