officeparser 7.1.0 → 7.2.1

package/dist/utils/moduleLoader.js

@@ -19,7 +19,7 @@ async function loadNodeEsmModule(specifier) {
     // This is especially important in Node 18 for sub-paths of packages.
     try {
         const { pathToFileURL } = await import('url');
-        // @ts-ignore - require.resolve is available in Node.js
+        // @ts-ignore - require.resolve is available in Node.js CJS context
         const absolutePath = require.resolve(specifier);
         const fileUrl = pathToFileURL(absolutePath).href;
         return import(fileUrl);
@@ -29,6 +29,16 @@ async function loadNodeEsmModule(specifier) {
         return import(specifier);
     }
 }
+/**
+ * Returns true if require.resolve is available in this runtime context.
+ * It is NOT available in native ESM (e.g. the .mjs wrapper) or browser environments.
+ * Checking this guards against a ReferenceError that would silently trigger
+ * the bundled fallback path for non-bundled ESM consumers.
+ */
+function isRequireAvailable() {
+    // @ts-ignore - require may not exist in ESM context
+    return typeof require !== 'undefined' && typeof require.resolve === 'function';
+}
 /**
  * Specialized loader for file-type
  */
@@ -36,10 +46,19 @@ async function loadFileType() {
     if (!envUtils_js_1.isBrowser) {
         // Ensure environment polyfills for Node.js 18 support
         (0, envUtils_js_1.ensureEnvPolyfills)();
-        // Node.js path: Use dynamic import wrapper for CJS compatibility
-        return loadNodeEsmModule('file-type');
+        if (isRequireAvailable()) {
+            // Check if node_modules is available at runtime.
+            // @ts-ignore - require.resolve is available in Node.js CJS context
+            require.resolve(String('file-type'));
+            // node_modules present: use the path-resolved loader for Node 18 ESM compatibility
+            return loadNodeEsmModule('file-type');
+        }
     }
-    // Browser path: standard dynamic import() is handled by bundlers (e.g. esbuild/Vite)
+    // Covers three cases with one return:
+    //   1. Node.js standalone/bundled (SEA or bundled CJS): bundler inlines the module at build time.
+    //   2. Node.js native ESM (no require available): runtime resolves the bare specifier.
+    //   3. Browser: bundler (esbuild/Vite) handles the static-looking dynamic import().
+    // Note: bypasses the Node 18 sub-path ESM fix in loadNodeEsmModule, but bundlers handle it.
     return import('file-type');
 }
 /**
@@ -49,14 +68,39 @@ async function loadPdfJs() {
     if (!envUtils_js_1.isBrowser) {
         // Ensure environment polyfills for Node.js 18 support
         (0, envUtils_js_1.ensureEnvPolyfills)();
-        // Node.js environment: require legacy build for stability with ESM-only main
-        try {
-            return await loadNodeEsmModule('pdfjs-dist/legacy/build/pdf.mjs');
-        }
-        catch {
-            return await loadNodeEsmModule('pdfjs-dist');
+        if (isRequireAvailable()) {
+            // @ts-ignore - require.resolve is available in Node.js CJS context
+            const pkgExists = (() => { try {
+                require.resolve(String('pdfjs-dist'));
+                return true;
+            }
+            catch {
+                return false;
+            } })();
+            if (pkgExists) {
+                // node_modules present: try the legacy build path first for stability with
+                // ESM-only main, then fall back to the package root.
+                // Errors from these loaders are NOT swallowed into the bundled-fallback path.
+                try {
+                    return await loadNodeEsmModule('pdfjs-dist/legacy/build/pdf.mjs');
+                }
+                catch {
+                    return await loadNodeEsmModule('pdfjs-dist');
+                }
+            }
         }
+        // Standalone/bundled fallback for Node.js (SEA, bundled CJS without node_modules, or native ESM context):
+        // Load both PDF.js and its worker, and register the worker on globalThis to enable the offline fake-worker.
+        // @ts-ignore - Ignore type check for local .mjs files in node_modules/packaging
+        const [pdfjs, pdfjsWorker] = await Promise.all([
+            // @ts-ignore - mjs imports may not have types
+            import('pdfjs-dist/legacy/build/pdf.mjs'),
+            // @ts-ignore - mjs imports may not have types
+            import('pdfjs-dist/legacy/build/pdf.worker.mjs')
+        ]);
+        globalThis.pdfjsWorker = pdfjsWorker;
+        return pdfjs;
     }
-    // Browser environment: esbuild handles standard static-looking dynamic import()
+    // Browser environment: bundler (esbuild/Vite) handles the standard dynamic import()
     return import('pdfjs-dist');
 }

package/dist/utils/xmlUtils.d.ts

@@ -144,6 +144,15 @@ export declare const parseOfficeMetadata: (xmlContent: string) => OfficeMetadata
  * ```
  */
 export declare const parseOOXMLCustomProperties: (xmlContent: string) => Record<string, string | number | boolean | Date>;
+/**
+ * Parses OOXML application properties from `docProps/app.xml`.
+ *
+ * Application properties contain document statistics and application settings.
+ *
+ * @param xmlContent - Raw XML string from `docProps/app.xml`
+ * @returns A record of property name -> typed value
+ */
+export declare const parseOOXMLAppProperties: (xmlContent: string) => Record<string, string | number | boolean>;
 /**
  * Decodes XML entities (standard named entities, decimal, and hexadecimal entities) in a string.
  * Useful when parsing content with regular expressions instead of a full DOM parser.

package/dist/utils/xmlUtils.js

@@ -11,7 +11,7 @@
  * @module xmlUtils
  */
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.decodeXmlEntities = exports.parseOOXMLCustomProperties = exports.parseOfficeMetadata = exports.getDirectChildren = exports.getAttribute = exports.getFirstElementByTagName = exports.getRawContent = exports.getSourceSubstring = exports.serializeXml = exports.getElementsByTagName = exports.parseXmlString = exports.isElement = void 0;
+exports.decodeXmlEntities = exports.parseOOXMLAppProperties = exports.parseOOXMLCustomProperties = exports.parseOfficeMetadata = exports.getDirectChildren = exports.getAttribute = exports.getFirstElementByTagName = exports.getRawContent = exports.getSourceSubstring = exports.serializeXml = exports.getElementsByTagName = exports.parseXmlString = exports.isElement = void 0;
 const xmldom_1 = require("@xmldom/xmldom");
 const dateUtils_js_1 = require("./dateUtils.js");
 /**
@@ -221,6 +221,13 @@ const parseOfficeMetadata = (xmlContent) => {
     // Check for OOXML Core Properties
     const coreProperties = (0, exports.getElementsByTagName)(xml, "cp:coreProperties")[0];
     if (coreProperties) {
+        metadata.nativeProperties = {};
+        for (let i = 0; i < coreProperties.childNodes.length; i++) {
+            const child = coreProperties.childNodes[i];
+            if ((0, exports.isElement)(child)) {
+                metadata.nativeProperties[child.tagName] = child.textContent;
+            }
+        }
         // Step 3: Extract title (Dublin Core element)
         const title = (0, exports.getElementsByTagName)(coreProperties, "dc:title")[0];
         if (title && title.textContent)
@@ -248,11 +255,21 @@ const parseOfficeMetadata = (xmlContent) => {
         const subject = (0, exports.getElementsByTagName)(coreProperties, "dc:subject")[0];
         if (subject && subject.textContent)
             metadata.subject = subject.textContent;
+        const keywords = (0, exports.getElementsByTagName)(coreProperties, "cp:keywords")[0];
+        if (keywords && keywords.textContent)
+            metadata.keywords = keywords.textContent;
         return metadata;
     }
     // Check for ODF Meta
     const officeMeta = (0, exports.getElementsByTagName)(xml, "office:meta")[0];
     if (officeMeta) {
+        metadata.nativeProperties = {};
+        for (let i = 0; i < officeMeta.childNodes.length; i++) {
+            const child = officeMeta.childNodes[i];
+            if ((0, exports.isElement)(child)) {
+                metadata.nativeProperties[child.tagName] = child.textContent;
+            }
+        }
         const title = (0, exports.getElementsByTagName)(officeMeta, "dc:title")[0];
         if (title && title.textContent)
             metadata.title = title.textContent;
@@ -265,6 +282,10 @@ const parseOfficeMetadata = (xmlContent) => {
         const subject = (0, exports.getElementsByTagName)(officeMeta, "dc:subject")[0];
         if (subject && subject.textContent)
             metadata.subject = subject.textContent;
+        const keywordElements = (0, exports.getElementsByTagName)(officeMeta, "meta:keyword");
+        if (keywordElements.length > 0) {
+            metadata.keywords = keywordElements.map(k => k.textContent).filter(Boolean).join(', ');
+        }
         const created = (0, exports.getElementsByTagName)(officeMeta, "meta:creation-date")[0];
         if (created && created.textContent)
             metadata.created = (0, dateUtils_js_1.parseOfficeDate)(created.textContent);
@@ -375,6 +396,37 @@ const parseOOXMLCustomProperties = (xmlContent) => {
     return result;
 };
 exports.parseOOXMLCustomProperties = parseOOXMLCustomProperties;
+/**
+ * Parses OOXML application properties from `docProps/app.xml`.
+ *
+ * Application properties contain document statistics and application settings.
+ *
+ * @param xmlContent - Raw XML string from `docProps/app.xml`
+ * @returns A record of property name -> typed value
+ */
+const parseOOXMLAppProperties = (xmlContent) => {
+    const xml = (0, exports.parseXmlString)(xmlContent);
+    const result = {};
+    const appProperties = (0, exports.getElementsByTagName)(xml, "Properties")[0];
+    if (appProperties) {
+        for (let i = 0; i < appProperties.childNodes.length; i++) {
+            const child = appProperties.childNodes[i];
+            if ((0, exports.isElement)(child)) {
+                const text = child.textContent || '';
+                if (text.toLowerCase() === 'true')
+                    result[child.tagName] = true;
+                else if (text.toLowerCase() === 'false')
+                    result[child.tagName] = false;
+                else if (!isNaN(Number(text)) && text.trim() !== '')
+                    result[child.tagName] = Number(text);
+                else
+                    result[child.tagName] = text;
+            }
+        }
+    }
+    return result;
+};
+exports.parseOOXMLAppProperties = parseOOXMLAppProperties;
 /**
  * Decodes XML entities (standard named entities, decimal, and hexadecimal entities) in a string.
  * Useful when parsing content with regular expressions instead of a full DOM parser.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "officeparser",
-  "version": "7.1.0",
+  "version": "7.2.1",
   "description": "A robust, strictly-typed Node.js and Browser library for parsing office files (.docx, .pptx, .xlsx, .odt, .odp, .ods, .pdf, .rtf, .csv, .md, .html) and generating high-fidelity outputs in Markdown, HTML, CSV, RTF, and RAG-focused chunks.",
   "funding": "https://github.com/sponsors/harshankur",
   "main": "dist/index.js",
@@ -31,16 +31,18 @@
     "sync:versions": "node scripts/sync-pdfjs-versions.js",
     "sync:docs": "mkdir -p docs/dist && cp dist/officeparser.browser.iife.js docs/dist/ && cp dist/officeparser.browser.mjs docs/dist/ && mkdir -p docs/test/files && cp test/files/* docs/test/files/",
     "lint": "eslint src",
-    "test": "npm run lint && npm run test:clean && npm run build && npm run test:license && npm run test:artifacts && npm run test:parser && npm run test:generator",
+    "test": "npm run lint && npm run test:clean && npm run build && npm run test:license && npm run test:artifacts && npm run test:parser && npm run test:generator && npm run test:cli",
     "test:baseline": "npm run test:parser:baseline && npm run test:generator:baseline",
     "test:parser": "npx tsx test/parser/testOfficeParser.ts",
     "test:parser:baseline": "npx tsx test/parser/testOfficeParser.ts baseline",
     "test:generator": "npx tsx test/generator/testOfficeGenerator.ts",
     "test:generator:baseline": "npx tsx test/generator/testOfficeGenerator.ts baseline",
     "test:artifacts": "npx tsx test/testShippingArtifacts.ts",
+    "test:cli": "npx tsx test/cli/testCli.ts",
     "test:visualizer": "node test/testVisualizer.js",
+    "test:integration": "node test/testIntegration.js",
     "test:license": "npm run sbom && node scripts/validate-licenses.js",
-    "test:clean": "rm -rf test/results test/generator/results test/generator/output test/parser/results test/parser/output",
+    "test:clean": "rm -rf test/results test/generator/results test/generator/output test/parser/results test/parser/output test/cli/results",
     "clean": "rm -rf dist && npm run test:clean",
     "sbom": "npx --yes @cyclonedx/cyclonedx-npm --output-format json --output-file dist/sbom.cdx.json --omit dev",
     "prepublishOnly": "npm run build",
@@ -123,6 +125,7 @@
     "esbuild-plugins-node-modules-polyfill": "^1.8.1",
     "eslint": "^10.3.0",
     "husky": "^9.1.7",
+    "postject": "^1.0.0-alpha.6",
     "process": "^0.11.10",
     "puppeteer": "^22.15.0",
     "tsx": "^4.21.0",