officeparser 7.2.0 → 7.2.2

package/dist/types.js

@@ -9,6 +9,8 @@ var OfficeErrorType;
 (function (OfficeErrorType) {
     /** Unsupported file extension */
     OfficeErrorType["EXTENSION_UNSUPPORTED"] = "EXTENSION_UNSUPPORTED";
+    /** Unsupported output generator format */
+    OfficeErrorType["FORMAT_UNSUPPORTED"] = "FORMAT_UNSUPPORTED";
     /** File appears to be corrupted or malformed */
     OfficeErrorType["FILE_CORRUPTED"] = "FILE_CORRUPTED";
     /** File could not be found at the specified path */
@@ -35,6 +37,14 @@ var OfficeErrorType;
     OfficeErrorType["MISSING_EMBEDDING_FUNCTION"] = "MISSING_EMBEDDING_FUNCTION";
     /** The operation was aborted */
     OfficeErrorType["OPERATION_ABORTED"] = "OPERATION_ABORTED";
+    /** ZIP entry count exceeds limit */
+    OfficeErrorType["ZIP_ENTRY_COUNT_LIMIT_EXCEEDED"] = "ZIP_ENTRY_COUNT_LIMIT_EXCEEDED";
+    /** ZIP entry missing a valid declared size */
+    OfficeErrorType["ZIP_ENTRY_INVALID_SIZE"] = "ZIP_ENTRY_INVALID_SIZE";
+    /** ZIP uncompressed size limit exceeded */
+    OfficeErrorType["ZIP_SIZE_LIMIT_EXCEEDED"] = "ZIP_SIZE_LIMIT_EXCEEDED";
+    /** Embedding call timed out */
+    OfficeErrorType["EMBEDDING_TIMEOUT"] = "EMBEDDING_TIMEOUT";
 })(OfficeErrorType || (exports.OfficeErrorType = OfficeErrorType = {}));
 /**
  * Standard warning types for OfficeParser.

package/dist/utils/configUtils.js

@@ -57,6 +57,12 @@ function isFullParserConfig(config) {
  */
 function resolveParserConfig(userConfig) {
     if (isFullParserConfig(userConfig)) {
+        if (!userConfig.decompressionLimits) {
+            userConfig.decompressionLimits = {
+                maxUncompressedBytes: 512 * 1024 * 1024,
+                maxZipEntries: 10000,
+            };
+        }
         return userConfig;
     }
     // 1. Start with full defaults (deep cloned)
@@ -65,9 +71,15 @@ function resolveParserConfig(userConfig) {
         return config;
     }
     // 2. Merge user config
-    // We handle ocrConfig specially to avoid shallow-overwriting the whole object
-    const { ocrConfig, ...rest } = userConfig;
+    // We handle ocrConfig and decompressionLimits specially to avoid shallow-overwriting the whole objects
+    const { ocrConfig, decompressionLimits, ...rest } = userConfig;
     Object.assign(config, rest);
+    if (decompressionLimits) {
+        config.decompressionLimits = {
+            ...config.decompressionLimits,
+            ...decompressionLimits,
+        };
+    }
     if (ocrConfig) {
         const { timeout, ...ocrRest } = ocrConfig;
         config.ocrConfig = {

package/dist/utils/errorUtils.js

@@ -17,6 +17,7 @@ const ERRORHEADER = "[OfficeParser]: ";
  */
 const ERROR_MESSAGES = {
     [types_js_1.OfficeErrorType.EXTENSION_UNSUPPORTED]: (ext) => `Sorry, OfficeParser currently supports docx, pptx, xlsx, odt, odp, ods, pdf, rtf, md, html, csv files only. Create a ticket in Issues on github to add support for ${ext} files. Stay tuned for further updates.`,
+    [types_js_1.OfficeErrorType.FORMAT_UNSUPPORTED]: (format) => `Sorry, OfficeGenerator does not support generating '${format}' files. Supported formats: json, text, md, html, csv, rtf, pdf, chunks.`,
     [types_js_1.OfficeErrorType.FILE_CORRUPTED]: (filepath) => `Your file ${filepath} seems to be corrupted. If you are sure it is fine, please create a ticket in Issues on github with the file to reproduce error.`,
     [types_js_1.OfficeErrorType.FILE_DOES_NOT_EXIST]: (filepath) => `File ${filepath} could not be found! Check if the file exists or verify if the relative path to the file is correct from your terminal's location.`,
     [types_js_1.OfficeErrorType.LOCATION_NOT_FOUND]: (location) => `Entered location ${location} is not reachable! Please make sure that the entered directory location exists. Check relative paths and reenter.`,
@@ -29,7 +30,11 @@ const ERROR_MESSAGES = {
     [types_js_1.OfficeErrorType.INVALID_SELECTOR]: (selector) => `Invalid selector: ${selector}`,
     [types_js_1.OfficeErrorType.INVALID_OUTPUT_MAPPING]: (output) => `Invalid output mapping: ${output}`,
     [types_js_1.OfficeErrorType.MISSING_EMBEDDING_FUNCTION]: `Semantic chunking requires an "embeddingFunction" to be provided in chunksConfig. This function must accept a string and return a Promise resolving to a number array (vector).`,
-    [types_js_1.OfficeErrorType.OPERATION_ABORTED]: `The operation was aborted.`
+    [types_js_1.OfficeErrorType.OPERATION_ABORTED]: `The operation was aborted.`,
+    [types_js_1.OfficeErrorType.ZIP_ENTRY_COUNT_LIMIT_EXCEEDED]: (limit) => `ZIP entry count exceeds limit (${limit})`,
+    [types_js_1.OfficeErrorType.ZIP_ENTRY_INVALID_SIZE]: `ZIP entry missing a valid declared size`,
+    [types_js_1.OfficeErrorType.ZIP_SIZE_LIMIT_EXCEEDED]: (limit) => `ZIP uncompressed size limit exceeded (${limit} bytes)`,
+    [types_js_1.OfficeErrorType.EMBEDDING_TIMEOUT]: (timeout) => `Embedding call timed out after ${timeout}ms`
 };
 /**
  * Lookup table for warning messages.

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/zipUtils.d.ts

@@ -13,6 +13,7 @@
  *
  * @module zipUtils
  */
+import { DecompressionLimits } from '../types.js';
 /**
  * Represents a file extracted from a ZIP archive.
  * Contains the file's path within the archive and its content as a Buffer.
@@ -69,5 +70,5 @@ interface ZipFileContent {
  *
  * @see https://pkware.cachefly.net/webdocs/casestudies/APPNOTE.TXT ZIP file format specification
  */
-export declare const extractFiles: (zipInput: Buffer, filterFn: (fileName: string) => boolean) => Promise<ZipFileContent[]>;
+export declare const extractFiles: (zipInput: Buffer, filterFn: (fileName: string) => boolean, limits: DecompressionLimits) => Promise<ZipFileContent[]>;
 export {};

package/dist/utils/zipUtils.js

@@ -17,6 +17,8 @@
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.extractFiles = void 0;
 const fflate_1 = require("fflate");
+const types_js_1 = require("../types.js");
+const errorUtils_js_1 = require("./errorUtils.js");
 /**
  * Extracts files from a ZIP archive with optional filtering.
  *
@@ -56,9 +58,39 @@ const fflate_1 = require("fflate");
  *
  * @see https://pkware.cachefly.net/webdocs/casestudies/APPNOTE.TXT ZIP file format specification
  */
-const extractFiles = (zipInput, filterFn) => {
+const extractFiles = (zipInput, filterFn, limits) => {
+    const maxUncompressedBytes = limits?.maxUncompressedBytes !== undefined && Number.isFinite(limits.maxUncompressedBytes) && limits.maxUncompressedBytes >= 0
+        ? limits.maxUncompressedBytes
+        : 512 * 1024 * 1024;
+    const maxZipEntries = limits?.maxZipEntries !== undefined && Number.isFinite(limits.maxZipEntries) && limits.maxZipEntries >= 0
+        ? limits.maxZipEntries
+        : 10000;
     return new Promise((resolve, reject) => {
-        (0, fflate_1.unzip)(new Uint8Array(zipInput.buffer, zipInput.byteOffset, zipInput.byteLength), { filter: (file) => filterFn(file.name) }, (err, decompressed) => {
+        let totalEntryCount = 0;
+        let entryCount = 0;
+        let declaredTotal = 0;
+        (0, fflate_1.unzip)(new Uint8Array(zipInput.buffer, zipInput.byteOffset, zipInput.byteLength), {
+            filter: (file) => {
+                totalEntryCount++;
+                if (totalEntryCount > maxZipEntries) {
+                    reject((0, errorUtils_js_1.getOfficeError)(types_js_1.OfficeErrorType.ZIP_ENTRY_COUNT_LIMIT_EXCEEDED, undefined, maxZipEntries));
+                    return false;
+                }
+                if (!filterFn(file.name))
+                    return false;
+                if (typeof file.originalSize !== 'number' || !Number.isFinite(file.originalSize) || file.originalSize < 0) {
+                    reject((0, errorUtils_js_1.getOfficeError)(types_js_1.OfficeErrorType.ZIP_ENTRY_INVALID_SIZE));
+                    return false;
+                }
+                entryCount++;
+                declaredTotal += file.originalSize;
+                if (declaredTotal > maxUncompressedBytes) {
+                    reject((0, errorUtils_js_1.getOfficeError)(types_js_1.OfficeErrorType.ZIP_SIZE_LIMIT_EXCEEDED, undefined, maxUncompressedBytes));
+                    return false;
+                }
+                return true;
+            }
+        }, (err, decompressed) => {
             if (err)
                 return reject(err);
             resolve(Object.entries(decompressed).map(([path, data]) => ({

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "officeparser",
-  "version": "7.2.0",
+  "version": "7.2.2",
   "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",