@damorris25/content-extractor

Extract text content and metadata from unstructured file formats — Office documents, OpenDocument, PDF, images, audio, email, CAD, and more. Inspired by Apache Tika, built browser-first with a single runtime dependency (fflate, MIT).

This is the TypeScript implementation of content-extractor. A Go implementation lives in the same repository; both produce identical output, enforced by golden-file parity tests.

Install

npm install @damorris25/content-extractor

Optional peer dependencies:

npm install pdfjs-dist    # PDF text extraction
npm install tesseract.js  # OCR for images

Quick Start

import { extract } from '@damorris25/content-extractor';

const bytes = new Uint8Array(await file.arrayBuffer()); // or fs.readFileSync
const result = await extract(bytes, { filename: 'report.docx' });

console.log(result.text);        // extracted text content
console.log(result.contentType); // detected MIME type
console.log(result.metadata);    // Record<string, string[]> — title, creator, dates, ...

Works in browsers and Node.js. Input is always a Uint8Array; extraction is async and accepts an optional AbortSignal:

const controller = new AbortController();
const result = await extract(bytes, {
  filename: 'big.xlsx',
  signal: controller.signal,
});

Supported Formats

| Family | Formats | Extracted | |--------|---------|-----------| | Microsoft Office | DOCX, XLSX, PPTX | Text + document metadata | | OpenDocument | ODT, ODS, ODP | Text + document metadata | | PDF | PDF (text layer) | Text (via pdfjs-dist peer) | | Web/Markup | XML, HTML, SVG | Visible text, titles | | Structured data | JSON, CSV, TSV, YAML, Markdown | Pass-through + detection | | Rich text | RTF | Text with Unicode support | | Email | EML (RFC 2822) | Body text + headers | | Images | JPEG, PNG, TIFF/GeoTIFF, BMP, GIF, WebP, HEIC, HEIF, AVIF | EXIF/metadata, optional OCR | | Audio | MP3, OGG, FLAC, WAV | ID3v2 / Vorbis / RIFF metadata | | CAD | DXF | Text entities, layers, metadata |

Security

Built for untrusted input: ZIP bomb protection (compression ratio + size caps), integer overflow checks, NUL byte stripping, and bounded parsing for ISOBMFF containers. Report vulnerabilities via GitHub Security Advisories.

Documentation

Full documentation, architecture contracts, benchmarks, and the Go implementation: github.com/damorris25/content-extractor

License

MIT © 2026 Dana Morris