extract2md/package.json

Client-side PDF to Markdown conversion with OCR and optional LLM rewrite. Core dependencies bundled for offline use.

{
  "name": "extract2md",
  "version": "2.0.0",
  "description": "Client-side PDF to Markdown conversion with OCR and optional LLM rewrite. Core dependencies bundled for offline use.",
  "main": "dist/assets/extract2md.umd.js",
  "module": "dist/assets/extract2md.esm.js",
  "type": "module",
  "types": "dist/assets/extract2md.d.ts",
  "scripts": {
    "build": "webpack",
    "prepublishOnly": "npm run build",
    "postinstall": "node scripts/postinstall.js",
    "test": "node test/simple.test.js"
  },
  "keywords": [
    "pdf",
    "markdown",
    "ocr",
    "tesseract.js",
    "pdf.js",
    "webllm",
    "llm",
    "client-side",
    "text-extraction",
    "pdf to markdown",
    "offline"
  ],
  "author": "Hashan Wickramasinghe <hashanwickramasinghe@gmail.com>",
  "license": "MIT",
  "dependencies": {
    "@mlc-ai/web-llm": "^0.2.79",
    "pdfjs-dist": "^5.2.133",
    "tesseract.js": "^5.0.5"
  },
  "devDependencies": {
    "@babel/core": "^7.24.0",
    "@babel/preset-env": "^7.24.0",
    "babel-loader": "^9.1.3",
    "copy-webpack-plugin": "^12.0.2",
    "webpack": "^5.90.3",
    "webpack-cli": "^5.1.4"
  },
  "files": [
    "dist",
    "scripts",
    "README.md",
    "MIGRATION.md"
  ],
  "homepage": "https://github.com/hashangit/Extract2MD#readme",
  "repository": {
    "type": "git",
    "url": "git+https://github.com/hashangit/Extract2MD.git"
  },
  "bugs": {
    "url": "https://github.com/hashangit/Extract2MD/issues"
  }
}