Reader

Reader turns uploaded PDFs into a private reading and listening workspace. It extracts text in the browser, reads it with the Web Speech API, and uses Gemini for summaries, quizzes, and sentence explanations.

What works

• Email/password authentication with Firebase Authentication
• Per-user PDF storage and library metadata with Firebase Storage and Firestore
• PDF text extraction with PDF.js
• Browser text-to-speech with pause, resume, and stop controls
• Gemini summaries, quizzes, and sentence explanations
• Responsive reader and library views

Only text-based PDFs are supported. Reader does not currently extract text from EPUB files, scanned PDFs, or image-only PDFs.

Development

Reader requires Node.js 22.12 or newer. Copy the environment template and add credentials for a Firebase project with Email/Password Authentication, Firestore, and Storage enabled.

cp .env.example .env.local
npm ci
npm run dev

The development server runs at http://localhost:9002. Gemini-backed features require GEMINI_API_KEY. Authenticated server routes use Application Default Credentials; Google hosting supplies them automatically, while local development can set GOOGLE_APPLICATION_CREDENTIALS to a service-account JSON file.

Deploy the checked-in Firebase rules before using a shared environment:

firebase deploy --only firestore:rules,storage

Checks

npm run format:check
npm run lint
npm run typecheck
npm test
npm run build

The unit suite covers the state hooks, logging, and error handling. Firebase, Gemini, and browser speech synthesis still require integration testing against configured services.