OCR App (Backend + React UI)

This repo is now a small monorepo:

• backend/: Flask + Tesseract OCR API (Cloud Run-friendly)
• frontend/: original React (Vite) UI (single image → /extract)
• frontend-v2/: “Book OCR (no login)” workflow UI (PDF/images → per-page OCR → proofread → export)

For production deployment, the recommended approach is decoupled:

• frontend-v2/ on Firebase Hosting
• backend/ on Cloud Run

See:

• docs/DEPLOYMENT.md
• docs/AUTHENTICATION.md

Local Development

1) Backend (API)

cd backend
export PORT=8080
export DISABLE_QUOTA=1
./venv/bin/python app.py

2) Frontend (React dev server)

cd frontend
npm install
npm run dev

Open http://localhost:5173/.

frontend/vite.config.js proxies /extract to http://localhost:8080 during development.

Tests (Backend)

cd backend
./venv/bin/python -m pytest -q

Deployment

• Recommended: docs/DEPLOYMENT.md
• Deep-dive / rationale: docs/DEPLOYMENT_APPROACH_1_STATIC_FRONTEND_CLOUDRUN_BACKEND.md
• Archived options (single-service Cloud Run, Vercel rewrites): docs/ARCHIVED_DEPLOYMENT_OPTIONS.md

Recommended runtime env vars:

• DAILY_LIMIT: daily request limit
• OCR_LANGUAGE: default OCR language
• FIRESTORE_DATABASE: Firestore DB name
• DISABLE_QUOTA: set to 0 in production

API

POST /extract

• Upload: multipart/form-data with image
• Optional: lang query param (tel, kan, hin, eng)

Response:

{ "status": "success", "text": "..." }