PDF Extracter

Complete source code for PDF Extracter, a project to extract and normalize data from PDFs.
Frontend is built with Next.js, and backend is powered by FastAPI.

---

Table of Contents

1. Project Overview
2. Features
3. Architecture
4. Models
5. Setup Instructions
6. Running Locally
7. Deployment
8. Environment Variables
9. License

---

Project Overview

PDF Extracter allows users to upload PDF documents and extract structured data using multiple extraction models. The system normalizes the extracted data into a consistent format for downstream usage.

---

Features

• Upload PDF files from frontend
• Choose extraction model (OmniDocs or Docling)
• Backend processing with FastAPI
• Normalized output for easy consumption
• OAuth login support via Google (NextAuth.js)
• Supports multiple file uploads
• Handles complex layouts and tables

---

Architecture

PDF Extracter
│
├── backend (FastAPI)
│   ├── app
│   │   ├── main.py              # FastAPI entrypoint
│   │   ├── extractor_factory.py # Factory for model selection
│   │   ├── facade.py            # Facade for extraction pipeline
│   │   ├── utils                # Helper functions, normalization
│   │   └── models               # Model handlers (OmniDocs, Docling)
│   ├── requirements.txt
│   └── deploy.py                # Modal deployment script
│
├── frontend (Next.js)
│   ├── app          # Pages and API routes
│   ├── components   # UI components
│   ├── lib          # API functions, utilities
│   ├── public
│   ├── .env.local   # Local environment variables
│   ├── package.json
│   └── next.config.mjs
│
└── README.md

---

Models

OmniDocs

• Proprietary PDF extraction model
• Handles structured and semi-structured documents
• Returns normalized JSON output

Docling

• Open-source document extraction model
• Good for text-heavy PDFs
• Returns normalized JSON output

---

Setup Instructions

Backend

cd backend
python -m venv venv

# Activate the virtual environment
# Windows
venv\Scripts\activate
# macOS/Linux
source venv/bin/activate

pip install -r requirements.txt

Frontend

cd frontend
npm install

Create a .env.local file with the following variables:

NEXTAUTH_URL=https://<your-vercel-subdomain>.vercel.app
NEXTAUTH_SECRET=<random-string>
GOOGLE_CLIENT_ID=<google-client-id>
GOOGLE_CLIENT_SECRET=<google-client-secret>

---

Running Locally

Backend

uvicorn app.main:app --reload

Access at: http://localhost:8000

Frontend

npm run dev

Access at: http://localhost:3000

---

Deployment

Frontend (Vercel)

1. Go to Vercel dashboard → Import Project from GitHub
2. Set Root Directory: frontend
3. Add environment variables (see above)
4. Deploy the project

Access live app at:

https://<your-vercel-subdomain>.vercel.app

Backend

• Deploy with Modal (deploy.py) or any cloud provider
• Ensure frontend points to the backend API URL

---

Environment Variables

Frontend (.env.local)

NEXTAUTH_URL=https://<your-vercel-subdomain>.vercel.app
NEXTAUTH_SECRET=<random-string>
GOOGLE_CLIENT_ID=<google-client-id>
GOOGLE_CLIENT_SECRET=<google-client-secret>

Backend (.env)

Add backend-specific credentials if needed.

Note: Ensure Google OAuth redirect URIs match your deployed domain:

https://<your-vercel-subdomain>.vercel.app/api/auth/callback/google

---

.gitignore

Ensure .gitignore excludes:

• node_modules
• .next
• venv
• __pycache__
• *.pyc
• .env.local
• .env

---

License

MIT License © Suhas-30

---

Need help? Open an issue or reach out to the maintainer.