🚀 DocGen – Code Documentation Generator

DocGen is a AI Project that generates complete professional documentation for any programming code or project file.

It works completely offline using:

• Ollama
• Qwen2.5-Coder Model
• Django REST Backend
• React + Tailwind Frontend
• PDF Export Support
• Doc Export Support

---

✨ Features

✅ Paste code and generate documentation instantly
✅ Upload code files (.py, .cpp, .java, .js)
✅ AI-generated structured Markdown documentation
✅ Export documentation as a perfectly aligned PDF
✅ Modern SaaS-grade UI (Gemini-level design)
✅ Fully Offline (No OpenAI / No Gemini API needed)

---

🏗️ Tech Stack

Layer	Technology
Frontend	React, Tailwind CSS, Framer Motion
Backend	Django, Django REST Framework
AI Model	Ollama + Qwen2.5-Coder:7B
PDF Generator	ReportLab (Platypus Engine)

---

⚙️ Setup Instructions (Run on Any PC)

---

✅ 1. Clone Repository

# Clone and enter repo
git clone https://github.com/SanjayMarathi/DocGen.git
cd DocGen

Start Ollama (pull model if needed)

ollama pull qwen2.5-coder:7b
ollama serve

Start backend (in another terminal)

cd backend
python -m venv venv
.\venv\Scripts\activate
pip install django djangorestframework django-cors-headers requests reportlab
python manage.py migrate
python manage.py runserver 8000

The venv activation above for powershell needs to be changed to venv/bin/activate if using bash.

Start frontend (in a new terminal)

cd frontend
npm install
npm start

Prerequisites

• Node.js + npm (for frontend)

• Python 3.10+ (for backend)

• Ollama installed and running locally (project uses qwen2.5-coder:7b)

• API endpoints:

  • POST http://127.0.0.1:8000/api/generate/ (streaming documentation)
  • POST http://127.0.0.1:8000/api/pdf/ (returns generated PDF)

Ollama (Local LLM)

The backend calls Ollama at http://localhost:11434.

Quick test

• Paste or upload code in the UI → Click Generate Documentation.
• Click EXPORT PDF to download the generated PDF.

Troubleshooting ⚠️

• Model not responding. Check Ollama. → Ensure Ollama is running and the model is available.
• If the frontend or backend use different host/ports, update frontend/src/App.js and backend/generator/views.py accordingly.
• CORS is already enabled in backend/settings.py (CORS_ALLOW_ALL_ORIGINS = True).

Convenience tip

If you want npm start to start the frontend from the repo root, add this script to the root package.json under scripts:

"scripts": {
  "start": "npm --prefix frontend start"
}

---