# recipe-document-converter

🍳 **A Docker-based import service for converting recipe documents (PDF, Word, Excel, Images) into structured JSON data.**

## 🎯 Project Overview

This repository contains a **microservice architecture** designed to:

1. **Extract** recipe content from multiple document formats
2. **Structure** unorganized data using LLM (Mistral) for intelligent parsing
3. **Integrate** seamlessly with recipe applications via REST API
4. **Scale** independently using Docker containerization

## 📦 Structure

```
recipe-document-converter/
├── recipe-document-converter/    # Main import service (this is the actual service)
│   ├── src/                     # TypeScript source code
│   ├── Dockerfile               # Container definition
│   ├── docker-compose.yml       # Multi-service orchestration
│   ├── package.json             # Dependencies
│   └── README.md                # Service documentation
└── README.md                    # This file
```

## 🚀 Quick Start

### Using Docker Compose (Recommended)

```bash
cd recipe-document-converter

# Start the service
docker-compose up -d

# Test the service
curl http://localhost:3000/health
```

### Local Development

```bash
cd recipe-document-converter

# Install dependencies
npm install

# Start development server
npm run start:dev
```

## 📖 API Endpoints

| Method | Endpoint      | Description                    |
|--------|---------------|--------------------------------|
| `GET`  | `/health`     | Health check                   |
| `POST` | `/import/pdf` | Import and extract PDF recipe  |

## 📚 Full Documentation

See [recipe-document-converter/README.md](recipe-document-converter/README.md) for complete documentation.

## 🔮 Planned Features

- [x] PDF extraction
- [x] Basic recipe structuring
- [ ] Mistral LLM integration
- [ ] Excel support
- [ ] Word support
- [ ] Image OCR support
- [ ] Web scraping

## 🛠️ Tech Stack

- **NestJS** — Node.js framework
- **TypeScript** — Type safety
- **Docker** — Containerization
- **pdf-parse** — PDF extraction
- **Zod** — Schema validation (coming soon)
- **Mistral AI** — LLM integration (coming soon)

## 🤝 Contributing

Contributions welcome! Open an issue or submit a PR.

## 📄 License

MIT