# Project Memory: cryptocurrency2

## Project Overview
**Name**: Crypto Intelligence Hub (cryptocurrency2)  
**Type**: Cryptocurrency data collection and analysis platform  
**Language**: Python  
**Framework**: Gradio (UI) + FastAPI (Backend)  
**Deployment**: Hugging Face Spaces (Docker)

## Key Features
- 📊 Comprehensive dashboard for crypto data sources and AI models
- 📚 Access to 200+ free cryptocurrency data sources
- 🤖 Integration with Hugging Face AI models for sentiment analysis
- 💭 Sentiment analysis using specialized financial and crypto models
- 🔌 FastAPI backend for advanced services
- 📡 WebSocket support for real-time data broadcasting

## Project Structure

### Main Entry Points
- `app.py` - Main Gradio application (Hugging Face Space)
- `api_server_extended.py` - FastAPI backend server
- `hf_unified_server.py` - Unified Hugging Face server
- `ai_models.py` - Hugging Face model management

### Key Directories
- `api/` - API modules (auth, endpoints, websocket, data services)
- `backend/` - Backend services and routers
- `collectors/` - Data collectors (market data, news, sentiment, on-chain, whale tracking)
- `database/` - Database management modules
- `api-resources/` - JSON files containing API resources and configurations
- `docs/` - Project documentation
- `archive/` - Old/archived files

### Configuration Files
- `config.py` - Application settings and Hugging Face model configurations
- `requirements.txt` - General Python dependencies
- `requirements_hf.txt` - Dependencies for Hugging Face Spaces
- `Dockerfile` - Docker configuration for Hugging Face Spaces
- `docker-compose.yml` - Docker Compose configuration

## AI Models Used
- **FinBERT** (`ProsusAI/finbert`) - Financial sentiment analysis
- **CryptoBERT** (`ElKulako/cryptobert`) - Cryptocurrency sentiment analysis
- **Twitter-RoBERTa** (`cardiffnlp/twitter-roberta-base-sentiment-latest`) - Social media sentiment
- **BART** (`facebook/bart-large-cnn`) - Text summarization

## Data Sources Supported

### Market Data
- CoinGecko (free)
- CoinMarketCap
- Binance Public API
- CoinCap
- And more...

### Blockchain Explorers
- Etherscan
- BscScan
- TronScan
- Blockchair
- And more...

### RPC Nodes
- Infura
- Alchemy
- Ankr
- PublicNode
- And more...

## Key Data Files
- `api-resources/crypto_resources_unified_2025-11-11.json` - Unified crypto resources
- `all_apis_merged_2025.json` - Merged API configurations
- `data/crypto_monitor.db` - SQLite database
- `data/feature_flags.json` - Feature flags configuration

## Environment Variables
- `HF_TOKEN` or `HUGGINGFACE_TOKEN` - Hugging Face authentication token
- `USE_FASTAPI_HTML` - Toggle FastAPI+HTML mode (default: true in Docker)
- `USE_GRADIO` - Toggle Gradio UI mode (default: false in Docker)
- `DOCKER_CONTAINER` - Docker environment detection

## Deployment
- **Primary**: Hugging Face Spaces (Docker)
- **Port**: 7860 (default)
- **SDK**: Docker
- **License**: MIT

## Technology Stack
- Python 3.x
- Gradio (UI framework)
- FastAPI (Backend API)
- SQLite (Database)
- Docker (Containerization)
- Hugging Face Transformers (AI models)
- Plotly (Visualization)
- Pandas (Data processing)
- WebSocket (Real-time communication)

## Key Components

### Collectors
- Market data collectors
- News collectors
- Sentiment analyzers
- On-chain data collectors
- Whale tracking
- RPC node managers
- Blockchain explorers

### API Services
- Authentication
- Data endpoints
- Pool endpoints
- WebSocket services
- Data broadcasting
- Monitoring services
- Integration services

### Database
- SQLite database (`crypto_monitor.db`)
- Database migrations
- Data access layer
- Models definitions

## Documentation Files
- `README.md` - Main project documentation (Persian/English)
- `PROJECT_STRUCTURE.md` - Project structure documentation
- `DEPLOYMENT_CHECKLIST.md` - Deployment guide
- `HF_DEPLOYMENT_QUICKSTART.md` - Hugging Face deployment quick start
- Various component-specific READMEs in `docs/` directory

## Notes
- Project supports both Persian and English languages
- Uses free API sources primarily
- Designed for Hugging Face Spaces deployment
- Supports both Gradio UI and FastAPI+HTML modes
- Real-time data updates via WebSocket
- Comprehensive sentiment analysis capabilities