Taranगिनी — Disaster Management Demo

Taranगिनी is a coastal disaster alert system that scrapes real-time alerts from INCOIS (Indian National Centre for Ocean Information Services) and displays them on an interactive web interface.

🏗️ Architecture & Data Flow

┌─────────────────┐
│ INCOIS Website  │
│ (incois.gov.in) │
└────────┬────────┘
         │
         │ Scrapes alerts
         ▼
┌─────────────────────┐
│ incois_scraper.py   │
│ (Python Script)     │
└─────────┬───────────┘
          │
          │ Saves to
          ▼
┌─────────────────────┐
│   alerts.json       │
│   (Data Storage)    │
└─────────┬───────────┘
          │
          │ Loaded by
          ▼
┌─────────────────────┐
│   main.py           │
│   (FastAPI Backend) │
│   - /alerts API     │
│   - /refresh-alerts │
└─────────┬───────────┘
          │
          │ Serves data via API
          ▼
┌─────────────────────┐
│   index.html        │
│   (Frontend UI)     │
│   - Displays alerts │
│   - Refresh button  │
└─────────────────────┘

🚀 Quick Start

1. Install Dependencies

python -m venv .venv
source .venv/bin/activate  # On Windows: .\.venv\Scripts\activate
pip install -r requirements.txt

2. Run the Scraper (Optional - generates sample data if offline)

python incois_scraper.py

This creates/updates alerts.json with the latest alerts from INCOIS.

3. Start the Web Server

python -m uvicorn main:app --reload

4. Open in Browser

Navigate to: http://127.0.0.1:8000

📡 API Endpoints

• GET / - Main web interface (frontend)
• GET /alerts - JSON API returning all alerts
• GET /refresh-alerts - Runs the scraper and refreshes alerts
• GET /reload - Reloads alerts from alerts.json
• GET /db-view - Database viewer interface
• GET /alerts-nearby?lat=X&lng=Y&radius_km=Z - Get alerts near a location

🔄 How to Update Alerts

Method 1: From the Web UI

Click the "🔄 Refresh Alerts" button on the homepage. This will:

1. Run the scraper in the background
2. Fetch latest alerts from INCOIS
3. Update the database
4. Refresh the display automatically

Method 2: Manual Scraper Run

python incois_scraper.py

The backend automatically reloads alerts.json on each page load.

Method 3: Via API

curl http://localhost:8000/refresh-alerts

📁 Project Structure

tarangini-alerts/
├── main.py                 # FastAPI backend server
├── incois_scraper.py       # INCOIS website scraper
├── alerts.json             # Scraped alerts data (auto-generated)
├── requirements.txt        # Python dependencies
├── templates/
│   ├── index.html          # Main frontend UI
│   └── db_view.html        # Database viewer
├── static/
│   └── logo.jpg            # Application logo
└── README.md               # This file

🛠️ Technologies Used

• Backend: FastAPI, SQLModel, SQLite
• Frontend: HTML, CSS, JavaScript (Vanilla)
• Scraper: BeautifulSoup4, Requests, Geopy, Feedparser
• Deployment: Uvicorn (ASGI server)

📊 Database Schema

The Alert model stores:

• id - Unique identifier
• source - Alert source (e.g., "INCOIS")
• severity - emergency, warning, info, or caution
• title - Alert headline
• message - Detailed alert message
• time - Timestamp when alert was recorded
• lat, lng - Geographic coordinates (optional)

🔍 Features

✅ Real-time alert scraping from INCOIS
✅ Interactive web interface with alert cards
✅ One-click refresh from UI
✅ Severity-based color coding
✅ Location-based filtering via API
✅ Responsive mobile-friendly design
✅ Automatic fallback to sample data when offline

📝 Notes

• The scraper generates sample alerts when INCOIS website is unreachable
• Alerts are automatically loaded from alerts.json on each page load
• The database (SQLite) is used as an intermediate layer for querying and filtering

🤝 Contributing

Feel free to open issues or submit pull requests to improve the system!

📄 License

This project is for educational and demonstration purposes.