Real estate scraper and visualization tool for nehnutelnosti.sk with multi-objective optimization (Pareto-optimal solutions).

• Web Scraping: Automated scraping of flats and houses from nehnutelnosti.sk
• Interactive Map: Leaflet map with marker clustering showing all properties
• Pareto Optimization: Highlights properties that are Pareto-optimal based on:
  • Minimize price
  • Minimize price per m²
  • Maximize rooms
  • Maximize year built
• Filtering: Filter by type, price, rooms, area, year, and map bounds
• Geocoding: Automatic geocoding of addresses using Nominatim (with caching)
• Scheduled Scraping: Optional scheduled scraping via APScheduler

docker-compose up --build

Services:

• API: http://localhost:8000/docs (FastAPI Swagger UI)
• Web: http://localhost:5173 (React frontend)
• Database: PostgreSQL on port 5432

Create a .env file (optional):

POSTGRES_USER=opti
POSTGRES_PASSWORD=opti
POSTGRES_DB=optihome
CORS_ORIGINS=http://localhost:5173
VITE_API_BASE_URL=http://localhost:8000

# Scheduler (optional)
ENABLE_SCHEDULER=false
SCHEDULE_HOUR=2
SCHEDULE_MINUTE=0
SCRAPE_PAGES_PER_RUN=5

Trigger scraping via API:

# Scrape flats (2 pages)
curl -X POST "http://localhost:8000/scrape/run?kind=flat&pages=2"

# Scrape houses (1 page)
curl -X POST "http://localhost:8000/scrape/run?kind=house&pages=1"

• GET /properties - List properties with filters
• GET /properties/{id} - Get property details
• GET /properties/pareto - Get Pareto-optimal properties
• POST /scrape/run - Trigger scraping

docker compose exec backend pytest

• Backend: FastAPI + SQLAlchemy + PostgreSQL
• Frontend: React + TypeScript + Vite + Leaflet
• Scraping: BeautifulSoup4 + httpx with rate limiting
• Geocoding: Nominatim with database caching
• Scheduler: APScheduler for periodic scraping

cd backend
pip install -r requirements.txt
uvicorn app.main:app --reload

cd frontend
npm install
npm run dev