aerobytes

A modern food delivery service that scrapes live data from UCSC dining locations.

Features

• ✅ Real-time scraping from UCSC nutrition website
• ✅ Server-side scraping (no CORS issues)
• ✅ Caching system (7-minute cache to reduce server load)
• ✅ Graceful error handling
• ✅ No mock data - all data comes from live scraping

Tech Stack

• Frontend: React 18
• Backend: Node.js + Express
• Scraping: Axios + Cheerio
• Caching: In-memory cache with TTL

Setup

Install Dependencies

npm install

Run Development Server

Start both backend and frontend:

npm run dev

Or run them separately:

Backend (Terminal 1):

npm run server

Frontend (Terminal 2):

npm start

API Endpoints

GET /api/dining-locations

Returns all available dining locations from UCSC.

Response:

[
  {
    "id": "01",
    "name": "Cowell/Stevenson Dining Hall",
    "type": "dining_hall",
    "supports_drone": true,
    "is_open": true
  }
]

GET /api/menu-items?locationId=<LOCATION_ID>

Returns menu items for a specific dining location.

Response:

[
  {
    "id": "01-pizza",
    "dining_location_id": "01",
    "name": "Pizza",
    "calories": 250,
    "is_available": true
  }
]

Architecture

Frontend (React)
   ↓ fetch()
Backend API (Express)
   ↓ axios + cheerio
https://nutrition.sa.ucsc.edu/

Caching

• Cache duration: 7 minutes (configurable in backend/cache.js)
• Cache is checked before each scrape
• Stale cache is returned as fallback on errors

Error Handling

• Network failures return empty arrays (graceful degradation)
• HTML structure changes are handled with multiple selector fallbacks
• Server never crashes - all errors are caught and logged

Environment Variables

Create a .env file (optional):

REACT_APP_API_URL=http://localhost:3001
PORT=3001

Notes

• All scraping happens server-side to avoid CORS issues
• No mock data is used - everything comes from live scraping
• The scraper handles multiple HTML structure variations
• Calories are extracted when available, otherwise set to null