Urban Diner Food Recommendation System

A Flask college project that recommends dishes from a local food dataset based on cuisine, diet, spice level, health goal, mood, budget, and search text.

Setup

1. Create and activate a virtual environment:

python3 -m venv .venv
source .venv/bin/activate

2. Install dependencies:

pip install -r requirements.txt

3. Run the app:

python3 app.py

4. Open the site:

http://127.0.0.1:8080

If port 8080 is busy, run with another port:

PORT=5099 python3 app.py

Optional Environment Variables

The app works without these for local demos.

SECRET_KEY=replace-with-a-local-secret
DATABASE_URL=sqlite:///food_recommendation.db
ADMIN_USERNAME=local-admin-name
ADMIN_PASSWORD=local-admin-password
OPENROUTER_API_KEY=optional-openrouter-api-key
OPENROUTER_MODEL=nvidia/nemotron-3-ultra-550b-a55b:free
OPENROUTER_SITE_URL=http://127.0.0.1:8080
OPENROUTER_TIMEOUT_SECONDS=55
ENABLE_DEV_OTP_RESET=1

Do not commit a real .env file. The repository ignores .env, local SQLite databases, virtual environments, and local zip archives.

Notes

• Account creation and login use the local SQLite database created in the Flask instance/ folder. New users are signed in immediately after registration.
• Password reset is disabled by default for production safety. Set ENABLE_DEV_OTP_RESET=1 locally to enable the development reset-code helper; it is ignored on Vercel/production.
• Food inventory is stored in the configured database. Missing dishes from datasets/foods.csv are added on startup without overwriting existing admin edits.
• The chatbot uses OpenRouter when OPENROUTER_API_KEY is set, otherwise it uses the built-in fallback recommender. Free models can be slow or rate-limited, so OPENROUTER_TIMEOUT_SECONDS controls how long the app waits before falling back.
• Admin inventory add, update, and delete actions require ADMIN_USERNAME and ADMIN_PASSWORD.

Tests

python3 -m unittest -v