Final Project

An exercise to put to practice software development teamwork, subsystem communication, containers, deployment, and CI/CD pipelines. See instructions for details.

NYU CS & Math Course Planner

A course recommender designed to help a Computer Science or Math student at NYU create a four-year plan by using an LLM to suggest courses from the CAS catalog that fulfill their requirements and classes relevant to their interests.

Live Deployment: http://159.65.190.132:5000/

Container Images

This project uses the following custom container images, hosted on Docker Hub:

• Web Application: apoorvib/web-app

The application also depends on the official MongoDB image:

• Database: mongo:7

Team Members

• Frontend: Anshu Aramandla
• Backend (LLM/recommendation): Apoorv Belgundi
• Backend (CRUD operations): Harrison Coon
• DB Setup & Login: Kylie Lin
• Docker & Integration: Jacob Ng

Instructions

Environment variables

Create a file at web-app/.env by copying web-app/.env.example and filling in values appropriate for your environment. Do not commit production secrets to version control — only commit web-app/.env.example with dummy/example values.

Example web-app/.env (copy into web-app/.env and edit):

MONGO_URI = mongodb://mongo:27017/course_planner
MONGO_DB_NAME = course_planner
ENVIRONMENT = development
WAIT_BEFORE_CONNECT = 2
OPENAI_API_KEY = sk-proj

• MONGO_URI: MongoDB connection string. When using Docker Compose, mongodb://mongo:27017 points to the mongo service in docker-compose.yml.
• MONGO_DB_NAME: name of the database used by the app.
• WAIT_BEFORE_CONNECT: seconds the seeder will wait before attempting a DB connection (helps when starting containers together).
• ENVIRONMENT: development or production — controls seeding/debug behavior.
• FLASK_SECRET: secret key for Flask session management. Keep this private in production.

If additional secrets/configuration files are required, include an example file (for example web-app/.env.example) and document exact steps for creating the real file(s) with the course admins.

Running the Webapp

You can run the app with Docker Compose (recommended) or directly in a local Python environment.

Run with Docker Compose (recommended)

1. Copy the example environment file and edit it as needed:

cp web-app/.env.example web-app/.env
# edit web-app/.env as needed

2. Build and start the services:

docker-compose up --build

3. Open http://127.0.0.1:5000 in your browser.

Notes:

• The web container runs start.sh, which attempts to seed the database (python -m database.seed) before starting the Flask app. If the database is already seeded, the script will warn and continue.
• To run the seeder manually while containers are running:

docker-compose run --rm web python -m database.seed

Run locally without Docker (optional)

1. From the repository root, change into web-app, create and activate a virtual environment, and install dependencies:

cd web-app
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

2. Ensure web-app/.env points to your MongoDB instance and seed the DB:

python -m database.seed

3. Start the Flask server:

python run.py

4. Open http://localhost:5000 to verify the app is running.

Stopping

• If started with Docker Compose: docker-compose down.
• If started locally: stop the server (Ctrl+C) and run deactivate to exit the virtual environment.

Testing

The web-app subsystem includes a comprehensive unit test suite targeting 80%+ code coverage of core business logic.

Test Suite Overview

• 44 unit tests covering user management, course parsing, semester planning, and database operations
• Coverage by module:
  • api/plan_utils.py: 84%
  • api/user_model.py: 80%
  • database/app_db.py: 97%
• Testing framework: pytest with mongomock for in-memory database testing

Running Tests

Install test dependencies (included in web-app/requirements.txt):

cd web-app
pip install -r requirements.txt

Run all tests:

pytest tests/

Run tests with coverage report:

pytest tests/ --cov=api --cov=database --cov-report=term-missing

Generate HTML coverage report:

pytest tests/ --cov=api --cov=database --cov-report=html

Then open htmlcov/index.html in a browser.

Test Structure

Tests are organized by module in web-app/tests/:

• test_user_model.py — User CRUD, authentication, profile management (13 tests)
• test_plan_utils.py — Course parsing, formatting, semester plan operations (18 tests)
• test_app_db.py — Database connection, seeding, indexing (12 tests)
• conftest.py — Shared pytest fixtures and environment setup
• README.md — Detailed testing documentation