FastAPI Development

Instructions

Build production-grade FastAPI applications following these patterns.

Async endpoint design:

• Use async def for endpoints performing I/O (database queries, HTTP calls, file operations)
• Use plain def for CPU-bound endpoints — FastAPI runs them in a thread pool automatically
• Await all async calls; never mix sync blocking calls inside async def endpoints

Pydantic models for request/response:

• Define request bodies as Pydantic BaseModel subclasses with type annotations and validators
• Define response models with response_model parameter on route decorators to control serialization
• Use Field(...) for validation constraints (min_length, ge, le, regex)
• Separate create, update, and response schemas (e.g., UserCreate, UserUpdate, UserResponse)

Dependency injection:

• Use Depends() for database sessions, authenticated users, configuration, and shared services
• Create generator dependencies with yield for resource lifecycle management (open/close DB sessions)
• Layer dependencies: auth depends on DB session, endpoint depends on auth
• Define reusable dependencies in a dedicated deps.py module

SQLAlchemy async with asyncpg:

• Use create_async_engine with asyncpg driver for PostgreSQL (postgresql+asyncpg://)
• Create sessions with async_sessionmaker and AsyncSession
• Use select() statements with await session.execute() — avoid legacy Query API
• Run migrations with Alembic configured for async (run_async)

Middleware:

• Add CORSMiddleware with explicit allow_origins, allow_methods, allow_headers
• Implement custom middleware for request logging, timing, and authentication token validation
• Order matters: middleware executes in reverse declaration order

Docker deployment:

• Use multi-stage builds: builder stage installs dependencies, runtime stage copies only what’s needed
• Run with Uvicorn: uvicorn app.main:app --host 0.0.0.0 --port 8000 --workers 4
• Set --workers based on CPU cores (2 × cores + 1 as starting point)
• Use health check endpoint (/health) in Docker Compose and orchestrators

Patriot University reference architecture:

• FastAPI application server with PostgreSQL 16 for relational data
• Redis 7 for caching, session storage, and rate limiting
• MinIO for S3-compatible object storage (documents, media)
• Docker Compose orchestration with service health checks and restart policies