How do you implement architectural decisions when working with FastAPI?

When working with FastAPI, I adhere to a modular architecture, dividing the project into layers: routing, business logic, data access, and models.

• Routes (routers) are organized by functional areas, using APIRouter for convenience and scalability.
• Services contain business logic to separate it from controllers.
• Pydantic models are used for data validation and serialization.
• For database interaction, I use an ORM (e.g., SQLAlchemy) with an asynchronous engine to avoid blocking request processing.
• Configuration and dependencies are placed in separate modules, applying Dependency Injection through FastAPI.

An example of a simple route with a dependency:

from fastapi import FastAPI, Depends

app = FastAPI()

def get_db():
    db = ...  # create session
    try:
        yield db
    finally:
        db.close()

@app.get("/items/{item_id}")
async def read_item(item_id: int, db=Depends(get_db)):
    item = db.query(Item).get(item_id)
    return item

This approach facilitates testing, maintenance, and extension of the application.