Async FastAPI SQLAlchemy Template

This is a ready-to-run FastAPI backend with async SQLAlchemy, which can be used as a template. It also uses alembic for database migrations and pytest for testing.

This project is heavily inspired by Full Stack FastAPI Template and demo-fastapi-async-sqlalchemy, but has the following features:

• The API implements basic CRUD functionality for the users as well as auth logic.
• No need in docker usage, the code can be run locally. Some instructions for setup, debugging, etc. can be found below.
• Tests basic implementation with pytest, suitable for async SQLAlchemy, with SQLite or Postgres powered by testing.postgresql as a database.

Setup

The project is structured so that all backend-related files are kept inside ./backend folder and not the project's root folder (besides the .env file which can be used for both parts - frontend and backend). This also means that all commands should be run from the backend folder and in further docs this is assumed unless the project's root folder is explicitly mentioned.

The code was written with Python 3.12, but 3.11 should also work, so ensure you have the supported version and install dependencies with pip install -r requirements-dev.txt.

The settings are configured via environment variables. The common way to set them up is via the .env file in the project's root folder. Check the config file to see which environment variables can or must be set up. The dialect and driver for DATABASE_URL should be postgresql+asyncpg.

Also, specify the PYTHONPATH environment variable as ./backend.

Database initialization

The database schema must be initialized via alembic. Ensure the DATABASE_URL environment variable is specified and run alembic upgrade head.

After the schema is initialized, the initial superuser must be created. This can be done with the python -m app.initial_data command.

Running the API

When dependency packages are installed, environ variables are set up, the database is initialized and accessible, the API can be run with fastapi dev app/main.py.

Running tests

Tests can be run with SQLite or Postgres as a database (see conftest) with python -m pytest ./app/tests.

To use Postgres, the PostgreSQL server must be installed and accessible via the PATH variable (see testing.postgresql README). It should not be difficult to implement Postgres support in a different way though.

VS Code settings

General settings

Set python.testing.pytestPath as python -m pytest. If you want to use flake8 as a linting tool, set flake8.cwd as ${workspaceFolder}/backend.

Debugging

For debug mode, you can use the following configuration:

{
    "name": "Python Debugger: FastAPI",
    "type": "debugpy",
    "request": "launch",
    "cwd": "${workspaceFolder}/backend",
    "module": "uvicorn",
    "args": [
        "app.main:app",
        "--reload"
    ],
    "jinja": true
}

Testing

With configured python.testing.pytestPath VS Code testing should be able to work. However, the global Run Tests option does not work properly because asyncio mode is not fetched from pytest.ini file. Solutions are welcome, but currently only Run Test can be used: