FastAPI endpoints

This repo hosts FastAPI endpoints that can be deployed in a container run on Google Cloud Run. Cloud Build is configured so that once a PR is merged into main, a new build process is triggered and the Cloud Run container is re-deployed.

Developing the code locally

We strongly recommend using a Python package management tool to manage the dependencies of this project. The tool chosen for this project is rye, a new Python packaging tool written in Rust with high performances. Since rye is compatible with PEP 621 pyproject.toml, other packaging management tools such as hatch or pdm should also work. It, however, won't work with poetry.

We also use docker to build the container image that you can test on your local machine. If you have not already installed docker, please click here to install it. Once you have installed docker, make sure that docker -v runs successfully on your command line.

Step 1: Create a repo from this template Clone this repo

First, please create a new repo using this template and use the following command to clone this repo to your local development machine:

git clone https://github.com/{YOUR_ORG_NAME}/{YOUR_REPO_NAME}.git
cd {YOUR_REPO_NAME}

Step 2: Inspect project names and path for the package

By default, the name for this project in pyproject.toml is fastapi-image. If this is not the name you prefer, please feel free to change it. Please also remember to change:

• the path of the project
• the paths in pyproject.toml
• the paths in Dockerfile
• the paths in the github workflow
• the package path in tests

Step 3: Install rye and dependencies

If you don't already have rye installed on your computer, please follow their installation guide to have it installed.

With rye, setting up a development environment is very easy. Just run this command:

rye sync

You will have a virtual environment automatically set up under your project directory with all dependencies and dev dependencies installed.

Step 4: Develop in dev mode

Run the following command to spin up a dev server:

rye run dev

This is equivalent to running

fastapi dev ./src/fastapi_image/main.py

Step 5: Test the code with pytest

Test scripts are located in the tests directory. To run tests, use

rye run test

Note: Simply calling the pytest cli or rye test will result in errors because the project is defined as "virtual" in rye, meaning that the project itself won't be installed as a package. We need to run python -m pytest to add the root directory to sys.path to allow the correct modules to be installed.

Step 6: Build the docker image and run in a container

We have also set up the commands to build the docker image and run it in a container:

rye run build
rye run container

The container will be available at http://127.0.0.1:8080/