FastAPI Demo

Getting Started

GitHub Codespaces

The easiest way to start working is to open your GitHub repository in GitHub Codespaces.

From within that environment, you have a full IDE to work with the code, run commands in the terminal, preview the application by running the ./preview.sh script, and perform normal git add/commit/push commands.

Local

After forking this repository for your own work, you may need to set up an isolated environment in Python. I suggest using pipenv for this:

# Install pipenv itself
python3 -m pip install pipenv

# Then from within the root directory of this project,
# create a new virtual environment
pipenv shell

# Exit the virtual environment at any point with "exit"
# Return to it by running "pipenv shell" from the directory.

Development

As typical with FastAPI development, run the local server as you code:

# cd into the app/ directory
cd app

# run the local uvicorn server (install locally first)
uvicorn main:app --reload

Your dev site is now running locally at http://127.0.0.1:8000/

Sample Endpoints and Methods

This template contains a variety of methods and endpoints:

http://127.0.0.1:8000/add/7/3

Adds two integers taken as URL path parameters

{
  "sum": 10
}

http://127.0.0.1:8000/items/1234567890?q=foo&s=bar

Takes an integer path parameter 1234567890 as well as two query string parameters foo and bar.

{
  "item_id": 1234567890,
  "q": "foo",
  "s": "bar"
}

For more on catching POST payloads in JSON, or form parameters, consult the FastAPI documentation.

Build the Container

Build locally using the docker build command:

docker build -t some_org/some_image:some_tag .

Run the Container Locally

Run the image locally and map the container port (80) to some host port (8080):

docker run -d -p 8080:80 --rm some_org/some_image:some_tag

Set up a Build Pipeline

Look in .build.yml for a sample template that completes the following steps:

1. Builds the container image.
2. Pushes the new image to the container registry of your choice (i.e. Docker Hub, GHCR, etc.)
3. Pushes an SQS message to Amazon triggering a re-deployment of the app in DCOS.

To enable this pipeline, place it in .github/workflows/ and git push. Some repository secrets must be set before builds will work.

post template