Example FastAPI Application

This WIP repository is an example application that uses what I consider good practices for developer experience, testing, and deployment. Best practices is to automate as much as possible, from pre-commit hooks, to automatic running of tests on every commit. Other good practices that cannot be shown in this repository easily is the presence of at least a production and staging environment, so there is a place to test and check changes before they go live.

FastAPI is a great way to build apis that keeps code efficient. It integrates use of pydantic for object models, makes definition of routes intuitive, and creates documentation out of the box.

This application includes endpoints for sensor registration and retrieval, which will save data to a postgres database.

The repository uses alembic to track changes in the database, and easily switch between versions.

TODO:

1. Tests could be configured to run on Github Actions with every commit.
2. Use certs and https in nginx. Ideally urls publicly available should use https.
3. Add user authentication. Ideally user authentication would be integrated into the middleware, so that it runs on every call. There would be a login endpoint, and an endpoint that provides a temporary token that would be stored in the Authorization header.

How to use locally

This application uses docker to deploy the application, and you will need a Docker account. Please ensure the prerequisites below are implemented before running the program.

To run the program, please run at command line,

docker-compose up -d --build

To shut down the application, run

docker-compose down

To interact with the application, you can use Postman to construct calls to the API with a GUI. The host is http://locahost, and the route is /sensors. For example, a call for all sensors would be GET http://localhost/sensors

This can also be done via command line with curl.

To add a sensor,

curl -X POST http://localhost/sensors -H 'Content-Type: application/json' -d '{"device_id":"device_id1"}'

To get one sensor with id 1,

curl http://localhost/sensors/1

To get all sensors,

curl http://localhost/sensors

Prerequisites

1. Install Docker and Docker Compose
2. Log into your docker account
3. copy .env-sample to .env, and feel free to change the password - this sets environment variables for the database

Development

Prerequisites

if you will develop the repo, a prerequisite is running pre-commit. pre-commit is already included as a dependency, so if all packages are installed, you just need to run at the command line

pre-commit install

to run the configured pre-commit hooks before a committing.

The hooks that are run are

• built in hooks to remove extra trailing white space, and an extra line return as needed at the end of a file
• isort sorts the imports
• black is an opinionated code formatter that will automatically edit your files. includes a default width of 88 characters
• flake8 is also a code formatter, that checks for errors, and complexity, asking you to fix the errors.

You'll need to have all hooks pass before the commit can happen. Pre-commit usually runs only on files added in the commit.

Configuring Pycharm

This project uses poetry for package management. You can select the Python interpreter to be a poetry interpreter.

To ensure the tests run correctly and are visible to Pycharm, you may need to add the application folder and test folders to PYTHONPATH. You can do this project-wide by going into the Settings, navigating to the Python interpreter. You'll select from a menu or drop down "Show all." Then click on the directory icon to see the paths the interpreter will search through. Manually add your application and test folders.