Book API

Introduction

This project is about creating a REST API using Fast API as a python framework. The main goal is showing how to create a development/testing docker based environment to run the API, its unit and integration tests in a easy and confortable way.

The API is a services useful to consult a review a Books database. It also has reviews from an outside source: Good Reads.

The service support user Sign In and Log In, look for certain a book based on its ISBN, title or author. The users can insert reviews and comments about the read books. All this data is stored in a PostgreSQL data base.

Running the API

Prerequisites

The project that system has installed:

• Docker
• Docker Compose
• Makefile
• Python >= 3.7

Running with docker

The first step is to create a app/.env file, for that it is recommended to use app/.example.env as a guide.

Then, it is necessary to build the images for the project. From the project root path run (it can take several minutes):

make build

One you have the images created you can start the API by running:

make run

and in you browser access to http://127.0.0.1:8000/docs. This URL provides documentation about the different endpoints available. Also, note that the code is mounted in the docker container as volume, so it is possible to develop the application without rebuilding the images with each code change.

Finally you can stop it with simply running:

make stop

Running with docker: Development and testing

Other alternative is to run the API with development and testing purposes. In order to do that you need to run:

make build_dev
make run_dev

The test are run with

make run_tests

The previous commands run unit tests for each service and integration test for the entire API

Details about docker images

The docker-compose file, which is used indirectly through the Makefile, creates 4 images/services:

1. The API based on FAST API
2. PostgreSQL database for development
3. PostgreSQL database for integration testing
4. PgAdmin as PostgreSQL database administrator

The project has 2 database because. One of them is used to run the integration tests and it is cleaned after each test process. In that way, the test guarantee that start always with the same conditions.

The other one is used to develop the app and do some manual trials. This database use a docker volume in order to be able to persist the data.

Finally, it is possible to enter to the database administrator by accessing to http://127.0.0.1:5050 with credentials:

• user: admin@example.com
• password: password

Yes, I know, security first!

Other details...

The project used poetry as a tool to install and manage python dependencies. Poetry bases on pyproject.toml file to obtain dependencies. This file also allows to configure other tools like:

• Black
• Flakehell
• and others...

The project uses black and flakehell (with some extensions) as pre-commit hooks. In order to run this hooks it is necessary to install the python dependencies locally inside a virtual env. All the dependencies and hooks can be installed and configured easily by running (assuming you have python >= 3.7 already installed):

make start_project

Next Steps