Records

The idea of this project is to show how Starlette, Marshmallow, and SQLAlchemy can be combined to create a RESTful HTTP API application that is modular, lightweight, and capable of dealing with many simultaneous requests.

Architecture

Uses a 3-tier architecture, illustrated below.

Setting up

Requires Python 3.9+, git, and poetry.

Building

$ git clone https://github.com/rbw/records.git
$ cd records
$ poetry update

Starting Postgres

The application requires an SQLAlchemy-supported relational database. This example uses pg.ARRAY in AlbumModel, making it compatible with Postgres only.

A docker-compose file for running a Postgres server is available in the project root.

$ docker-compose up

Starting Records

$ poetry shell
$ python -m records

Usage

Get all albums

$ curl http://localhost:5000/albums

Create an album

$ curl -X POST --data '{"title": "test", "release_date": "2035-01-20", "stores": ["APPLE", "YOUTUBE"], "tracks": ["TEST000000001", "TEST000000002"], "upc": "00000000000005"}' http://localhost:5000/albums

Show an album

$ curl http://localhost:5000/albums/00000000000005

Todo

• Add Tracks API [GET/POST/DELETE => /tracks,/tracks:trackid:]
• Implement add-tracks-to-album [POST => /albums/:albumid:/tracks]
• Implement del-track-from-album [DELETE => /albums/:albumid:/tracks/:trackid:]
• Implement query filtering
• Add CLI
• Project docs
• Code docs
• Automated tests

Funding

Consider leaving a donation if you like this software.

Author

Robert Wikman <[email protected]>