This is a python 3.7 prod-ready (almost) app backend. It supports http operations on User, Song, Notification.
Formatting is done with the Intellij default formatter (should be PEP8)
To run, ensure you have a running Postgres database.
Run the api in dev mode with python3 app.py
- Run
pip install -r requirements.txt - Create a database before running migrations
We use a postgres database.
By default, migrations are run on the jellynote database (cf yoyo.ini for tables and credentials)
To initialize it, run yoyo apply -b
Unit/Intergation tests are available in the tests package.
To run them, you need an available DB and then:
~/dev/jellynote-backend > python3 -m pytest .
To build the container, run
docker build . -t jellynote-api:latest
To boot the system, run
docker-compose up
It starts Postgres + the latest version of the backend. The api should be available on localhost:8000
TODO
dataclass-json does not support yet serde for enums, the commit to handle it have been done a few days ago (26 May) so there will be a few boilerplate lines waiting for the last version.
Technical:
- project setup
- dataclasses
- serializers
- persistence (users, songs)
- persistence (notification)
- db migrations
- configuration
- routes (users, songs)
- routes notifications
- tests (users, songs)
- tests notifications
- gen notifications
- prod run configuration with gunicorn
- docker compose
Functional:
- create, update, delete users
- create, update, delete songs
- docker image + docker compose
- on song creation, generate notifications
- access notifications
- scenario data + script
Start the
Create a few users :
for usr in $(ls examples/user*) ; do curl -X POST http://localhost:8000/users -H 'Content-Type: application/json' -d @$usr ; done
Post a few songs :
for song in $(ls examples/song*) ; do curl -X POST http://localhost:8000/songs -H 'Content-Type: application/json' -d @$song ; done
Check for notifications :
curl -X GET localhost:8000/notifications/<userid>