unit-tests.yml

name: ListenBrainz Server Tests

on:
  push:
    branches: [ master ]
  pull_request:
    branches: [ '*' ]
    types: [ opened, synchronize, reopened, ready_for_review ]
    paths-ignore:
      - 'listenbrainz_spark/**'
      - 'requirements_spark.txt'
      - '**/Dockerfile.spark*'
      - '**.jsx?'
      - '**.tsx?'
      - '**.snap'
      - 'docker/docker-compose.spark*.yml'
      - 'docker/prod/**'
      - 'docker/services/**'
      - 'docker/spark-cluster-config/**'
      - 'docs/**'
      - '**.less'
      - '**.md'

concurrency:
    group: ${{ github.workflow }}-${{ github.ref }}
    cancel-in-progress: true

jobs:

  test:

    runs-on: ubuntu-latest

    if: github.event.pull_request.draft == false

    steps:
    - uses: actions/checkout@v3

    - name: Create configuration file
      run: cp listenbrainz/config.py.sample listenbrainz/config.py

    - name: Login to Docker Hub
      run: echo ${{ secrets.DOCKER_HUB_PASSWORD }} | docker login -u ${{ secrets.DOCKER_HUB_USERNAME }} --password-stdin
      continue-on-error: true

    # by default, docker-compose tries to pull the `listenbrainz` image used by timescale_writer as well.
    # but this is built locally for the listenbrainz service and reused for timescale writer, hence the pull fails.
    # to avoid this we specifically mention the services to pull images for here,
    - name: Pull docker images
      run: docker-compose -f docker/docker-compose.test.yml pull lb_db redis rabbitmq couchdb

    - name: Run tests
      run: ./test.sh