airbnb/caravel

# Python Presto/Hive unit tests
name: Python Presto/Hive

on:
  push:
    branches:
      - "master"
      - "[0-9].[0-9]"
  pull_request:
    types: [synchronize, opened, reopened, ready_for_review]

# cancel previous workflow jobs for PRs
concurrency:
  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.run_id }}
  cancel-in-progress: true

jobs:
  test-postgres-presto:
    runs-on: ubuntu-20.04
    env:
      PYTHONPATH: ${{ github.workspace }}
      SUPERSET_CONFIG: tests.integration_tests.superset_test_config
      REDIS_PORT: 16379
      SUPERSET__SQLALCHEMY_DATABASE_URI: postgresql+psycopg2://superset:superset@127.0.0.1:15432/superset
      SUPERSET__SQLALCHEMY_EXAMPLES_URI: presto://localhost:15433/memory/default
    services:
      postgres:
        image: postgres:15-alpine
        env:
          POSTGRES_USER: superset
          POSTGRES_PASSWORD: superset
        ports:
          # Use custom ports for services to avoid accidentally connecting to
          # GitHub action runner's default installations
          - 15432:5432
      presto:
        image: starburstdata/presto:350-e.6
        env:
          POSTGRES_USER: superset
          POSTGRES_PASSWORD: superset
        ports:
          # Use custom ports for services to avoid accidentally connecting to
          # GitHub action runner's default installations
          - 15433:8080
      redis:
        image: redis:7-alpine
        ports:
          - 16379:6379
    steps:
      - name: "Checkout ${{ github.ref }} ( ${{ github.sha }} )"
        uses: actions/checkout@v4
        with:
          persist-credentials: false
          submodules: recursive
      - name: Check for file changes
        id: check
        uses: ./.github/actions/change-detector/
        with:
          token: ${{ secrets.GITHUB_TOKEN }}
      - name: Setup Python
        uses: ./.github/actions/setup-backend/
        if: steps.check.outputs.python == 'true'
      - name: Setup Postgres
        if: steps.check.outputs.python
        uses: ./.github/actions/cached-dependencies
        with:
          run: |
            echo "${{ steps.check.outputs.python }}"
            setup-postgres
      - name: Run celery
        if: steps.check.outputs.python
        run: celery --app=superset.tasks.celery_app:app worker -Ofair -c 2 &
      - name: Python unit tests (PostgreSQL)
        if: steps.check.outputs.python
        run: |
          ./scripts/python_tests.sh -m 'chart_data_flow or sql_json_flow'
      - name: Upload code coverage
        uses: codecov/codecov-action@v4
        with:
          flags: python,presto
          token: ${{ secrets.CODECOV_TOKEN }}
          verbose: true

  test-postgres-hive:
    runs-on: ubuntu-20.04
    env:
      PYTHONPATH: ${{ github.workspace }}
      SUPERSET_CONFIG: tests.integration_tests.superset_test_config
      REDIS_PORT: 16379
      SUPERSET__SQLALCHEMY_DATABASE_URI: postgresql+psycopg2://superset:superset@127.0.0.1:15432/superset
      SUPERSET__SQLALCHEMY_EXAMPLES_URI: hive://localhost:10000/default
      UPLOAD_FOLDER: /tmp/.superset/uploads/
    services:
      postgres:
        image: postgres:15-alpine
        env:
          POSTGRES_USER: superset
          POSTGRES_PASSWORD: superset
        ports:
          # Use custom ports for services to avoid accidentally connecting to
          # GitHub action runner's default installations
          - 15432:5432
      redis:
        image: redis:7-alpine
        ports:
          - 16379:6379
    steps:
      - name: "Checkout ${{ github.ref }} ( ${{ github.sha }} )"
        uses: actions/checkout@v4
        with:
          persist-credentials: false
          submodules: recursive
      - name: Check for file changes
        id: check
        uses: ./.github/actions/change-detector/
        with:
          token: ${{ secrets.GITHUB_TOKEN }}
      - name: Create csv upload directory
        if: steps.check.outputs.python
        run: sudo mkdir -p /tmp/.superset/uploads
      - name: Give write access to the csv upload directory
        if: steps.check.outputs.python
        run: sudo chown -R $USER:$USER /tmp/.superset
      - name: Start hadoop and hive
        if: steps.check.outputs.python
        run: docker compose -f scripts/databases/hive/docker-compose.yml up -d
      - name: Setup Python
        uses: ./.github/actions/setup-backend/
        if: steps.check.outputs.python
      - name: Setup Postgres
        if: steps.check.outputs.python
        uses: ./.github/actions/cached-dependencies
        with:
          run: setup-postgres
      - name: Run celery
        if: steps.check.outputs.python
        run: celery --app=superset.tasks.celery_app:app worker -Ofair -c 2 &
      - name: Python unit tests (PostgreSQL)
        if: steps.check.outputs.python
        run: |
          ./scripts/python_tests.sh -m 'chart_data_flow or sql_json_flow'
      - name: Upload code coverage
        uses: codecov/codecov-action@v4
        with:
          flags: python,hive
          token: ${{ secrets.CODECOV_TOKEN }}
          verbose: true