e2e-tests.yml

name: E2E Tests

on:
  workflow_dispatch:
  push:
    branches:
      - main
  pull_request:
    paths:
    - '.github/workflows/e2e-tests.yml'
    - 'build/**'
    - 'includes/**'
    - 'src/**'
    - 'test/e2e/**'
    - '*.php'
    - 'package.*'

# Cancels all previous workflow runs for pull requests that have not completed.
concurrency:
  # The concurrency group contains the workflow name and the branch name for pull requests
  # or the commit hash for any other events.
  group: ${{ github.workflow }}-${{ github.event_name == 'pull_request' && github.head_ref || github.sha }}
  cancel-in-progress: true


jobs:
  e2e-tests:
    runs-on: ubuntu-latest
    strategy:
      fail-fast: false
      max-parallel: 1 # Prevent parallel runs to make use of the caching for node_modules and playwright browsers
      matrix:
          php: [
            '7.4',
            '8.3'
          ]
          wp: [
            '6.4',
            '6.5',
            '6.6',
            'latest'
          ]

    steps:
    - name: Checkout
      uses: actions/checkout@v4

    - uses: actions/cache@v4
      id: playwright-cache
      with:
        path: |
          ~/.cache/ms-playwright
        key: ${{ runner.os }}-playwright-${{ hashFiles('**/package-lock.json') }}

    - name: Setup Node
      uses: actions/setup-node@v4
      with:
        node-version-file: '.nvmrc'
        # Enable built-in functionality for caching and restoring dependencies, which is disabled by default.
        # The actions/setup-node uses actions/cache under the hood.
        # https://github.com/actions/setup-node#caching-global-packages-data
        cache: 'npm'

    # Restoring the short lived node_modules cache
    # to be used across all workflows running on the last commit.
    # https://github.com/actions/cache/blob/main/caching-strategies.md#creating-a-short-lived-cache
    - uses: actions/cache/restore@v4
      id: node_modules-cache
      with:
        path: |
          ./node_modules
        key: ${{ runner.os }}-node_modules-${{ github.sha }}-${{ hashFiles('package-lock.json') }}

    - name: NPM install
      if: steps.node_modules-cache.outputs.cache-hit != 'true'
      run: npm ci --legacy-peer-deps

    # Creating a short lived node_modules cache
    - uses: actions/cache/save@v4
      if: steps.node_modules-cache.outputs.cache-hit != 'true'
      with:
        path: |
          ./node_modules
        key: ${{ steps.node_modules-cache.outputs.cache-primary-key }}

    - name: Install Playwright dependencies
      run: npx playwright install --with-deps
      if: steps.playwright-cache.outputs.cache-hit != 'true'
    - run: npx playwright install-deps
      if: steps.playwright-cache.outputs.cache-hit == 'true'

    - name: Prepare preferredVersions for blueprint
      # playground/cli ignores --php & --wp flags when --blueprint is set
      #
      # I created an issue at:
      # https://github.com/WordPress/playground-tools/issues/352
      run: |
        preferredVersions='{
          "php": "${{ matrix.php }}",
          "wp": "${{ matrix.wp }}"
        }'
        echo "Use jq to prepend the preferred versions to the existing blueprint JSON file."
        jq --argjson preferredVersions "$preferredVersions" '{preferredVersions: $preferredVersions} + .' ./test/e2e/blueprint.json > temp.blueprint.json

    - name: Starting Playground, staring Playwright & running the tests
      # This is kind of a hack PART 1/2,
      # to make sure Playwright DOES NOT start the webserver on its own.
      #
      # Part 2/2 is the "command: undefined," declaration
      # in test/e2e/playwright.config.ts
      #
      # While auto-loading the webserver when needed sounded nice, it introduced a race-condition
      # between the setup of Playground and Playwrights own start event.
      # Playwright listens for the availability of the webserver relatively simple,
      # as soon as there is a status code 200, Playwright starts all engines.
      #
      # Unfortunately Playground is not ready at this point, it hast started WordPress
      # and is going to start stepping through the blueprint, but hasn't loaded GatherPress nor imported any data;
      # Resulting in failing tests.
      #
      # It was not possible (for me) to keep the setup of Playground in a separate step,
      # why this "run > sleep > run" became necessary.
      # The setup process usually takes about 20sec, so 60 is just a good extra, to not run into errors.
      #
      # The sleep step, should help making sure Playwright runs only, when this is DONE & READY.
      #
      # Because it seems to be a problem to "wait on webServer.command" https://github.com/microsoft/playwright/issues/11811
      # & "it seems that globalSetup runs before webServer is started." https://github.com/microsoft/playwright/issues/11811#issuecomment-1040732201

      run: |
        npm run playground:mount -- --blueprint=./temp.blueprint.json & \
        sleep 60 && \
        echo 'Playground is ready now, lets run some end-to-end tests.' && \
        # DEBUG=pw:api,pw:webserver \
        xvfb-run --auto-servernum --server-args="-screen 0 1280x960x24" -- \
        npm run test:e2e


    - name: Archive debug artifacts (screenshots, traces)
      uses: actions/upload-artifact@v4
      if: ${{ ! cancelled() }}
      with:
        name: failures-artifacts--${{ matrix.php }}
        path: artifacts/test-results/
        if-no-files-found: ignore


    - name: Archive flaky tests report
      uses: actions/upload-artifact@v4
      if: ${{ ! cancelled() }}
      with:
          name: flaky-tests-report--${{ matrix.php }}
          path: flaky-tests
          if-no-files-found: ignore


  # merge-artifacts:
  #   name: Merge Artifacts
  #   if: ${{ ! cancelled() }}
  #   needs: [e2e-tests]
  #   runs-on: ubuntu-latest
  #   outputs:
  #     has-flaky-test-report: ${{ !!steps.merge-flaky-tests-reports.outputs.artifact-id }}
  #   steps:
  #     - name: Merge failures artifacts
  #       uses: actions/upload-artifact/merge@v4
  #       # Don't fail the job if there aren't any artifacts to merge.
  #       continue-on-error: true
  #       with:
  #         name: failures-artifacts
  #         # Retain the merged artifacts in case of a rerun.
  #         pattern: failures-artifacts*
  #         delete-merged: true

  #     - name: Merge flaky tests reports
  #       id: merge-flaky-tests-reports
  #       uses: actions/upload-artifact/merge@v4
  #       continue-on-error: true
  #       with:
  #         name: flaky-tests-report
  #         pattern: flaky-tests-report*
  #         delete-merged: true