docker-crawlab.yml

name: "Docker Image CI: crawlab"

on:
  push:
    branches: [ develop, test, main ]
  release:
    types: [ published ]
  workflow_dispatch:
  repository_dispatch:
    types: [ docker-crawlab ]

env:
  IMAGE_PATH_CRAWLAB_BACKEND: backend
  IMAGE_PATH_CRAWLAB_FRONTEND: frontend
  GH_PKG_NAME_CRAWLAB_BASE: ${{ github.repository_owner }}/crawlab-base
  GH_PKG_NAME_CRAWLAB_BACKEND: ${{ github.repository_owner }}/crawlab-backend
  GH_PKG_NAME_CRAWLAB_FRONTEND: ${{ github.repository_owner }}/crawlab-frontend
  GH_PKG_NAME_CRAWLAB: ${{ github.repository_owner }}/crawlab
  IMAGE_NAME_CRAWLAB_BASE: ghcr.io/${{ github.repository_owner }}/crawlab-base
  IMAGE_NAME_CRAWLAB_BACKEND: ghcr.io/${{ github.repository_owner }}/crawlab-backend
  IMAGE_NAME_CRAWLAB_FRONTEND: ghcr.io/${{ github.repository_owner }}/crawlab-frontend
  IMAGE_NAME_CRAWLAB_GH: ghcr.io/${{ github.repository_owner }}/crawlab
  IMAGE_NAME_CRAWLAB_DOCKERHUB: crawlabteam/crawlab
  IMAGE_NAME_CRAWLAB_TENCENT: ccr.ccs.tencentyun.com/crawlab/crawlab
  E2E_TESTS_IMAGE_NAME: ghcr.io/${{ github.repository_owner }}/e2e-tests
  E2E_TESTS_WORKSPACE: ${{ github.workspace }}/playwright-report

jobs:
  setup:
    name: Setup
    runs-on: ubuntu-latest
    outputs:
      backend_changed: ${{ steps.check_changed_files.outputs.backend_changed }}
      frontend_changed: ${{ steps.check_changed_files.outputs.frontend_changed }}
      docker_changed: ${{ steps.check_changed_files.outputs.docker_changed }}
      workflow_changed: ${{ steps.check_changed_files.outputs.workflow_changed }}
      base_image_changed: ${{ steps.check_changed_files.outputs.base_image_changed }}
      version: ${{ steps.version.outputs.version }}
      test_script: ${{ steps.test_config.outputs.test_script }}

    steps:
      - uses: actions/checkout@v4

      - name: Get changed files
        id: changed_files
        uses: tj-actions/changed-files@v45
        with:
          files_yaml: |
            backend:
              - 'backend/**'
              - 'core/**'
              - 'fs/**'
              - 'grpc/**'
              - 'vcs/**'
              - 'trace/**'
            frontend:
              - '${{ env.IMAGE_PATH_CRAWLAB_FRONTEND }}/**'
            docker:
              - 'Dockerfile'
              - 'docker/bin/**'
              - 'docker/nginx/**'
            workflow:
              - '.github/workflows/docker-crawlab.yml'
            base_image:
              - 'docker/base-image/**'

      - id: check_changed_files
        name: Check changed files
        run: |
          # set outputs
          if [[ "${{ github.ref }}" == "refs/tags/"* ]]; then
            echo "backend_changed=true" >> $GITHUB_OUTPUT
            echo "frontend_changed=true" >> $GITHUB_OUTPUT
            echo "docker_changed=true" >> $GITHUB_OUTPUT
            echo "workflow_changed=true" >> $GITHUB_OUTPUT
            echo "base_image_changed=true" >> $GITHUB_OUTPUT
          else
            echo "backend_changed=${{ steps.changed_files.outputs.backend_any_changed }}" >> $GITHUB_OUTPUT
            echo "frontend_changed=${{ steps.changed_files.outputs.frontend_any_changed }}" >> $GITHUB_OUTPUT
            echo "docker_changed=${{ steps.changed_files.outputs.docker_any_changed }}" >> $GITHUB_OUTPUT
            echo "workflow_changed=${{ steps.changed_files.outputs.workflow_any_changed }}" >> $GITHUB_OUTPUT
            echo "base_image_changed=${{ steps.changed_files.outputs.base_image_any_changed }}" >> $GITHUB_OUTPUT
          fi
          
          # Display change status for each component
          echo "Backend changed: ${{ steps.changed_files.outputs.backend_any_changed }}"
          echo "Frontend changed: ${{ steps.changed_files.outputs.frontend_any_changed }}"
          echo "Docker changed: ${{ steps.changed_files.outputs.docker_any_changed }}"
          echo "Workflow changed: ${{ steps.changed_files.outputs.workflow_any_changed }}"
          echo "Base image changed: ${{ steps.changed_files.outputs.base_image_any_changed }}"

      - id: version
        name: Get version
        run: |
          # Strip git ref prefix from version
          VERSION=$(echo "${{ github.ref }}" | sed -e 's,.*/\(.*\),\1,')
          
          # Strip "v" prefix from tag name
          [[ "${{ github.ref }}" == "refs/tags/"* ]] && VERSION=$(echo $VERSION | sed -e 's/^v//')
          
          # Use Docker `latest` tag convention
          [ "$VERSION" == "main" ] && VERSION=latest
          
          echo "version=$VERSION" >> $GITHUB_OUTPUT

      - id: test_config
        name: Set test configuration
        run: |
          if [[ "${{ github.ref }}" == "refs/heads/main" ]]; then
            echo "test_script=test:full" >> $GITHUB_OUTPUT
          elif [[ "${{ github.ref }}" == "refs/heads/test" ]]; then
            echo "test_script=test:extended" >> $GITHUB_OUTPUT
          else
            echo "test_script=test:normal" >> $GITHUB_OUTPUT
          fi

  build_base_image:
    name: Build base image
    needs: [ setup ]
    if: needs.setup.outputs.base_image_changed == 'true' || needs.setup.outputs.workflow_changed == 'true'
    runs-on: ubuntu-latest
    outputs:
      failed: ${{ steps.set_output.outputs.failed }}
    steps:
      - uses: actions/checkout@v4
      - name: Set up Docker Buildx
        uses: docker/setup-buildx-action@v3
      - name: Log in to GitHub Container Registry
        uses: docker/login-action@v3
        with:
          registry: ghcr.io
          username: ${{ github.actor }}
          password: ${{ secrets.GITHUB_TOKEN }}
      - name: Build and push image
        uses: docker/build-push-action@v5
        with:
          context: ./docker/base-image
          push: true
          tags: ${{ env.IMAGE_NAME_CRAWLAB_BASE }}:${{ needs.setup.outputs.version }}
      - name: Set output
        id: set_output
        if: failure()
        run: echo "failed=true" >> $GITHUB_OUTPUT

  test_backend:
    name: Test backend
    needs: [ setup ]
    if: needs.setup.outputs.backend_changed == 'true' || needs.setup.outputs.workflow_changed == 'true'
    runs-on: ubuntu-latest
    outputs:
      failed: ${{ steps.set_output.outputs.failed }}
    services:
      mongo:
        image: mongo:5
        ports:
          - 27017:27017
    strategy:
      matrix:
        package: [core]
    steps:
      - uses: actions/checkout@v4
      - name: Set up Go
        uses: actions/setup-go@v5
        with:
          go-version-file: '${{ matrix.package }}/go.mod'
          cache-dependency-path: '${{ matrix.package }}/go.sum'
      - name: Run tests
        working-directory: ${{ matrix.package }}
        run: |
          # Find all directories containing *_test.go files
          test_dirs=$(find . -name "*_test.go" -exec dirname {} \; | sort -u)
          # Run go test on each directory
          for dir in $test_dirs
          do
            echo "Running tests in $dir"
            go test ./$dir
          done
      - name: Set output
        id: set_output
        if: failure()
        run: echo "failed=true" >> $GITHUB_OUTPUT

  build_backend:
    name: Build backend
    needs: [ setup, test_backend ]
    if: needs.test_backend.result == 'success' || needs.setup.outputs.workflow_changed == 'true'
    runs-on: ubuntu-latest
    outputs:
      failed: ${{ steps.set_output.outputs.failed }}
    steps:
      - uses: actions/checkout@v4
      - name: Log in to GitHub Container Registry
        uses: docker/login-action@v3
        with:
          registry: ghcr.io
          username: ${{ github.actor }}
          password: ${{ secrets.GITHUB_TOKEN }}
      - name: Build and push image
        uses: docker/build-push-action@v5
        with:
          context: .
          file: ${{ env.IMAGE_PATH_CRAWLAB_BACKEND }}/Dockerfile
          push: true
          tags: ${{ env.IMAGE_NAME_CRAWLAB_BACKEND }}:${{ needs.setup.outputs.version }}
      - name: Set output
        id: set_output
        if: failure()
        run: echo "failed=true" >> $GITHUB_OUTPUT

  build_frontend:
    name: Build frontend
    needs: [ setup ]
    if: needs.setup.outputs.frontend_changed == 'true' || needs.setup.outputs.workflow_changed == 'true'
    runs-on: ubuntu-latest
    outputs:
      failed: ${{ steps.set_output.outputs.failed }}
    steps:
      - uses: actions/checkout@v4
        with:
          submodules: 'recursive'
      - name: Log in to GitHub Container Registry
        uses: docker/login-action@v3
        with:
          registry: ghcr.io
          username: ${{ github.actor }}
          password: ${{ secrets.GITHUB_TOKEN }}
      - name: Build and push image
        uses: docker/build-push-action@v5
        with:
          context: ${{ env.IMAGE_PATH_CRAWLAB_FRONTEND }}
          push: true
          tags: ${{ env.IMAGE_NAME_CRAWLAB_FRONTEND }}:${{ needs.setup.outputs.version }}
      - name: Set output
        id: set_output
        if: failure()
        run: echo "failed=true" >> $GITHUB_OUTPUT

  build_crawlab:
    name: Build crawlab
    needs: [setup, build_base_image, test_backend, build_backend, build_frontend]
    if: |
      always() &&
      (
        needs.test_backend.outputs.failed != 'true' &&
        needs.build_backend.outputs.failed != 'true' &&
        needs.build_frontend.outputs.failed != 'true' &&
        needs.build_base_image.outputs.failed != 'true'
      ) &&
      (
        needs.setup.outputs.backend_changed == 'true' ||
        needs.setup.outputs.frontend_changed == 'true' ||
        needs.setup.outputs.docker_changed == 'true' ||
        needs.setup.outputs.base_image_changed == 'true' ||
        needs.setup.outputs.workflow_changed == 'true'
      )
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - name: Update Dockerfile
        run: |
          IMAGE_NAMES=(
            "crawlab-base"
            "crawlab-backend"
            "crawlab-frontend"
          )
          for name in "${IMAGE_NAMES[@]}"; do
            IMAGE_NAME="ghcr.io/${{ github.repository_owner }}/$name"
            OLD_IMAGE="crawlabteam/${name}:"
            NEW_IMAGE="${IMAGE_NAME}:"
            sed -i "s|${OLD_IMAGE}|${NEW_IMAGE}|" Dockerfile
          done
      - name: Log in to GitHub Container Registry
        uses: docker/login-action@v3
        with:
          registry: ghcr.io
          username: ${{ github.actor }}
          password: ${{ secrets.GITHUB_TOKEN }}
      - name: Build and push image
        uses: docker/build-push-action@v5
        with:
          context: .
          file: ./Dockerfile
          build-args: |
            CRAWLAB_TAG=${{ needs.setup.outputs.version }}
          push: true
          tags: |
            ${{ env.IMAGE_NAME_CRAWLAB_GH }}:${{ needs.setup.outputs.version }}

  test_crawlab:
    name: Test crawlab
    needs: [setup, build_crawlab]
    if: ${{ always() && needs.build_crawlab.result == 'success' }}
    runs-on: ubuntu-latest
    services:
      mongo:
        image: mongo:5
        options: >-
          --health-cmd "mongosh --eval 'db.adminCommand(\"ping\")' || exit 1"
          --health-interval 10s
          --health-timeout 5s
          --health-retries 5
      master:
        image: ghcr.io/${{ github.repository_owner }}/crawlab:${{ needs.setup.outputs.version }}
        env:
          CRAWLAB_NODE_MASTER: Y
          CRAWLAB_MONGO_HOST: mongo 
          CRAWLAB_MONGO_PORT: 27017
        ports:
          - 8080:8080
      worker:
        image: ghcr.io/${{ github.repository_owner }}/crawlab:${{ needs.setup.outputs.version }}
        env:
          CRAWLAB_NODE_MASTER: N
          CRAWLAB_MASTER_HOST: master
    steps:
      - name: Checkout repository
        uses: actions/checkout@v4
      - name: Log in to GitHub Container Registry
        uses: docker/login-action@v3
        with:
          registry: ghcr.io
          username: ${{ github.actor }}
          password: ${{ secrets.GITHUB_TOKEN }}
      - name: Pull e2e test image
        run: |
          docker pull ${{ env.E2E_TESTS_IMAGE_NAME }}:latest
      - name: Run e2e test image
        run: |
          docker run --network host \
            -e BASE_URL=http://localhost:8080 \
            -e TEST_SCRIPT=${{ needs.setup.outputs.test_script }} \
            -v ${{ env.E2E_TESTS_WORKSPACE }}:/app/playwright-report \
            ${{ env.E2E_TESTS_IMAGE_NAME }}:latest
      - name: Upload test results
        if: ${{ success() || failure() }}
        uses: actions/upload-artifact@v4
        with:
          name: playwright-report-${{ needs.setup.outputs.version }}
          path: ${{ env.E2E_TESTS_WORKSPACE }}
          retention-days: 1
          overwrite: true

  push_images:
    name: Push images
    if: ${{ always() && needs.test_crawlab.result == 'success' }}
    needs: [setup, test_crawlab]
    runs-on: ubuntu-latest
    strategy:
      matrix:
        registry: [dockerhub, tencent]
    steps:
      - name: Login to GitHub Container Registry
        uses: docker/login-action@v3
        with:
          registry: ghcr.io
          username: ${{ github.actor }}
          password: ${{ secrets.GITHUB_TOKEN }}
      - name: Pull Docker image from GitHub Container Registry
        run: docker pull ${{ env.IMAGE_NAME_CRAWLAB_GH }}:${{ needs.setup.outputs.version }}
      - name: Login to DockerHub or Tencent Registry
        uses: docker/login-action@v3
        with:
          registry: ${{ (matrix.registry == 'dockerhub' && 'https://index.docker.io/v1/') || (matrix.registry == 'tencent' && 'ccr.ccs.tencentyun.com') }}
          username: ${{ (matrix.registry == 'dockerhub' && secrets.DOCKER_USERNAME) || (matrix.registry == 'tencent' && secrets.DOCKER_TENCENT_USERNAME) }}
          password: ${{ (matrix.registry == 'dockerhub' && secrets.DOCKER_PASSWORD) || (matrix.registry == 'tencent' && secrets.DOCKER_TENCENT_PASSWORD) }}
      - name: Tag and push image
        if: ${{ matrix.registry == 'dockerhub' || (matrix.registry == 'tencent' && github.ref != 'refs/heads/develop') }}
        run: |
          docker tag ${{ env.IMAGE_NAME_CRAWLAB_GH }}:${{ needs.setup.outputs.version }} ${{ (matrix.registry == 'dockerhub' && env.IMAGE_NAME_CRAWLAB_DOCKERHUB) || (matrix.registry == 'tencent' && env.IMAGE_NAME_CRAWLAB_TENCENT) }}:${{ needs.setup.outputs.version }}
          docker push ${{ (matrix.registry == 'dockerhub' && env.IMAGE_NAME_CRAWLAB_DOCKERHUB) || (matrix.registry == 'tencent' && env.IMAGE_NAME_CRAWLAB_TENCENT) }}:${{ needs.setup.outputs.version }}