build_apache_tika.yaml

on:
  push:
    branches:
      - main
    paths:
      - 'Dockerfile_apache_tika'
      - '.github/workflows/build_apache_tika.yaml'
    tags:
      - "v*"
  issue_comment:
    types: [created]
  workflow_dispatch:
    inputs:
      tika_version:
        description: 'Apache Tika version (e.g. 3.2.2)'
        required: false
        default: '3.2.2'

name: Build Apache Tika container image

permissions:
  contents: read
  packages: write

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

jobs:
  test-apache-tika:
    name: Test Apache Tika on multiple architectures
    if: |
      (github.event_name != 'issue_comment') ||
      (github.event.issue.pull_request != null && contains(github.event.comment.body, 'cicd/build'))
    strategy:
      fail-fast: false
      matrix:
        include:
          - platform: linux/amd64
            runner: ubuntu-latest
            arch: amd64
          - platform: linux/arm64
            runner: ubuntu-24.04-arm
            arch: arm64
    runs-on: ${{ matrix.runner }}
    timeout-minutes: 30
    steps:
      - name: Checkout code
        uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4.3.1
        with:
          ref: ${{ github.event_name == 'issue_comment' && format('refs/pull/{0}/head', github.event.issue.number) || github.ref }}

      - name: Free up disk space
        run: ./.github/scripts/free_disk_space.sh

      - name: Set up Docker Buildx
        uses: docker/setup-buildx-action@8d2750c68a42422c14e847fe6c8ac0403b4cbd6f # v3.12.0
        with:
          driver-opts: |
            image=moby/buildkit:v0.12.5

      - name: Build Apache Tika test image for ${{ matrix.platform }}
        uses: docker/build-push-action@ca052bb54ab0790a636c9b5f226502c73d547a25 # v5.4.0
        with:
          context: .
          file: ./Dockerfile_apache_tika
          platforms: ${{ matrix.platform }}
          load: ${{ matrix.platform == 'linux/amd64' }}
          cache-from: type=gha,scope=tika-test-${{ matrix.arch }}
          cache-to: type=gha,mode=min,scope=tika-test-${{ matrix.arch }},ignore-error=true
          build-args: |
            TIKA_VERSION=${{ inputs.tika_version || '3.2.2' }}
          tags: |
            test-apache-tika:${{ matrix.arch }}

      - name: Test Apache Tika server
        run: |
          # Start Tika server
          CONTAINER_ID=$(docker run -d -p 9998:9998 test-apache-tika:${{ matrix.arch }})
          echo "Waiting for Tika server to start on ${{ matrix.platform }}..."
          sleep 15

          # Test if Tika is responding (with retry)
          TIKA_READY=false
          for i in {1..6}; do
            if curl -f http://localhost:9998/version > /dev/null 2>&1; then
              TIKA_READY=true
              break
            fi
            echo "Attempt $i/6: Tika not ready yet, waiting 5 more seconds..."
            sleep 5
          done

          if [ "$TIKA_READY" = true ]; then
            echo "✅ Apache Tika server is responding on ${{ matrix.platform }}"
            TIKA_VERSION=$(curl -s http://localhost:9998/version)
            echo "✅ Tika version: $TIKA_VERSION"
          else
            echo "❌ Apache Tika server is not responding after 45 seconds on ${{ matrix.platform }}"
            echo "Container logs:"
            docker logs $CONTAINER_ID
            docker stop $CONTAINER_ID
            exit 1
          fi

          # Cleanup
          docker stop $CONTAINER_ID

      - name: Clean up test images
        if: always()
        run: |
          docker system prune -f
          docker image prune -af

  build-apache-tika:
    name: Build Apache Tika container image
    strategy:
      fail-fast: false
      matrix:
        include:
          - platform: linux/amd64
            runner: ubuntu-latest
            arch: amd64
          - platform: linux/arm64
            runner: ubuntu-24.04-arm
            arch: arm64
    runs-on: ${{ matrix.runner }}
    needs: [test-apache-tika]
    timeout-minutes: 30
    steps:
      - name: Checkout code
        uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4.3.1
        with:
          ref: ${{ github.event_name == 'issue_comment' && format('refs/pull/{0}/head', github.event.issue.number) || github.ref }}

      - name: Free up disk space
        run: ./.github/scripts/free_disk_space.sh

      - name: Set up Docker Buildx
        uses: docker/setup-buildx-action@8d2750c68a42422c14e847fe6c8ac0403b4cbd6f # v3.12.0
        with:
          driver-opts: |
            image=moby/buildkit:v0.12.5

      - name: Login to GitHub Container Registry
        uses: docker/login-action@c94ce9fb468520275223c153574b00df6fe4bcc9 # v3.7.0
        with:
          registry: ghcr.io
          username: ${{ github.repository_owner }}
          password: ${{ secrets.GITHUB_TOKEN }}

      - name: Generate build metadata
        id: meta
        run: |
          echo "build_date=$(date -u +"%Y-%m-%dT%H:%M:%SZ")" >> $GITHUB_OUTPUT
          echo "vcs_ref=$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
          echo "version=$(git describe --tags --always 2>/dev/null || echo 'latest')" >> $GITHUB_OUTPUT

      - name: Build and push Apache Tika development container image
        if: ${{ startsWith(github.ref, 'refs/heads/') || github.event_name == 'issue_comment' }}
        uses: docker/build-push-action@ca052bb54ab0790a636c9b5f226502c73d547a25 # v5.4.0
        with:
          context: .
          file: ./Dockerfile_apache_tika
          platforms: ${{ matrix.platform }}
          push: true
          cache-from: type=gha,scope=tika-main-${{ matrix.arch }}
          cache-to: type=gha,mode=max,scope=tika-main-${{ matrix.arch }},ignore-error=true
          build-args: |
            TIKA_VERSION=${{ inputs.tika_version || '3.2.2' }}
            BUILD_DATE=${{ steps.meta.outputs.build_date }}
            VCS_REF=${{ steps.meta.outputs.vcs_ref }}
          tags: |
            ghcr.io/${{ github.repository_owner }}/querido-diario-apache-tika-server:latest-${{ matrix.arch }}

      - name: Build and push Apache Tika tagged container image
        if: ${{ startsWith(github.ref, 'refs/tags/') }}
        uses: docker/build-push-action@ca052bb54ab0790a636c9b5f226502c73d547a25 # v5.4.0
        with:
          context: .
          file: ./Dockerfile_apache_tika
          platforms: ${{ matrix.platform }}
          push: true
          cache-from: type=gha,scope=tika-tag-${{ matrix.arch }}
          cache-to: type=gha,mode=max,scope=tika-tag-${{ matrix.arch }},ignore-error=true
          build-args: |
            TIKA_VERSION=${{ inputs.tika_version || '3.2.2' }}
            BUILD_DATE=${{ steps.meta.outputs.build_date }}
            VCS_REF=${{ steps.meta.outputs.vcs_ref }}
          tags: |
            ghcr.io/${{ github.repository_owner }}/querido-diario-apache-tika-server:${{ github.ref_name }}-${{ matrix.arch }}

  create-apache-tika-manifest:
    name: Create Apache Tika multi-arch manifest
    runs-on: ubuntu-latest
    needs: build-apache-tika
    timeout-minutes: 15
    steps:
      - name: Set up Docker Buildx
        uses: docker/setup-buildx-action@8d2750c68a42422c14e847fe6c8ac0403b4cbd6f # v3.12.0

      - name: Login to GitHub Container Registry
        uses: docker/login-action@c94ce9fb468520275223c153574b00df6fe4bcc9 # v3.7.0
        with:
          registry: ghcr.io
          username: ${{ github.repository_owner }}
          password: ${{ secrets.GITHUB_TOKEN }}

      - name: Verify single-arch images availability (branch/PR)
        if: ${{ startsWith(github.ref, 'refs/heads/') || github.event_name == 'issue_comment' }}
        run: |
          IMAGE="ghcr.io/${{ github.repository_owner }}/querido-diario-apache-tika-server"
          for tag in latest-amd64 latest-arm64; do
            for i in {1..20}; do
              if docker buildx imagetools inspect $IMAGE:$tag > /dev/null 2>&1; then
                echo "Found $IMAGE:$tag";
                break;
              fi
              echo "Waiting for $IMAGE:$tag to be available ($i/20)...";
              sleep 3;
            done
          done

      - name: Create and push Apache Tika development manifest
        if: ${{ startsWith(github.ref, 'refs/heads/') || github.event_name == 'issue_comment' }}
        run: |
          IMAGE="ghcr.io/${{ github.repository_owner }}/querido-diario-apache-tika-server"
          docker buildx imagetools create \
            -t $IMAGE:latest \
            $IMAGE:latest-amd64 \
            $IMAGE:latest-arm64

      - name: Verify single-arch images availability (tag)
        if: ${{ startsWith(github.ref, 'refs/tags/') }}
        run: |
          IMAGE="ghcr.io/${{ github.repository_owner }}/querido-diario-apache-tika-server"
          for arch in amd64 arm64; do
            for i in {1..20}; do
              if docker buildx imagetools inspect $IMAGE:${{ github.ref_name }}-$arch > /dev/null 2>&1; then
                echo "Found $IMAGE:${{ github.ref_name }}-$arch";
                break;
              fi
              echo "Waiting for $IMAGE:${{ github.ref_name }}-$arch to be available ($i/20)...";
              sleep 3;
            done
          done

      - name: Create and push Apache Tika tagged manifest
        if: ${{ startsWith(github.ref, 'refs/tags/') }}
        run: |
          IMAGE="ghcr.io/${{ github.repository_owner }}/querido-diario-apache-tika-server"
          docker buildx imagetools create \
            -t $IMAGE:${{ github.ref_name }} \
            $IMAGE:${{ github.ref_name }}-amd64 \
            $IMAGE:${{ github.ref_name }}-arm64