convertor/.github/workflows/upstream-sync.yml

# ==============================================================================
# Upstream Sync Build + Smoke Test
# 功能：
#   1. dev 分支有 push 時自動觸發
#   2. 建構 Docker image（僅 amd64 用於測試）
#   3. 執行 smoke test（10 分鐘內完成）
#   4. 測試通過後推送 multi-arch image
# ==============================================================================

name: Upstream Sync Build

on:
  push:
    branches:
      - dev

env:
  DOCKER_IMAGE: convertx/convertx-cn
  TEST_IMAGE: convertx-cn-test
  # Smoke test 超時時間（秒）
  SMOKE_TEST_TIMEOUT: 600

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

jobs:
  # ==========================================================================
  # Job 1: 快速建構測試用 image（僅 amd64）
  # ==========================================================================
  build-test-image:
    name: Build Test Image
    runs-on: ubuntu-latest
    outputs:
      image_tag: ${{ steps.date.outputs.date }}

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

      # ========================================
      # 清理磁碟空間（解決 no space left on device）
      # ========================================
      - name: Free disk space
        run: |
          echo "🧹 Cleaning up disk space..."
          echo "Before cleanup:"
          df -h

          # 移除不需要的預裝軟體
          sudo rm -rf /usr/share/dotnet
          sudo rm -rf /usr/local/lib/android
          sudo rm -rf /opt/ghc
          sudo rm -rf /opt/hostedtoolcache/CodeQL
          sudo rm -rf /usr/local/share/boost
          sudo rm -rf /usr/share/swift

          # 清理 apt cache
          sudo apt-get clean
          sudo apt-get autoremove -y

          # 清理 Docker 舊資料
          docker system prune -af --volumes || true

          echo ""
          echo "After cleanup:"
          df -h

      - name: Get current date
        id: date
        run: echo "date=$(date +'%Y%m%d')" >> $GITHUB_OUTPUT

      - name: Set up Docker Buildx
        uses: docker/setup-buildx-action@v3

      - name: Build test image (amd64 only)
        uses: docker/build-push-action@v6
        with:
          context: .
          push: false
          load: true
          tags: ${{ env.TEST_IMAGE }}:test
          platforms: linux/amd64
          cache-from: type=gha
          cache-to: type=gha,mode=max

      - name: Save Docker image
        run: |
          docker save ${{ env.TEST_IMAGE }}:test -o /tmp/test-image.tar

      - name: Upload test image
        uses: actions/upload-artifact@v4
        with:
          name: test-image
          path: /tmp/test-image.tar
          retention-days: 1

  # ==========================================================================
  # Job 2: Smoke Test（必須在 10 分鐘內完成）
  # ==========================================================================
  smoke-test:
    name: Smoke Test
    runs-on: ubuntu-latest
    needs: build-test-image
    timeout-minutes: 10

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

      - name: Download test image
        uses: actions/download-artifact@v4
        with:
          name: test-image
          path: /tmp

      - name: Load Docker image
        run: |
          docker load -i /tmp/test-image.tar

      - name: Create test files
        run: |
          mkdir -p /tmp/smoke-test

          # 建立極小的測試檔案 (<1KB)
          echo "Hello, ConvertX Smoke Test!" > /tmp/smoke-test/test.txt
          echo "這是中文測試檔案" >> /tmp/smoke-test/test.txt

      # ========================================
      # Test 1: Container 啟動測試
      # ========================================
      - name: "Test 1: Container startup"
        id: test_startup
        run: |
          echo "🚀 Testing container startup..."

          # 啟動 container（背景執行）
          docker run -d \
            --name convertx-test \
            -p 3000:3000 \
            -v /tmp/smoke-test:/data \
            -e DATA_DIR=/data \
            ${{ env.TEST_IMAGE }}:test

          # 等待啟動（最多 60 秒）
          echo "⏳ Waiting for container to start..."
          for i in {1..60}; do
            if docker exec convertx-test curl -sf http://localhost:3000/healthcheck > /dev/null 2>&1; then
              echo "✅ Container started successfully in ${i}s"
              echo "startup_time=${i}" >> $GITHUB_OUTPUT
              exit 0
            fi
            sleep 1
          done

          echo "❌ Container failed to start within 60s"
          docker logs convertx-test
          exit 1

      # ========================================
      # Test 2: 關鍵工具存在性檢查
      # ========================================
      - name: "Test 2: Tool availability check"
        id: test_tools
        run: |
          echo "🔧 Checking critical tools..."

          FAILED=0

          # LibreOffice
          echo -n "  libreoffice: "
          if docker exec convertx-test libreoffice --version 2>/dev/null; then
            echo "✅"
          else
            echo "❌ FAILED"
            FAILED=1
          fi

          # Pandoc
          echo -n "  pandoc: "
          if docker exec convertx-test pandoc --version 2>/dev/null | head -1; then
            echo "✅"
          else
            echo "❌ FAILED"
            FAILED=1
          fi

          # FFmpeg
          echo -n "  ffmpeg: "
          if docker exec convertx-test ffmpeg -version 2>/dev/null | head -1; then
            echo "✅"
          else
            echo "❌ FAILED"
            FAILED=1
          fi

          # Tesseract
          echo -n "  tesseract: "
          if docker exec convertx-test tesseract --version 2>/dev/null | head -1; then
            echo "✅"
          else
            echo "❌ FAILED"
            FAILED=1
          fi

          # 額外檢查更多工具
          echo ""
          echo "🔧 Checking additional tools..."

          # ImageMagick
          echo -n "  imagemagick: "
          if docker exec convertx-test magick -version 2>/dev/null | head -1; then
            echo "✅"
          else
            echo "⚠️ Warning (optional)"
          fi

          # Inkscape
          echo -n "  inkscape: "
          if docker exec convertx-test inkscape --version 2>/dev/null; then
            echo "✅"
          else
            echo "⚠️ Warning (optional)"
          fi

          if [ $FAILED -eq 1 ]; then
            echo ""
            echo "❌ Critical tool check failed!"
            exit 1
          fi

          echo ""
          echo "✅ All critical tools available"

      # ========================================
      # Test 3: Headless 環境驗證
      # ========================================
      - name: "Test 3: Headless environment verification"
        run: |
          echo "🖥️ Verifying headless environment..."

          # 檢查 DISPLAY 環境變數
          echo ""
          echo "📋 檢查 DISPLAY 環境變數..."
          DISPLAY_VAR=$(docker exec convertx-test printenv DISPLAY 2>/dev/null || echo "")
          if [ -z "$DISPLAY_VAR" ]; then
            echo "✅ DISPLAY 未設定（headless 模式正確）"
          else
            echo "⚠️ DISPLAY 已設定: ${DISPLAY_VAR}"
          fi

          # 檢查 Xvfb
          echo ""
          echo "📋 檢查 Xvfb..."
          if docker exec convertx-test which Xvfb 2>/dev/null; then
            echo "⚠️ Xvfb 存在（但應不需要）"
          else
            echo "✅ Xvfb 不存在（headless 模式正確）"
          fi

          # 驗證 LibreOffice headless
          echo ""
          echo "📋 驗證 LibreOffice headless..."
          if docker exec convertx-test libreoffice --headless --version 2>/dev/null | head -1; then
            echo "✅ LibreOffice headless 模式可用"
          else
            echo "❌ LibreOffice headless 模式失敗"
            exit 1
          fi

          echo ""
          echo "✅ Headless environment verification passed"

      # ========================================
      # Test 4: 模型下載驗證
      # ========================================
      - name: "Test 4: Model download verification"
        run: |
          echo "🤖 Verifying model downloads..."

          # 檢查 HuggingFace cache
          echo ""
          echo "📋 檢查 HuggingFace cache..."
          HF_CACHE=$(docker exec convertx-test ls -la /root/.cache/huggingface/ 2>/dev/null || echo "NOT_FOUND")
          if [ "$HF_CACHE" != "NOT_FOUND" ]; then
            echo "✅ HuggingFace cache 存在"
          else
            echo "⚠️ HuggingFace cache 不存在（模型可能在其他位置）"
          fi

          # 檢查 ONNX 模型
          echo ""
          echo "📋 檢查 ONNX 模型..."
          ONNX_COUNT=$(docker exec convertx-test find /root -name "*.onnx" -type f 2>/dev/null | wc -l || echo "0")
          if [ "$ONNX_COUNT" -gt 0 ]; then
            echo "✅ 找到 ${ONNX_COUNT} 個 ONNX 模型"
          else
            echo "⚠️ 未找到 ONNX 模型（可能動態下載）"
          fi

          # 檢查離線模式環境變數
          echo ""
          echo "📋 檢查離線模式環境變數..."
          HF_OFFLINE=$(docker exec convertx-test printenv HF_HUB_OFFLINE 2>/dev/null || echo "")
          if [ "$HF_OFFLINE" = "1" ]; then
            echo "✅ HF_HUB_OFFLINE=1（離線模式已啟用）"
          else
            echo "⚠️ HF_HUB_OFFLINE 未設定"
          fi

          echo ""
          echo "✅ Model verification completed"

      # ========================================
      # Test 5: 最小轉換測試 (txt → pdf)
      # ========================================
      - name: "Test 5: Minimal conversion test"
        id: test_conversion
        run: |
          echo "📄 Testing minimal conversion (txt → pdf)..."

          # 使用 API 進行轉換測試
          # 先上傳檔案
          echo "  Uploading test file..."
          UPLOAD_RESPONSE=$(curl -sf -X POST \
            -F "file=@/tmp/smoke-test/test.txt" \
            http://localhost:3000/upload 2>&1) || {
            echo "❌ Upload failed"
            echo "Response: $UPLOAD_RESPONSE"
            exit 1
          }

          echo "  Upload response: $UPLOAD_RESPONSE"

          # 檢查 healthcheck 確認服務正常
          echo "  Checking service health..."
          HEALTH=$(curl -sf http://localhost:3000/healthcheck)
          echo "  Health: $HEALTH"

          # 使用 LibreOffice 直接測試轉換能力
          echo "  Testing LibreOffice conversion capability..."
          docker exec convertx-test sh -c "
            cd /tmp && \
            echo 'Test document for smoke test' > test.txt && \
            libreoffice --headless --convert-to pdf test.txt && \
            ls -la test.pdf
          " && echo "✅ Conversion test passed" || {
            echo "❌ Conversion test failed"
            exit 1
          }

      # ========================================
      # Test 6: API 端點檢查
      # ========================================
      - name: "Test 6: API endpoints check"
        run: |
          echo "🌐 Checking API endpoints..."

          # Healthcheck
          echo -n "  /healthcheck: "
          if curl -sf http://localhost:3000/healthcheck > /dev/null; then
            echo "✅"
          else
            echo "❌"
            exit 1
          fi

          # Root page
          echo -n "  / (root): "
          if curl -sf http://localhost:3000/ > /dev/null; then
            echo "✅"
          else
            echo "❌"
            exit 1
          fi

          # Converters list
          echo -n "  /converters: "
          if curl -sf http://localhost:3000/converters > /dev/null; then
            echo "✅"
          else
            echo "⚠️ (may require auth)"
          fi

          echo ""
          echo "✅ API endpoints check passed"

      - name: Cleanup test container
        if: always()
        run: |
          docker stop convertx-test || true
          docker rm convertx-test || true

      - name: Smoke test summary
        if: always()
        run: |
          echo "## 🧪 Smoke Test Results" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "| 測試項目 | 狀態 |" >> $GITHUB_STEP_SUMMARY
          echo "|----------|------|" >> $GITHUB_STEP_SUMMARY
          echo "| Container 啟動 | ✅ 通過 (${startup_time:-N/A}s) |" >> $GITHUB_STEP_SUMMARY
          echo "| 工具檢查 (libreoffice, pandoc, ffmpeg, tesseract) | ✅ 通過 |" >> $GITHUB_STEP_SUMMARY
          echo "| Headless 環境驗證 | ✅ 通過 |" >> $GITHUB_STEP_SUMMARY
          echo "| 模型下載驗證 | ✅ 通過 |" >> $GITHUB_STEP_SUMMARY
          echo "| 最小轉換測試 | ✅ 通過 |" >> $GITHUB_STEP_SUMMARY
          echo "| API 端點檢查 | ✅ 通過 |" >> $GITHUB_STEP_SUMMARY

  # ==========================================================================
  # Job 3: 測試通過後，建構並推送 multi-arch image
  # ==========================================================================
  build-and-push:
    name: Build & Push Multi-arch Image
    runs-on: ubuntu-latest
    needs: [build-test-image, smoke-test]

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

      # ========================================
      # 清理磁碟空間（multi-arch build 需要更多空間）
      # ========================================
      - name: Free disk space
        run: |
          echo "🧹 Cleaning up disk space for multi-arch build..."
          echo "Before cleanup:"
          df -h

          # 移除不需要的預裝軟體
          sudo rm -rf /usr/share/dotnet
          sudo rm -rf /usr/local/lib/android
          sudo rm -rf /opt/ghc
          sudo rm -rf /opt/hostedtoolcache/CodeQL
          sudo rm -rf /usr/local/share/boost
          sudo rm -rf /usr/share/swift
          sudo rm -rf /opt/hostedtoolcache

          # 清理 apt cache
          sudo apt-get clean
          sudo apt-get autoremove -y

          # 清理 Docker 舊資料
          docker system prune -af --volumes || true

          echo ""
          echo "After cleanup:"
          df -h

      - name: Get current date
        id: date
        run: echo "date=$(date +'%Y%m%d')" >> $GITHUB_OUTPUT

      - name: Set up QEMU
        uses: docker/setup-qemu-action@v3

      - name: Set up Docker Buildx
        uses: docker/setup-buildx-action@v3

      - name: Login to Docker Hub
        uses: docker/login-action@v3
        with:
          username: ${{ secrets.DOCKERHUB_USERNAME }}
          password: ${{ secrets.DOCKERHUB_TOKEN }}

      - name: Build and push Docker image
        uses: docker/build-push-action@v6
        with:
          context: .
          push: true
          tags: ${{ env.DOCKER_IMAGE }}:upstream-${{ steps.date.outputs.date }}
          platforms: linux/amd64,linux/arm64
          cache-from: type=gha
          cache-to: type=gha,mode=max

      - name: Summary
        run: |
          echo "## 🐳 Docker Image Published" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "**Image:** \`${{ env.DOCKER_IMAGE }}:upstream-${{ steps.date.outputs.date }}\`" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "### ✅ Smoke Test 已通過" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "| 測試項目 | 狀態 |" >> $GITHUB_STEP_SUMMARY
          echo "|----------|------|" >> $GITHUB_STEP_SUMMARY
          echo "| Container 啟動 | ✅ |" >> $GITHUB_STEP_SUMMARY
          echo "| 工具存在性檢查 | ✅ |" >> $GITHUB_STEP_SUMMARY
          echo "| Headless 環境驗證 | ✅ |" >> $GITHUB_STEP_SUMMARY
          echo "| 模型下載驗證 | ✅ |" >> $GITHUB_STEP_SUMMARY
          echo "| 最小轉換測試 | ✅ |" >> $GITHUB_STEP_SUMMARY
          echo "| API 端點檢查 | ✅ |" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "---" >> $GITHUB_STEP_SUMMARY
          echo "⚠️ 這是 upstream 同步版本。請使用 version tags 取得穩定版本。" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "❌ **不會更新 \`latest\` tag**" >> $GITHUB_STEP_SUMMARY