xgj/cache-state/examples/error-handling.yml

# 错误处理和恢复示例
# 演示如何使用状态缓存进行错误处理、状态恢复和重试逻辑

name: Cache State - 错误处理示例

on:
  workflow_dispatch:
    inputs:
      simulate_error:
        description: '模拟错误类型'
        required: false
        default: 'none'
        type: choice
        options:
          - none
          - network_error
          - build_failure
          - deployment_error
      recovery_mode:
        description: '恢复模式'
        required: false
        default: 'auto'
        type: choice
        options:
          - auto
          - manual
          - skip

jobs:
  initialize-operation:
    runs-on: ubuntu-latest
    outputs:
      operation-id: ${{ steps.operation-setup.outputs.operation-id }}
      should-continue: ${{ steps.check-previous.outputs.should-continue }}
    steps:
      - name: 检出代码
        uses: actions/checkout@v4

      - name: 设置操作ID
        id: operation-setup
        run: |
          OPERATION_ID="op-$(date +%s)-${{ github.run_number }}"
          echo "operation-id=${OPERATION_ID}" >> $GITHUB_OUTPUT
          echo "🆔 操作ID: ${OPERATION_ID}"

      - name: 检查上次操作状态
        id: last-operation
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'last-operation-status'
          default-value: 'none'
          action: 'get'
          cache-prefix: 'operation'

      - name: 检查是否有失败的操作需要恢复
        id: check-previous
        run: |
          LAST_STATUS="${{ steps.last-operation.outputs.state-value }}"
          RECOVERY_MODE="${{ github.event.inputs.recovery_mode }}"
          
          echo "🔍 检查上次操作状态: ${LAST_STATUS}"
          echo "🔧 恢复模式: ${RECOVERY_MODE}"
          
          if [[ "${LAST_STATUS}" == "failed" || "${LAST_STATUS}" == "partial" ]]; then
            echo "⚠️ 检测到失败的操作需要处理"
            
            case "${RECOVERY_MODE}" in
              "auto")
                echo "🔄 自动恢复模式，将继续操作"
                echo "should-continue=true" >> $GITHUB_OUTPUT
                ;;
              "manual")
                echo "✋ 手动恢复模式，需要人工干预"
                echo "should-continue=false" >> $GITHUB_OUTPUT
                exit 1
                ;;
              "skip")
                echo "⏭️ 跳过恢复，开始新操作"
                echo "should-continue=true" >> $GITHUB_OUTPUT
                ;;
            esac
          else
            echo "✅ 没有失败的操作，正常继续"
            echo "should-continue=true" >> $GITHUB_OUTPUT
          fi

      - name: 记录操作开始
        if: steps.check-previous.outputs.should-continue == 'true'
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'current-operation-id'
          state-value: ${{ steps.operation-setup.outputs.operation-id }}
          action: 'set'
          cache-prefix: 'operation'

      - name: 记录操作状态为进行中
        if: steps.check-previous.outputs.should-continue == 'true'
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'last-operation-status'
          state-value: 'running'
          action: 'set'
          cache-prefix: 'operation'

  build-with-retry:
    needs: initialize-operation
    if: needs.initialize-operation.outputs.should-continue == 'true'
    runs-on: ubuntu-latest
    steps:
      - name: 检出代码
        uses: actions/checkout@v4

      - name: 检查构建重试状态
        id: retry-status
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'build-retry-count-${{ needs.initialize-operation.outputs.operation-id }}'
          default-value: '0'
          action: 'get'
          cache-prefix: 'retry'

      - name: 准备构建重试
        id: prepare-retry
        run: |
          CURRENT_RETRY=${{ steps.retry-status.outputs.state-value }}
          MAX_RETRIES=3
          
          echo "🔄 当前重试次数: ${CURRENT_RETRY}/${MAX_RETRIES}"
          
          if [[ ${CURRENT_RETRY} -ge ${MAX_RETRIES} ]]; then
            echo "❌ 已达到最大重试次数"
            echo "can-retry=false" >> $GITHUB_OUTPUT
            exit 1
          else
            echo "✅ 可以继续重试"
            echo "can-retry=true" >> $GITHUB_OUTPUT
            echo "next-retry=$((CURRENT_RETRY + 1))" >> $GITHUB_OUTPUT
          fi

      - name: 模拟构建过程
        id: build
        run: |
          echo "🔨 开始构建过程..."
          
          # 根据输入模拟不同类型的错误
          case "${{ github.event.inputs.simulate_error }}" in
            "build_failure")
              echo "❌ 模拟构建失败"
              exit 1
              ;;
            "network_error")
              echo "🌐 模拟网络错误"
              sleep 2
              exit 1
              ;;
            *)
              echo "✅ 构建成功"
              ;;
          esac

      - name: 更新重试计数（失败时）
        if: failure()
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'build-retry-count-${{ needs.initialize-operation.outputs.operation-id }}'
          state-value: ${{ steps.prepare-retry.outputs.next-retry }}
          action: 'set'
          cache-prefix: 'retry'

      - name: 记录构建失败状态
        if: failure()
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'build-status-${{ needs.initialize-operation.outputs.operation-id }}'
          state-value: 'failed'
          action: 'set'
          cache-prefix: 'build'

      - name: 记录构建失败时间
        if: failure()
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'build-failure-time-${{ needs.initialize-operation.outputs.operation-id }}'
          state-value: ${{ github.event.head_commit.timestamp }}
          action: 'set'
          cache-prefix: 'build'

      - name: 清理重试计数（成功时）
        if: success()
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'build-retry-count-${{ needs.initialize-operation.outputs.operation-id }}'
          state-value: ''
          action: 'set'
          cache-prefix: 'retry'

      - name: 记录构建成功状态
        if: success()
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'build-status-${{ needs.initialize-operation.outputs.operation-id }}'
          state-value: 'success'
          action: 'set'
          cache-prefix: 'build'

  deploy-with-rollback:
    needs: [initialize-operation, build-with-retry]
    if: success()
    runs-on: ubuntu-latest
    steps:
      - name: 检出代码
        uses: actions/checkout@v4

      - name: 获取上次成功的部署版本
        id: last-deploy
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'last-successful-deployment'
          default-value: 'v1.0.0'
          action: 'get'
          cache-prefix: 'deploy'

      - name: 备份当前部署状态
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'deployment-backup-${{ needs.initialize-operation.outputs.operation-id }}'
          state-value: ${{ steps.last-deploy.outputs.state-value }}
          action: 'set'
          cache-prefix: 'backup'

      - name: 模拟部署过程
        id: deploy
        run: |
          echo "🚀 开始部署..."
          echo "📦 部署版本: ${{ github.sha }}"
          echo "🔙 备份版本: ${{ steps.last-deploy.outputs.state-value }}"
          
          # 模拟部署错误
          if [[ "${{ github.event.inputs.simulate_error }}" == "deployment_error" ]]; then
            echo "❌ 模拟部署失败"
            exit 1
          fi
          
          echo "✅ 部署成功"

      - name: 记录部署成功
        if: success()
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'last-successful-deployment'
          state-value: ${{ github.sha }}
          action: 'set'
          cache-prefix: 'deploy'

      - name: 清理备份（成功时）
        if: success()
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'deployment-backup-${{ needs.initialize-operation.outputs.operation-id }}'
          state-value: ''
          action: 'set'
          cache-prefix: 'backup'

      - name: 执行回滚（失败时）
        if: failure()
        run: |
          echo "🔄 部署失败，开始回滚..."
          ROLLBACK_VERSION="${{ steps.last-deploy.outputs.state-value }}"
          echo "回滚到版本: ${ROLLBACK_VERSION}"
          
          # 这里执行实际的回滚操作
          echo "✅ 回滚完成"

      - name: 记录回滚状态
        if: failure()
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'rollback-executed-${{ needs.initialize-operation.outputs.operation-id }}'
          state-value: 'true'
          action: 'set'
          cache-prefix: 'rollback'

  finalize-operation:
    needs: [initialize-operation, build-with-retry, deploy-with-rollback]
    if: always() && needs.initialize-operation.outputs.should-continue == 'true'
    runs-on: ubuntu-latest
    steps:
      - name: 检出代码
        uses: actions/checkout@v4

      - name: 确定最终操作状态
        id: final-status
        run: |
          BUILD_RESULT="${{ needs.build-with-retry.result }}"
          DEPLOY_RESULT="${{ needs.deploy-with-rollback.result }}"
          
          echo "🔍 构建结果: ${BUILD_RESULT}"
          echo "🔍 部署结果: ${DEPLOY_RESULT}"
          
          if [[ "${BUILD_RESULT}" == "success" && "${DEPLOY_RESULT}" == "success" ]]; then
            FINAL_STATUS="success"
          elif [[ "${BUILD_RESULT}" == "failure" ]]; then
            FINAL_STATUS="build_failed"
          elif [[ "${DEPLOY_RESULT}" == "failure" ]]; then
            FINAL_STATUS="deploy_failed"
          elif [[ "${BUILD_RESULT}" == "success" && "${DEPLOY_RESULT}" == "skipped" ]]; then
            FINAL_STATUS="partial"
          else
            FINAL_STATUS="unknown"
          fi
          
          echo "final-status=${FINAL_STATUS}" >> $GITHUB_OUTPUT
          echo "🏁 最终状态: ${FINAL_STATUS}"

      - name: 记录最终操作状态
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'last-operation-status'
          state-value: ${{ steps.final-status.outputs.final-status }}
          action: 'set'
          cache-prefix: 'operation'

      - name: 记录操作完成时间
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'last-operation-time'
          state-value: ${{ github.event.head_commit.timestamp }}
          action: 'set'
          cache-prefix: 'operation'

      - name: 生成错误处理报告
        run: |
          echo "📋 错误处理和恢复报告"
          echo "========================"
          echo "操作ID: ${{ needs.initialize-operation.outputs.operation-id }}"
          echo "最终状态: ${{ steps.final-status.outputs.final-status }}"
          echo "模拟错误类型: ${{ github.event.inputs.simulate_error }}"
          echo "恢复模式: ${{ github.event.inputs.recovery_mode }}"
          echo ""
          
          case "${{ steps.final-status.outputs.final-status }}" in
            "success")
              echo "✅ 操作成功完成"
              ;;
            "build_failed")
              echo "❌ 构建失败，请检查构建日志"
              ;;
            "deploy_failed")
              echo "❌ 部署失败，已执行回滚"
              ;;
            "partial")
              echo "⚠️ 部分完成，可能需要手动干预"
              ;;
            *)
              echo "❓ 状态未知，需要进一步调查"
              ;;
          esac

  cleanup-expired-states:
    runs-on: ubuntu-latest
    steps:
      - name: 检出代码
        uses: actions/checkout@v4

      - name: 清理过期的重试状态
        run: |
          echo "🧹 清理过期的状态信息..."
          # 这里可以实现清理逻辑
          # 例如，清理超过一定时间的重试计数、备份等临时状态

      - name: 清理示例 - 重置错误计数
        uses: .actions/xgj/cache-state@v1
        with:
          state-key: 'error-count-global'
          state-value: '0'
          action: 'set'
          cache-prefix: 'cleanup'