translation-zh.yaml

name: Sync TiDB Cloud Docs from EN to ZH

concurrency:
  group: translation-zh-cloud-${{ github.event_name }}
  cancel-in-progress: true

on:
  schedule:
    - cron: "0 9 * * 3" # Runs at 17:00 every Wednesday (Beijing time, UTC+8)
  workflow_dispatch:
    inputs:
      file_names:
        description: "Specify Cloud files to translate (comma-separated list)"
        required: true
        type: string
      source_files_translation_mode:
        description: "Translate specified files incrementally or as full files"
        required: true
        type: choice
        default: incremental
        options:
          - incremental
          - full

env:
  SOURCE_REPO: pingcap/docs
  SOURCE_BRANCH: release-8.5
  TERMS_BRANCH: master
  CN_CLOUD_BRANCH: i18n-zh-release-8.5
  CLOUD_TOC_FILES: TOC-tidb-cloud.md,TOC-tidb-cloud-starter.md,TOC-tidb-cloud-essential.md,TOC-tidb-cloud-premium.md,TOC-tidb-cloud-releases.md
  CLOUD_INDEX_FILES: tidb-cloud/dedicated/_index.md,tidb-cloud/essential/_index.md,tidb-cloud/premium/_index.md,tidb-cloud/releases/_index.md,tidb-cloud/starter/_index.md
  AI_TRANSLATOR_REPO: qiancai/ai-pr-translator
  AI_TRANSLATOR_REF: main

jobs:
  translate:
    if: github.repository == 'pingcap/docs'
    runs-on: ubuntu-latest

    permissions:
      contents: write
      pull-requests: write

    steps:
      - uses: actions/checkout@v6
        name: Checkout docs target branch
        with:
          ref: ${{ env.CN_CLOUD_BRANCH }}
          path: docs

      - uses: actions/checkout@v6
        name: Checkout docs source branch
        with:
          repository: ${{ env.SOURCE_REPO }}
          ref: ${{ env.SOURCE_BRANCH }}
          path: docs-source
          fetch-depth: 0
          persist-credentials: false

      - uses: actions/checkout@v6
        name: Checkout docs terms branch
        with:
          repository: ${{ env.SOURCE_REPO }}
          ref: ${{ env.TERMS_BRANCH }}
          path: docs-terms
          persist-credentials: false

      - uses: actions/checkout@v6
        name: Checkout ai-pr-translator
        with:
          repository: ${{ env.AI_TRANSLATOR_REPO }}
          ref: ${{ env.AI_TRANSLATOR_REF }}
          path: ai-pr-translator
          persist-credentials: false

      - uses: actions/setup-python@v6
        name: Setup Python
        with:
          python-version: "3.9"
          cache: pip
          cache-dependency-path: ai-pr-translator/scripts/requirements.txt

      - name: Install Python dependencies
        shell: bash
        run: |
          set -euo pipefail
          python -m pip install --upgrade pip
          pip install -r ai-pr-translator/scripts/requirements.txt

      - name: Resolve commit range
        id: commits
        shell: bash
        working-directory: docs
        run: |
          set -euo pipefail

          readarray -t cursor_values < <(python - <<'PY'
          import json
          from pathlib import Path

          data = json.loads(Path("latest_translation_commit.json").read_text(encoding="utf-8"))
          print((data.get("target") or "").strip())
          print((data.get("sha") or "").strip())
          PY
          )

          cursor_target="${cursor_values[0]:-}"
          base_ref="${cursor_values[1]:-}"
          head_ref="$(git -C "${GITHUB_WORKSPACE}/docs-source" rev-parse HEAD)"

          if [ -z "${base_ref}" ]; then
            echo "latest_translation_commit.json does not contain a source sha" >&2
            exit 1
          fi

          if [ -n "${cursor_target}" ] && [ "${cursor_target}" != "${SOURCE_BRANCH}" ]; then
            echo "latest_translation_commit.json target is ${cursor_target}, expected ${SOURCE_BRANCH}" >&2
            exit 1
          fi

          echo "base_ref=${base_ref}" >> "${GITHUB_OUTPUT}"
          echo "head_ref=${head_ref}" >> "${GITHUB_OUTPUT}"

      - name: Resolve Cloud source file filter
        id: source_files
        shell: bash
        env:
          INPUT_FILE_NAMES: ${{ inputs.file_names || '' }}
          BASE_REF: ${{ steps.commits.outputs.base_ref }}
          HEAD_REF: ${{ steps.commits.outputs.head_ref }}
          DOCS_SOURCE_PATH: ${{ github.workspace }}/docs-source
          CLOUD_TOC_FILES: ${{ env.CLOUD_TOC_FILES }}
          CLOUD_INDEX_FILES: ${{ env.CLOUD_INDEX_FILES }}
        run: |
          set -euo pipefail
          python ai-pr-translator/scripts/resolve_cloud_source_files.py

      - name: Run commit sync workflow
        id: sync
        if: steps.source_files.outputs.has_source_changes == 'true'
        continue-on-error: true
        shell: bash
        env:
          GITHUB_TOKEN: ${{ github.token }}
          AZURE_OPENAI_KEY: ${{ secrets.AZURE_OPENAI_KEY }}
          OPENAI_BASE_URL: ${{ secrets.AZURE_OPENAI_BASE_URL }}
          SOURCE_REPO: ${{ env.SOURCE_REPO }}
          TARGET_REPO: ${{ env.SOURCE_REPO }}
          SOURCE_BRANCH: ${{ env.SOURCE_BRANCH }}
          SOURCE_BASE_REF: ${{ steps.commits.outputs.base_ref }}
          SOURCE_HEAD_REF: ${{ steps.commits.outputs.head_ref }}
          SOURCE_FOLDER: ""
          SOURCE_FILES: ${{ steps.source_files.outputs.files }}
          SOURCE_FILES_TRANSLATION_MODE: ${{ inputs.source_files_translation_mode || 'incremental' }}
          SOURCE_LANGUAGE: English
          TARGET_LANGUAGE: Chinese
          IGNORE_RESOURCE_CARD_SECTION: "Yes"
          SOURCE_REPO_PATH: ${{ github.workspace }}/docs-source
          TARGET_REF: ${{ env.CN_CLOUD_BRANCH }}
          PREFER_LOCAL_TARGET_FOR_READ: "true"
          TARGET_REPO_PATH: ${{ github.workspace }}/docs
          AI_PROVIDER: azure
          TERMS_PATH: ${{ github.workspace }}/docs-terms/resources/terms.md
          FAIL_ON_TRANSLATION_ERROR: "true"
          SKIP_TRANSLATING_CLOUD_DOCS_TO_ZH: "false"
          SKIP_TRANSLATING_AI_DOCS_TO_ZH: "true"
        run: |
          set -euo pipefail
          cd ai-pr-translator/scripts
          python commit_sync_workflow.py

      - name: Prepare translated changes
        id: changes
        if: always() && steps.source_files.outputs.has_source_changes == 'true'
        shell: bash
        working-directory: docs
        env:
          HEAD_REF: ${{ steps.commits.outputs.head_ref }}
          SOURCE_BRANCH: ${{ env.SOURCE_BRANCH }}
          FAILURE_REPORT: ${{ github.workspace }}/ai-pr-translator/scripts/temp_output/translation-failures.md
        run: |
          set -euo pipefail

          if [ -s "${FAILURE_REPORT}" ]; then
            {
              echo "failure_summary<<EOF"
              cat "${FAILURE_REPORT}"
              echo
              echo "> Before merging this PR, manually translate the failed files above or rerun this workflow with workflow_dispatch and file_names set to those paths."
              echo "EOF"
            } >> "${GITHUB_OUTPUT}"
          fi

          if [ -z "$(git status --porcelain)" ]; then
            echo "has_changes=false" >> "${GITHUB_OUTPUT}"
            exit 0
          fi

          if [ "${GITHUB_EVENT_NAME}" = "schedule" ]; then
            python -c 'import json, os; from pathlib import Path; Path("latest_translation_commit.json").write_text(json.dumps({"target": os.environ["SOURCE_BRANCH"], "sha": os.environ["HEAD_REF"]}, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")'
          else
            echo "Skipping latest_translation_commit.json update for ${GITHUB_EVENT_NAME} run."
          fi

          echo "has_changes=true" >> "${GITHUB_OUTPUT}"

      - name: Set build metadata
        id: build_meta
        if: steps.changes.outputs.has_changes == 'true'
        shell: bash
        run: |
          echo "date=$(TZ=Asia/Shanghai date +'%Y-%m-%d')" >> "${GITHUB_OUTPUT}"
          echo "id=$(TZ=Asia/Shanghai date +'%Y%m%d')-$(date +%s)" >> "${GITHUB_OUTPUT}"

      - name: Create PR
        if: steps.changes.outputs.has_changes == 'true'
        uses: peter-evans/create-pull-request@v8
        with:
          path: docs
          token: ${{ github.token }}
          branch: zh-translation/cloud-${{ steps.build_meta.outputs.id }}
          base: ${{ env.CN_CLOUD_BRANCH }}
          title: "${{ env.CN_CLOUD_BRANCH }}: translate Cloud doc changes from ${{ env.SOURCE_REPO }} ${{ env.SOURCE_BRANCH }} on ${{ steps.build_meta.outputs.date }}"
          labels: |
            translation/no-need
          body: |
            ### What is changed, added or deleted? (Required)

            Translate Cloud documentation changes from `pingcap/docs` `${{ env.SOURCE_BRANCH }}` to Chinese via `ai-pr-translator` commit-diff sync.

            English commit diff:

            https://github.com/${{ env.SOURCE_REPO }}/compare/${{ steps.commits.outputs.base_ref }}...${{ steps.commits.outputs.head_ref }}

            ### Which TiDB version(s) do your changes apply to? (Required)

            - [x] ${{ env.CN_CLOUD_BRANCH }}

            ### What is the related PR or file link(s)?

            - Source repo: `${{ env.SOURCE_REPO }}`
            - Source branch: `${{ env.SOURCE_BRANCH }}`
            - Cloud TOCs: `${{ env.CLOUD_TOC_FILES }}`
            - Source files: `${{ steps.source_files.outputs.files }}`
            - Sync outcome: `${{ steps.sync.outcome }}`

            ${{ steps.changes.outputs.failure_summary }}

            ### Do your changes match any of the following descriptions?

            - [ ] Delete files
            - [ ] Change aliases
            - [ ] Need modification after applied to another branch
            - [ ] Might cause conflicts after applied to another branch
          delete-branch: true