docs-pr.translate.yaml

name: GPT Translate by PR

on:
  workflow_dispatch:
    inputs:
      pr:
        description: "Pull request URL or number (e.g., 123 or https://github.com/org/repo/pull/123)"
        required: true

permissions:
  id-token: write
  pull-requests: write
  checks: write
  statuses: write
  contents: write

jobs:
  gpt_translate:
    runs-on: ubuntu-latest

    steps:
      - name: Parse PR input
        id: pr
        uses: actions/github-script@v7
        env:
          PR_INPUT: ${{ inputs.pr }}
        with:
          script: |
            const raw = process.env.PR_INPUT || '';
            if (!raw || raw.trim() === '') {
              throw new Error('PR input is required.');
            }

            const trimmed = raw.trim();
            let prNumber = null;

            if (/^\d+$/.test(trimmed)) {
              prNumber = parseInt(trimmed, 10);
            } else {
              const matches = trimmed.match(/\d+/g);
              if (matches && matches.length > 0) {
                prNumber = parseInt(matches[matches.length - 1], 10);
              }
            }

            if (!prNumber || Number.isNaN(prNumber)) {
              throw new Error(`Unable to extract pull request number from input: ${trimmed}`);
            }

            const { data: pr } = await github.rest.pulls.get({
              owner: context.repo.owner,
              repo: context.repo.repo,
              pull_number: prNumber,
            });

            core.setOutput('pr_number', String(prNumber));
            core.setOutput('head_ref', pr.head.ref);
            core.setOutput('head_sha', pr.head.sha);
            core.setOutput('base_ref', pr.base.ref);
            core.setOutput('title', pr.title);
            core.setOutput('html_url', pr.html_url);

      - name: Collect changed documentation files
        id: collect
        uses: actions/github-script@v7
        env:
          PR_NUMBER: ${{ steps.pr.outputs.pr_number }}
        with:
          script: |
            const rawPrNumber = process.env.PR_NUMBER || '';
            const prNumber = Number.parseInt(rawPrNumber, 10);
            if (!prNumber) {
              throw new Error(`PR number is missing or invalid: "${rawPrNumber}"`);
            }

            const isDoc = (path) =>
              path.startsWith('docs/en/') &&
              (path.endsWith('.md') || path.endsWith('.json'));

            const toCnPath = (path) =>
              `docs/cn/${path.slice('docs/en/'.length)}`;

            const files = await github.paginate(
              github.rest.pulls.listFiles,
              {
                owner: context.repo.owner,
                repo: context.repo.repo,
                pull_number: prNumber,
                per_page: 100,
              }
            );

            const inputSet = new Set();
            const removedSet = new Set();

            for (const file of files) {
              const { filename, status, previous_filename: prev } = file;

              if (status === 'removed' && isDoc(filename)) {
                removedSet.add(toCnPath(filename));
                continue;
              }

              if (status === 'renamed' && prev && isDoc(prev)) {
                removedSet.add(toCnPath(prev));
              }

              if (isDoc(filename) && status !== 'removed') {
                inputSet.add(`./${filename}`);
              }
            }

            core.setOutput('input_files', Array.from(inputSet).join(' '));
            core.setOutput('removed_cn', Array.from(removedSet).join(' '));
            core.setOutput('has_inputs', inputSet.size > 0 ? 'true' : 'false');
            core.setOutput('has_removals', removedSet.size > 0 ? 'true' : 'false');

      - name: Exit if no documentation changes
        if: steps.collect.outputs.has_inputs != 'true' && steps.collect.outputs.has_removals != 'true'
        run: |
          echo "No English documentation additions, updates, or deletions detected for PR #${{ steps.pr.outputs.pr_number }} (${{ steps.pr.outputs.html_url }})."
          exit 0

      - name: Checkout repository
        uses: actions/checkout@v4
        with:
          fetch-depth: 0
          ref: main

      - name: Populate PR content
        if: steps.collect.outputs.has_inputs == 'true'
        env:
          PR: ${{ steps.pr.outputs.pr_number }}
          INPUT_FILES: ${{ steps.collect.outputs.input_files }}
        run: |
          set -euo pipefail
          git fetch --no-tags --depth=1 origin pull/${PR}/head:refs/remotes/origin/pr-${PR}
          for file in $INPUT_FILES; do
            clean_path=${file#./}
            mkdir -p "$(dirname "$clean_path")"
            git show origin/pr-${PR}:"$clean_path" > "$clean_path"
          done

      - name: Snapshot existing translation branches
        if: steps.collect.outputs.has_inputs == 'true'
        id: snapshot
        run: |
          git ls-remote --heads origin 'translation-*' | awk '{print $2}' | sed 's#refs/heads/##' | sort > /tmp/translation-branches-before.txt
          echo "before=/tmp/translation-branches-before.txt" >> "$GITHUB_OUTPUT"

      - name: Run GPT Translate
        if: steps.collect.outputs.has_inputs == 'true'
        uses: BohuTANG/gpt-translate-refine@v1.4.4
        with:
          github_token: ${{ secrets.GITHUB_TOKEN }}
          api_key: ${{ secrets.API_KEY }}
          base_url: ${{ secrets.BASE_URL }}
          ai_model: ${{ secrets.LLM_MODEL }}
          refine_ai_model: ${{ secrets.REFINE_LLM_MODEL }}
          target_lang: "Simplified-Chinese"
          system_prompt: ".github/workflows/prompt.txt"
          refine_system_prompt: ".github/workflows/refine_prompt.txt"
          temperature: ${{ secrets.TEMPERATURE }}
          refine_temperature: ${{ secrets.REFINE_TEMPERATURE }}
          input_files: "${{ steps.collect.outputs.input_files }}"
          output_files: "docs/cn/**/*.{md,json}"
          pr_title: "Add LLM Translations V2 for PR #${{ steps.pr.outputs.pr_number }}"

      - name: Identify translation branch
        if: steps.collect.outputs.has_inputs == 'true'
        id: branch
        env:
          SNAPSHOT_FILE: ${{ steps.snapshot.outputs.before }}
        run: |
          git ls-remote --heads origin 'translation-*' | awk '{print $2}' | sed 's#refs/heads/##' | sort > /tmp/translation-branches-after.txt
          comm -13 "$SNAPSHOT_FILE" /tmp/translation-branches-after.txt > /tmp/new-translation-branches.txt
          branch=$(tail -n 1 /tmp/new-translation-branches.txt)
          if [ -n "$branch" ]; then
            echo "Discovered translation branch: $branch"
            echo "branch=$branch" >> "$GITHUB_OUTPUT"
          else
            echo "Unable to determine translation branch created by GPT workflow."
            echo "branch=" >> "$GITHUB_OUTPUT"
          fi

      - name: Annotate translation PR
        if: steps.branch.outputs.branch != ''
        uses: actions/github-script@v7
        env:
          SOURCE_PR: ${{ steps.pr.outputs.pr_number }}
          SOURCE_PR_URL: ${{ steps.pr.outputs.html_url }}
          REMOVED_FILES: ${{ steps.collect.outputs.removed_cn }}
        with:
          script: |
            const branch = '${{ steps.branch.outputs.branch }}';
            if (!branch) {
              core.info('No translation branch detected, skipping annotation.');
              return;
            }

            const owner = context.repo.owner;
            const repo = context.repo.repo;

            const prs = await github.paginate(github.rest.pulls.list, {
              owner,
              repo,
              head: `${owner}:${branch}`,
              state: 'open',
              per_page: 100,
            });

            if (prs.length === 0) {
              core.warning(`Unable to find translation PR for branch ${branch}`);
              return;
            }

            const translationPr = prs[0];
            const sourcePrNumber = process.env.SOURCE_PR;
            const sourcePrUrl = process.env.SOURCE_PR_URL;

            const removedFilesRaw = (process.env.REMOVED_FILES || '').trim();
            const removedFilesList = removedFilesRaw
              ? removedFilesRaw.split(/\s+/).map((file) => `- \`${file}\``).join('\n')
              : '- *(none)*';

            const annotationBlock = [
              '---',
              '### 🔗 Source Pull Request',
              `- PR #${sourcePrNumber}`,
              `- ${sourcePrUrl}`,
              '',
              '### 🗑️ Synchronized CN Deletions',
              removedFilesList,
              '---',
            ].join('\n');

            let updatedBody = translationPr.body || '';

            const marker = '\n---\n### 🔗 Source Pull Request';
            if (updatedBody.includes(marker)) {
              updatedBody = updatedBody.replace(/---\n### 🔗 Source Pull Request[\s\S]*?---/m, annotationBlock);
            } else {
              updatedBody = `${updatedBody.trim()}\n\n${annotationBlock}`.trim();
            }

            const updatedTitle = translationPr.title.includes(`#${sourcePrNumber}`)
              ? translationPr.title
              : `${translationPr.title} (from #${sourcePrNumber})`;

            await github.rest.pulls.update({
              owner,
              repo,
              pull_number: translationPr.number,
              title: updatedTitle,
              body: updatedBody,
            });
      - name: Finalize translation branch
        if: steps.branch.outputs.branch != ''
        env:
          REMOVED_FILES: ${{ steps.collect.outputs.removed_cn }}
          TRANSLATION_BRANCH: ${{ steps.branch.outputs.branch }}
        run: |
          set -euo pipefail
          git fetch origin "$TRANSLATION_BRANCH"
          git checkout "$TRANSLATION_BRANCH"

          if command -v sudo >/dev/null 2>&1; then
            sudo chown -R "$(id -u)":"$(id -g)" .git docs || true
          fi

          changed=0

          if [ -f ".translation-init" ]; then
            git rm -f .translation-init
            echo "Removed .translation-init"
            changed=1
          fi

          if [ -n "$REMOVED_FILES" ]; then
            for file in $REMOVED_FILES; do
              if [ -f "$file" ]; then
                rm -f "$file"
                echo "Removed $file"
                changed=1
              fi
            done
            if [ -d docs/cn ]; then
              find docs/cn -mindepth 1 -type d -empty -print -delete
            fi
          fi

          if [ "$changed" -eq 1 ]; then
            git config user.name "github-actions[bot]"
            git config user.email "github-actions[bot]@users.noreply.github.com"
            git add -A
            git commit -m "chore: finalize translation for PR #${{ steps.pr.outputs.pr_number }}"
            git push origin "$TRANSLATION_BRANCH"
          else
            echo "No cleanup changes to commit."
          fi

      - name: Prepare deletion-only changes
        if: steps.collect.outputs.has_inputs != 'true' && steps.collect.outputs.has_removals == 'true'
        env:
          REMOVED_FILES: ${{ steps.collect.outputs.removed_cn }}
        run: |
          set -euo pipefail
          for file in $REMOVED_FILES; do
            if [ -f "$file" ]; then
              rm -f "$file"
              echo "Removed $file"
            fi
          done
          find docs/cn -mindepth 1 -type d -empty -print -delete

      - name: Open deletion-only translation PR
        if: steps.collect.outputs.has_inputs != 'true' && steps.collect.outputs.has_removals == 'true'
        uses: peter-evans/create-pull-request@v6
        with:
          token: ${{ secrets.GITHUB_TOKEN }}
          branch: translation-pr-${{ steps.pr.outputs.pr_number }}
          base: main
          commit-message: "chore: sync deletions for PR #${{ steps.pr.outputs.pr_number }}"
          title: "AI Translate cleanup for PR #${{ steps.pr.outputs.pr_number }}"
          body: |
            This automated PR removes translated files that no longer have an English source from PR #${{ steps.pr.outputs.pr_number }}.