docs-commit.translate.yaml

name: GPT Translate per PR

on:
  workflow_dispatch:
    inputs:
      pr_number:
        description: "Optional: PR number or link to sync"
        required: false

jobs:
  gpt_translate:
    runs-on: ubuntu-latest

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

      - name: Collect changed documentation files
        id: collect
        uses: actions/github-script@v7
        env:
          PR_INPUT: ${{ inputs.pr_number }}
        with:
          script: |
            const isDoc = (path) =>
              path.startsWith('docs/en/') &&
              (path.endsWith('.md') || path.endsWith('.json'));

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

            const rawInput = (process.env.PR_INPUT || '').trim();
            let prNumber = null;
            if (rawInput) {
              const match = rawInput.match(/\d+$/);
              if (match) {
                prNumber = Number.parseInt(match[0], 10);
              }
            }
            if (!prNumber) {
              core.setFailed('Please provide the PR number (or a link ending with the number).');
              return;
            }

            core.setOutput('pr_number', String(prNumber));

            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}`);
              }
            }

            const inputs = Array.from(inputSet).join(' ');
            const removals = Array.from(removedSet).join(' ');

            core.setOutput('input_files', inputs);
            core.setOutput('removed_cn', removals);
            core.setOutput('has_inputs', inputSet.size > 0 ? 'true' : 'false');
            core.setOutput('has_removals', removedSet.size > 0 ? 'true' : 'false');
            core.setOutput('pr_number', String(prNumber));

      - 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 in PR #${{ steps.collect.outputs.pr_number }}."
          exit 0

      - 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"

      - 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: Apply deletions to translation branch
        if: >
          steps.collect.outputs.has_inputs == 'true' &&
          steps.collect.outputs.has_removals == 'true' &&
          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"

          for file in $REMOVED_FILES; do
            if [ -f "$file" ]; then
              rm -f "$file"
              echo "Removed $file"
            fi
          done

          # Clean up empty directories under docs/cn
          find docs/cn -mindepth 1 -type d -empty -print -delete

          if git status --porcelain | grep .; 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: sync deletions for PR #${{ steps.collect.outputs.pr_number }}"
            git push origin "$TRANSLATION_BRANCH"
          else
            echo "No deletions 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.collect.outputs.pr_number }}
          base: main
          commit-message: "chore: sync deletions for PR #${{ steps.collect.outputs.pr_number }}"
          title: "AI Translate cleanup for PR #${{ steps.collect.outputs.pr_number }}"
          body: |
            This automated PR removes translated files that no longer have an English source from PR #${{ steps.collect.outputs.pr_number }}.