ja-sync-check.yml

name: Japanese Translation Sync Check

on:
  pull_request:
    branches:
      - main
    paths:
      - '**/*.mdx'
      - '!ja/**/*.mdx'
      - '!snippets/ja/**/*.mdx'

jobs:
  check-ja-translations:
    runs-on: ubuntu-latest

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

      - name: Setup Node.js
        uses: actions/setup-node@v4
        with:
          node-version: '20'

      - name: Install dependencies
        run: npm ci

      - name: Check changed MDX files have ja equivalents and check for moved files
        id: check-translations
        run: |
          # Get list of changed MDX files
          CHANGED_FILES=$(git diff --name-only --diff-filter=ACMRT ${{ github.event.pull_request.base.sha }} ${{ github.event.pull_request.head.sha }} | grep "\.mdx$" | grep -v "^ja/" | grep -v "^zh/" | grep -v "^snippets/zh/" | grep -v "^snippets/ja/" || true)

          # Get list of deleted MDX files (potentially moved)
          DELETED_FILES=$(git diff --name-only --diff-filter=D ${{ github.event.pull_request.base.sha }} ${{ github.event.pull_request.head.sha }} | grep "\.mdx$" | grep -v "^ja/" | grep -v "^zh/" | grep -v "^snippets/zh/" | grep -v "^snippets/ja/" || true)

          # Get list of added MDX files (potentially destinations of moves)
          ADDED_FILES=$(git diff --name-only --diff-filter=A ${{ github.event.pull_request.base.sha }} ${{ github.event.pull_request.head.sha }} | grep "\.mdx$" | grep -v "^ja/" | grep -v "^zh/" | grep -v "^snippets/zh/" | grep -v "^snippets/ja/" || true)

          # Exit if no MDX files were changed, deleted or added
          if [ -z "$CHANGED_FILES" ] && [ -z "$DELETED_FILES" ] && [ -z "$ADDED_FILES" ]; then
            echo "No MDX files changed outside of ja and snippets/ja directories. Skipping check."
            exit 0
          fi

          # Check for moved files - these would be deleted and added in the same PR
          MOVED_FILES=""
          MOVE_DESTINATIONS=""

          if [ ! -z "$DELETED_FILES" ] && [ ! -z "$ADDED_FILES" ]; then
            echo "Potential file moves detected:"
            echo "------------------------"
            echo "Deleted files:"
            echo "$DELETED_FILES"
            echo "------------------------"
            echo "Added files:"
            echo "$ADDED_FILES"
            echo "------------------------"

            # Check if redirects are added in docs.json for deleted files
            DOCS_JSON_CHANGED=$(git diff --name-only ${{ github.event.pull_request.base.sha }} ${{ github.event.pull_request.head.sha }} | grep -q "docs.json" && echo "true" || echo "false")

            if [ "$DOCS_JSON_CHANGED" = "false" ]; then
              echo "Warning: docs.json was not modified - if files were moved, redirects should be added."
            fi

            # Check docs.json content for redirects if it was changed
            if [ "$DOCS_JSON_CHANGED" = "true" ]; then
              echo "Checking docs.json for redirects..."

              # Use Node.js to check redirects with wildcard support
              MISSING_REDIRECTS=$(node -e "
                const fs = require('fs');
                const docsJson = JSON.parse(fs.readFileSync('docs.json', 'utf8'));
                const deletedFiles = process.argv.slice(1);
                const redirects = docsJson.redirects || [];

                function matchesPattern(filePath, pattern) {
                  pattern = pattern.replace(/^\//, '');
                  filePath = filePath.replace(/^\//, '');
                  const regexPattern = pattern
                    .replace(/\//g, '\\\\/')
                    .replace(/:slug\*/g, '.*')
                    .replace(/:slug/g, '[^/]+')
                    .replace(/\*/g, '.*');
                  const regex = new RegExp(\`^\${regexPattern}$\`);
                  return regex.test(filePath);
                }

                const missing = [];
                for (const file of deletedFiles) {
                  const sourcePath = file.replace(/\.mdx$/, '');
                  const hasRedirect = redirects.some(redirect => {
                    const redirectSource = redirect.source.replace(/^\//, '');
                    const filePathNormalized = sourcePath.replace(/^\//, '');
                    return redirectSource === filePathNormalized || matchesPattern(filePathNormalized, redirectSource);
                  });

                  if (hasRedirect) {
                    console.log(\`Found redirect for moved file: \${file}\`);
                  } else {
                    console.log(\`Missing redirect in docs.json for: \${file}\`);
                    missing.push(file);
                  }
                }

                if (missing.length > 0) {
                  process.exit(1);
                }
              " $DELETED_FILES 2>&1 | tee /tmp/redirect_check.log || true)

              # Check if there were missing redirects
              if echo "$MISSING_REDIRECTS" | grep -q "Missing redirect"; then
                MOVED_FILES=$(echo "$MISSING_REDIRECTS" | grep "Missing redirect" | sed 's/.*for: //')
              else
                MOVED_FILES=""
              fi
            else
              # If docs.json wasn't changed, all deleted files are suspect
              MOVED_FILES="$DELETED_FILES"
            fi
          fi

          echo "Changed MDX files outside ja directory:"
          echo "$CHANGED_FILES"
          echo "------------------------"

          # Check each file for corresponding ja updates
          MISSING_TRANSLATIONS=""
          for file in $CHANGED_FILES; do
            # Determine the corresponding ja path based on whether it's a snippet or not
            if [[ $file == snippets/* ]]; then
              # For snippets, the Japanese version should be in snippets/ja/
              ja_file="${file/snippets\//snippets\/ja\/}"
            else
              # For regular docs, the Japanese version should be in ja/
              ja_file="ja/${file}"
            fi

            # Check if the corresponding ja file was also modified
            if git diff --name-only --diff-filter=ACMRT ${{ github.event.pull_request.base.sha }} ${{ github.event.pull_request.head.sha }} | grep -q "^${ja_file}$"; then
              echo "Found corresponding change: ${ja_file}"
            else
              # Check if the ja file exists at all
              if [ -f "${ja_file}" ]; then
                echo "Missing update to existing file: ${ja_file}"
                MISSING_TRANSLATIONS="$MISSING_TRANSLATIONS ${ja_file}"
              else
                echo "Missing equivalent file: ${ja_file}"
                MISSING_TRANSLATIONS="$MISSING_TRANSLATIONS ${ja_file}"
              fi
            fi
          done

          # For added files, check corresponding ja files
          for file in $ADDED_FILES; do
            # Determine the corresponding ja path based on whether it's a snippet or not
            if [[ $file == snippets/* ]]; then
              # For snippets, the Japanese version should be in snippets/ja/
              ja_file="${file/snippets\//snippets\/ja\/}"
            else
              # For regular docs, the Japanese version should be in ja/
              ja_file="ja/${file}"
            fi

            # Check if the corresponding ja file was also added or renamed
            if git diff --name-only --diff-filter=AR ${{ github.event.pull_request.base.sha }} ${{ github.event.pull_request.head.sha }} | grep -q "^${ja_file}$"; then
              echo "Found corresponding added/renamed file: ${ja_file}"
            # Also check with rename detection to catch renamed files
            elif git diff --name-status --find-renames ${{ github.event.pull_request.base.sha }} ${{ github.event.pull_request.head.sha }} | grep -E "^R[0-9]+\s+" | awk '{print $3}' | grep -q "^${ja_file}$"; then
              echo "Found corresponding renamed file: ${ja_file}"
            else
              echo "Missing equivalent for added file: ${ja_file}"
              MISSING_TRANSLATIONS="$MISSING_TRANSLATIONS ${ja_file}"
            fi
          done

          # Determine if we need to fail the check
          SHOULD_FAIL="false"

          # Check for missing translations (warning only, does not fail the build)
          if [ "$MISSING_TRANSLATIONS" != "" ]; then
            echo "------------------------"
            echo "The following translations need to be updated:"
            for missing in $MISSING_TRANSLATIONS; do
              echo "- $missing"
            done
            echo "::warning::The following ja translation files need to be updated to stay in sync with the English version:$MISSING_TRANSLATIONS"
          fi

          # Check for moved files without redirects
          if [ "$MOVED_FILES" != "" ]; then
            SHOULD_FAIL="true"
            echo "------------------------"
            echo "The following files were moved but missing redirects in docs.json:"
            for moved in $MOVED_FILES; do
              echo "- $moved"
            done
            echo "------------------------"
            echo "Please add redirects in docs.json for moved files using the format:"
            echo '{"redirects":[{"source":"/path/to/old-file","destination":"/path/to/new-file"}]}'
            echo "------------------------"
          fi

          # Exit with error if we found issues
          if [ "$SHOULD_FAIL" = "true" ]; then
            echo "------------------------"
            echo "Please fix the issues above before merging this PR."
            exit 1
          else
            echo "------------------------"
            echo "All checks passed!"
          fi