abp/.github/workflows/auto-add-seo.yml

name: Auto Add SEO Descriptions

on:
  pull_request:
    paths:
      - 'docs/en/**/*.md'
    branches:
      - 'rel-*'
      - 'dev'
    types: [closed]

jobs:
  add-seo-descriptions:
    if: |
      github.event.pull_request.merged == true &&
      !startsWith(github.event.pull_request.head.ref, 'auto-docs-seo/')      
    runs-on: ubuntu-latest
    permissions:
      contents: write
      pull-requests: write
    
    steps:
      - name: Checkout code
        uses: actions/checkout@v4
        with:
          ref: ${{ github.event.pull_request.merge_commit_sha }}
          fetch-depth: 0
          token: ${{ secrets.GITHUB_TOKEN }}

      - name: Setup Python
        uses: actions/setup-python@v5
        with:
          python-version: '3.11'

      - name: Install dependencies
        run: |
                    pip install openai

      - name: Create new branch for SEO updates
        run: |
          git config --local user.email "github-actions[bot]@users.noreply.github.com"
          git config --local user.name "github-actions[bot]"
          
          # Checkout base branch first
          git checkout ${{ github.event.pull_request.base.ref }}
          
          # Create new branch from base
          BRANCH_NAME="auto-docs-seo/${{ github.event.pull_request.number }}"
          git checkout -b $BRANCH_NAME
          echo "BRANCH_NAME=$BRANCH_NAME" >> $GITHUB_ENV          

      - name: Get changed markdown files from merged PR
        id: changed-files
        run: |
          # Get the list of commits in the PR
          PR_HEAD_SHA="${{ github.event.pull_request.head.sha }}"
          PR_BASE_SHA="${{ github.event.pull_request.base.sha }}"
          
          echo "PR commits range: $PR_BASE_SHA..$PR_HEAD_SHA"
          
          # Get files changed in the PR commits only (only Added and Modified, exclude Deleted)
          FILES=$(git diff --name-only --diff-filter=AM $PR_BASE_SHA..$PR_HEAD_SHA | grep 'docs/en/.*\.md$' || true)
          
          echo "Files changed in the merged PR (added/modified only):"
          echo "$FILES"
          echo ""
          
          echo "changed_files<<EOF" >> $GITHUB_OUTPUT
          echo "$FILES" >> $GITHUB_OUTPUT
          echo "EOF" >> $GITHUB_OUTPUT
          
          if [ -z "$FILES" ]; then
            echo "has_files=false" >> $GITHUB_OUTPUT
            echo "No markdown files changed in docs/en/"
          else
            echo "has_files=true" >> $GITHUB_OUTPUT
            
            # Checkout the changed files from merge commit to get the merged content
            echo ""
            echo "Checking out changed files from merge commit..."
            while IFS= read -r file; do
              if [ -n "$file" ]; then
                echo "  Checking out: $file"
                # Create directory if it doesn't exist
                mkdir -p "$(dirname "$file")"
                # Checkout the file from merge commit
                if ! git show "${{ github.event.pull_request.merge_commit_sha }}:$file" > "$file" 2>err.log; then
                  echo "  Warning: Could not checkout $file"
                  echo "    Reason: $(cat err.log)"
                fi
              fi
            done <<< "$FILES"
            
            echo ""
            echo "Files now in working directory:"
            git status --short
          fi          

      - name: Process changed files and add SEO descriptions
        if: steps.changed-files.outputs.has_files == 'true'
        env:
          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
          IGNORED_FOLDERS: ${{ vars.DOCS_SEO_IGNORED_FOLDERS }}
          CHANGED_FILES: ${{ steps.changed-files.outputs.changed_files }}
        run: |
                    python3 .github/scripts/add_seo_descriptions.py


      - name: Commit and push changes
        if: steps.changed-files.outputs.has_files == 'true'
        run: |
          git add -A docs/en/
          
          if git diff --staged --quiet; then
            echo "No changes to commit"
            echo "has_commits=false" >> $GITHUB_ENV
          else
            git commit -m "docs: Add SEO descriptions to modified documentation files" -m "Related to PR #${{ github.event.pull_request.number }}"
            git push origin ${{ env.BRANCH_NAME }}
            echo "has_commits=true" >> $GITHUB_ENV
          fi          

      - name: Create Pull Request
        if: env.has_commits == 'true'
        uses: actions/github-script@v7
        with:
          script: |
            const fs = require('fs');
            const stats = fs.readFileSync('/tmp/seo_stats.txt', 'utf8').split('\n');
            const processedCount = parseInt(stats[0]) || 0;
            const skippedCount = parseInt(stats[1]) || 0;
            const skippedTooShort = parseInt(stats[2]) || 0;
            const skippedIgnored = parseInt(stats[3]) || 0;
            const prNumber = ${{ github.event.pull_request.number }};
            const baseRef = '${{ github.event.pull_request.base.ref }}';
            const branchName = '${{ env.BRANCH_NAME }}';
            
            if (processedCount > 0) {
              // Read the actually updated files list (not all changed files)
              const updatedFilesStr = fs.readFileSync('/tmp/seo_updated_files.txt', 'utf8');
              const updatedFiles = updatedFilesStr.trim().split('\n').filter(f => f.trim());
              
              let prBody = '🤖 **Automated SEO Descriptions**\n\n';
              prBody += `This PR automatically adds SEO descriptions to documentation files that were modified in PR #${prNumber}.\n\n`;
              prBody += '## 📊 Summary\n';
              prBody += `- ✅ **Updated:** ${processedCount} file(s)\n`;
              prBody += `- ⏭️ **Skipped (total):** ${skippedCount} file(s)\n`;
              if (skippedTooShort > 0) {
                prBody += `  - ⏭️ Content < 200 chars: ${skippedTooShort} file(s)\n`;
              }
              if (skippedIgnored > 0) {
                prBody += `  - 🚫 Ignored folders: ${skippedIgnored} file(s)\n`;
              }
              prBody += '\n## 📝 Modified Files\n';
              prBody += updatedFiles.slice(0, 20).map(f => `- \`${f}\``).join('\n');
              if (updatedFiles.length > 20) {
                prBody += `\n- ... and ${updatedFiles.length - 20} more`;
              }
              prBody += '\n\n## 🔧 Details\n';
              prBody += `- **Related PR:** #${prNumber}\n\n`;
              prBody += 'These descriptions were automatically generated to improve SEO and search engine visibility. 🚀';

              const { data: pr } = await github.rest.pulls.create({
                owner: context.repo.owner,
                repo: context.repo.repo,
                title: `docs: Add SEO descriptions (from PR ${prNumber})`,
                head: branchName,
                base: baseRef,
                body: prBody
              });
              
              console.log(`✅ Created PR: ${pr.html_url}`);
              
              // Add reviewers to the PR (from GitHub variable)
              const reviewersStr = '${{ vars.DOCS_SEO_REVIEWERS || '' }}';
              const reviewers = reviewersStr.split(',').map(r => r.trim()).filter(r => r);
              
              if (reviewers.length === 0) {
                console.log('⚠️ No reviewers specified in DOCS_SEO_REVIEWERS variable.');
                return;
              }
              
              try {
                await github.rest.pulls.requestReviewers({
                  owner: context.repo.owner,
                  repo: context.repo.repo,
                  pull_number: pr.number,
                  reviewers: reviewers,
                  team_reviewers: []
                });
                console.log(`✅ Added reviewers (${reviewers.join(', ')}) to PR ${pr.number}`);
              } catch (error) {
                console.log(`⚠️ Could not add reviewers: ${error.message}`);
              }
            }