gw_comment_trigger.yml

name: Bugspots Analysis via Comment

on:
  issue_comment:
    types: [created]

jobs:
  parse-comment:
    if: github.event.issue.pull_request == null && contains(github.event.comment.body, 'gw --repo')
    runs-on: ubuntu-latest
    permissions:
      contents: read
      issues: write
      
    outputs:
      repo_url: ${{ steps.parse.outputs.repo_url }}
      limit: ${{ steps.parse.outputs.limit }}
      branch: ${{ steps.parse.outputs.branch }}
      
    steps:
      - name: Parse comment
        id: parse
        env:
          COMMENT_BODY: ${{ github.event.comment.body }}
        run: |
          echo "Comment body: $COMMENT_BODY"
          
          # Extract repository URL
          # Extract repository URL using environment variable
          repo_url=$(echo "$COMMENT_BODY" | grep -oP 'gw --repo \K[^\s]+' || echo "")
          if [ -z "$repo_url" ]; then
            echo "Error: No repository URL found in comment"
            exit 1
          fi
          
          # Validate repository URL format
          if ! echo "$repo_url" | grep -qE '^https://github\.com/[a-zA-Z0-9._-]+/[a-zA-Z0-9._-]+/?$'; then
            echo "Error: Invalid repository URL format. Must be a GitHub repository URL."
            exit 1
          fi
          
          # Extract limit
          limit_raw=$(echo "$COMMENT_BODY" | grep -oP -- '--limit\s+\K\d+' || echo "")
          
          # Validate and set limit (default to 10 if not specified or invalid)
          if [ -n "$limit_raw" ] && [ "$limit_raw" -gt 0 ] && [ "$limit_raw" -le 100 ]; then
            limit="$limit_raw"
          else
            limit="10"
            if [ -n "$limit_raw" ]; then
              echo "Warning: Invalid limit value '$limit_raw', using default 10"
            fi
          fi
          
          # Extract branch
          branch_raw=$(echo "$COMMENT_BODY" | grep -oP -- '--branch\s+\K[^\s]+' || echo "")
          if [ -n "$branch_raw" ]; then
            branch="$branch_raw"
          else
            branch=""
          fi
          
          echo "repo_url=$repo_url" >> $GITHUB_OUTPUT
          echo "limit=$limit" >> $GITHUB_OUTPUT
          echo "branch=$branch" >> $GITHUB_OUTPUT
          echo "Parsed repo: $repo_url, limit: $limit, branch: $branch"

  run-bugspots:
    needs: parse-comment
    runs-on: ubuntu-latest
    permissions:
      contents: read
      issues: write
      
    steps:
      - name: Checkout repository
        uses: actions/checkout@v4
        
      - name: Set up Ruby
        uses: ruby/setup-ruby@v1
        with:
          ruby-version: '2.7'
          
      - name: Install bugspots gem
        run: gem install bugspots
        
      - name: Run Bugspots Comment Analyzer
        id: bugspots
        continue-on-error: true
        env:
          REPO_URL: ${{ needs.parse-comment.outputs.repo_url }}
          LIMIT: ${{ needs.parse-comment.outputs.limit }}
          BRANCH: ${{ needs.parse-comment.outputs.branch }}
        run: |
          chmod +x BugPredict/gw.sh
          BugPredict/gw.sh "$REPO_URL" "$LIMIT" "$BRANCH"
          
      - name: Check for repository errors
        id: check-errors
        env:
          REPO_URL: ${{ needs.parse-comment.outputs.repo_url }}
          BRANCH: ${{ needs.parse-comment.outputs.branch }}
        run: |
          repo_name=$(basename "$REPO_URL" .git)
          
          # Check if the bugspots step failed
          if [ "${{ steps.bugspots.outcome }}" = "failure" ]; then
            echo "analysis_failed=true" >> $GITHUB_OUTPUT
            
            # Check for specific error types
            if [ -f "bugspots-results/bugspots-${repo_name}.err" ]; then
              error_content=$(cat "bugspots-results/bugspots-${repo_name}.err")
              if [[ "$error_content" == *"Clone failed"* ]]; then
                echo "error_type=clone_failed" >> $GITHUB_OUTPUT
              elif [[ "$error_content" == *"Invalid Git repository"* ]]; then
                echo "error_type=invalid_repo" >> $GITHUB_OUTPUT
              elif [[ "$error_content" == *"No commits found"* ]]; then
                echo "error_type=no_commits" >> $GITHUB_OUTPUT
              elif [[ "$error_content" == *"no such branch"* ]]; then
                echo "error_type=invalid_branch" >> $GITHUB_OUTPUT
              else
                echo "error_type=unknown" >> $GITHUB_OUTPUT
              fi
            else
              echo "error_type=unknown" >> $GITHUB_OUTPUT
            fi
          else
            echo "analysis_failed=false" >> $GITHUB_OUTPUT
          fi
          
      - name: Prepare error comment
        if: steps.check-errors.outputs.analysis_failed == 'true'
        env:
          REPO_URL: ${{ needs.parse-comment.outputs.repo_url }}
          ERROR_TYPE: ${{ steps.check-errors.outputs.error_type }}
        run: |
          echo "## ❌ Repository Analysis Failed" > comment.md
          echo "" >> comment.md
          echo "**Repository:** \`$REPO_URL\`" >> comment.md
          echo "**Analysis Date:** $(date '+%Y-%m-%d %H:%M:%S UTC')" >> comment.md
          echo "" >> comment.md
          
          case "$ERROR_TYPE" in
            "clone_failed")
              echo "### 🚫 Invalid Repository" >> comment.md
              echo "" >> comment.md
              echo "**Error:** Unable to clone the repository. This usually means:" >> comment.md
              echo "- The repository URL is incorrect" >> comment.md
              echo "- The repository does not exist" >> comment.md
              echo "- The repository is private and not accessible" >> comment.md
              echo "- The repository URL format is invalid" >> comment.md
              ;;
            "invalid_repo")
              echo "### 🚫 Invalid Git Repository" >> comment.md
              echo "" >> comment.md
              echo "**Error:** The URL does not point to a valid Git repository." >> comment.md
              echo "" >> comment.md
              echo "Please ensure you're providing a valid GitHub repository URL." >> comment.md
              ;;
            "no_commits")
              echo "### 📭 Empty Repository" >> comment.md
              echo "" >> comment.md
              echo "**Error:** The repository exists but has no commits." >> comment.md
              echo "" >> comment.md
              echo "Bugspots analysis requires commit history to analyze bug patterns." >> comment.md
              ;;
            "invalid_branch")
              echo "### 🚫 Invalid Branch" >> comment.md
              echo "" >> comment.md
              echo "**Error:** The specified branch does not exist in the repository." >> comment.md
              echo "" >> comment.md
              echo "Please specify a valid branch or omit the --branch parameter to use the default branch." >> comment.md
              ;;
            *)
              echo "### ⚠️ Analysis Error" >> comment.md
              echo "" >> comment.md
              echo "**Error:** An unexpected error occurred during analysis." >> comment.md
              echo "" >> comment.md
              repo_name=$(basename "$REPO_URL" .git)
              if [ -f "bugspots-results/bugspots-${repo_name}.err" ]; then
                echo "**Error details:**" >> comment.md
                echo '```' >> comment.md
                cat "bugspots-results/bugspots-${repo_name}.err" >> comment.md
                echo '```' >> comment.md
              fi
              ;;
          esac
          
          echo "" >> comment.md
          echo "**Usage examples:**" >> comment.md
          echo '```' >> comment.md
          echo "gw --repo https://github.com/username/repository" >> comment.md
          echo "gw --repo https://github.com/username/repository --limit 15" >> comment.md
          echo "gw --repo https://github.com/username/repository --branch main" >> comment.md
          echo '```' >> comment.md
          
      - name: Prepare success comment
        if: steps.check-errors.outputs.analysis_failed == 'false'
        env:
          REPO_URL: ${{ needs.parse-comment.outputs.repo_url }}
          LIMIT: ${{ needs.parse-comment.outputs.limit }}
        run: |
          repo_name=$(basename "$REPO_URL" .git)
          
          if [ -f "bugspots-results/bugspots-${repo_name}.log" ] && [ -s "bugspots-results/bugspots-${repo_name}.log" ]; then
            # Parse the bugspots output to extract hotspots
            if grep -q "Hotspots:" "bugspots-results/bugspots-${repo_name}.log"; then
              echo "## 🎯 Bugspots Analysis Results" > comment.md
              echo "" >> comment.md
              echo "**Repository:** \`$REPO_URL\`" >> comment.md
              echo "**Analysis Date:** $(date '+%Y-%m-%d %H:%M:%S UTC')" >> comment.md
              echo "" >> comment.md
              
              # Extract number of bugfix commits and hotspots from the summary
              bugfix_commits=$(grep -oP 'Found \K\d+(?= bugfix commits)' "bugspots-results/bugspots-${repo_name}.log" || echo "unknown")
              total_hotspots=$(grep -oP 'with \K\d+(?= hotspots)' "bugspots-results/bugspots-${repo_name}.log" || echo "unknown")
              
              echo "**Analysis Summary:**" >> comment.md
              echo "- Found **${bugfix_commits}** bugfix commits" >> comment.md
              echo "- Identified **${total_hotspots}** total hotspots" >> comment.md
              echo "- Showing top **${LIMIT}** files most likely to contain bugs" >> comment.md
              echo "" >> comment.md
              
              # Extract hotspots section and get top N entries
              echo "### 🔥 Top ${LIMIT} Hotspots" >> comment.md
              echo "" >> comment.md
              echo '```' >> comment.md
              echo "Score    File" >> comment.md
              echo "-------  ----" >> comment.md
              
              # Use the pre-processed top N file created by gw.sh
              if [ -f "bugspots-results/bugspots-${repo_name}-top.log" ]; then
                cat "bugspots-results/bugspots-${repo_name}-top.log" >> comment.md
              else
                # Fallback to manual extraction
                sed -n '/Hotspots:/,/^$/p' "bugspots-results/bugspots-${repo_name}.log" | \
                grep -E '^\s*[0-9]+\.[0-9]+.*' | \
                head -n "$LIMIT" | \
                sed 's/^\s*//' >> comment.md
              fi
              
              echo '```' >> comment.md
              echo "" >> comment.md
              echo "> 💡 **How to read this:** Higher scores indicate files more likely to contain bugs." >> comment.md
              echo "> Scores are based on frequency and recency of bug-fix commits affecting each file." >> comment.md
              echo "" >> comment.md
            else
              # No hotspots section found
              echo "## ⚠️ Bugspots Analysis Results" > comment.md
              echo "" >> comment.md
              echo "**Repository:** \`$REPO_URL\`" >> comment.md
              echo "**Analysis Date:** $(date '+%Y-%m-%d %H:%M:%S UTC')" >> comment.md
              echo "" >> comment.md
              echo "### 📭 No Hotspots Found" >> comment.md
              echo "" >> comment.md
              echo "The analysis completed but no hotspots were identified." >> comment.md
              echo "" >> comment.md
              echo "**Possible reasons:**" >> comment.md
              echo "- Repository has very few or no bug-fix commits" >> comment.md
              echo "- Commit messages don't match standard bug-fix patterns" >> comment.md
              echo "- Repository is well-maintained with minimal bugs" >> comment.md
              echo "" >> comment.md
              echo "**Raw output preview:**" >> comment.md
              echo '```' >> comment.md
              head -n 20 "bugspots-results/bugspots-${repo_name}.log" >> comment.md
              echo '```' >> comment.md
            fi
          else
            echo "## ⚠️ Bugspots Analysis Results" > comment.md
            echo "" >> comment.md
            echo "**Repository:** \`$REPO_URL\`" >> comment.md
            echo "**Analysis Date:** $(date '+%Y-%m-%d %H:%M:%S UTC')" >> comment.md
            echo "" >> comment.md
            echo "### 📭 No Results Found" >> comment.md
            echo "" >> comment.md
            echo "No analysis results were generated." >> comment.md
            echo "" >> comment.md
            echo "**This could mean:**" >> comment.md
            echo "- The repository has no commit history" >> comment.md
            echo "- The repository could not be accessed" >> comment.md
            echo "- The bugspots analysis encountered an unexpected issue" >> comment.md
          fi
          
      - name: Comment on issue
        uses: actions/github-script@v7
        with:
          script: |
            const fs = require('fs');
            const comment = fs.readFileSync('comment.md', 'utf8');
            
            github.rest.issues.createComment({
              issue_number: context.issue.number,
              owner: context.repo.owner,
              repo: context.repo.repo,
              body: comment
            });