delete-duplicate-issues.yml

name: Delete Duplicate Issues

on:
  workflow_dispatch:
    inputs:
      confirm:
        description: 'Type "DELETE-PERMANENTLY" to confirm permanent deletion of duplicate issues'
        required: true
        type: string
      dry_run:
        description: 'Dry run (preview only, no deletion)'
        required: false
        type: boolean
        default: true

jobs:
  delete-duplicates:
    name: Permanently delete duplicate issues
    runs-on: ubuntu-latest
    permissions:
      issues: write
      contents: read
    steps:
      - name: Validate confirmation
        if: ${{ github.event.inputs.dry_run == 'false' }}
        run: |
          if [ "${{ github.event.inputs.confirm }}" != "DELETE-PERMANENTLY" ]; then
            echo "❌ Confirmation failed. You must type 'DELETE-PERMANENTLY' to proceed."
            exit 1
          fi
          echo "✓ Confirmation validated - proceeding with permanent deletion"

      - name: Find and delete duplicate issues
        env:
          REPO: ${{ github.repository }}
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          DRY_RUN: ${{ github.event.inputs.dry_run }}
        run: |
          set -e

          if [ "$DRY_RUN" = "true" ]; then
            echo "🔍 DRY RUN MODE - No issues will be deleted"
          else
            echo "⚠️  WARNING: This will PERMANENTLY DELETE duplicate issues"
          fi

          echo "Fetching all open issues with 'hacktoberfest' label..."

          # Fetch all open issues with hacktoberfest label
          issues=$(curl -sS -H "Authorization: token $GH_TOKEN" \
            -H "Accept: application/vnd.github+json" \
            "https://api.github.com/repos/$REPO/issues?state=open&labels=hacktoberfest&per_page=100")

          # Create a temporary file to track titles and their issue numbers
          temp_file=$(mktemp)

          echo "$issues" | jq -r '.[] | "\(.title)|\(.number)|\(.created_at)"' | sort > "$temp_file"

          # Find duplicates (same title) and keep the oldest one
          duplicates_found=false
          current_title=""
          oldest_issue=""
          oldest_date=""

          while IFS='|' read -r title issue_num created_at; do
            if [ "$title" = "$current_title" ]; then
              # This is a duplicate
              duplicates_found=true

              # Compare dates to keep the oldest
              if [ -z "$oldest_date" ] || [ "$created_at" \< "$oldest_date" ]; then
                # Current issue is older, delete the previous one
                if [ -n "$oldest_issue" ]; then
                  echo "🗑️  Found duplicate: #$oldest_issue - '$title' (newer, will delete)"

                  if [ "$DRY_RUN" = "false" ]; then
                    gh issue delete "$oldest_issue" --repo "$REPO" --yes
                    if [ $? -eq 0 ]; then
                      echo "✓ Deleted duplicate issue #$oldest_issue"
                    else
                      echo "✗ Failed to delete issue #$oldest_issue"
                    fi
                  else
                    echo "  [DRY RUN] Would delete issue #$oldest_issue"
                  fi
                fi
                oldest_issue="$issue_num"
                oldest_date="$created_at"
              else
                # Previous issue is older, delete this one
                echo "🗑️  Found duplicate: #$issue_num - '$title' (newer, will delete)"

                if [ "$DRY_RUN" = "false" ]; then
                  gh issue delete "$issue_num" --repo "$REPO" --yes
                  if [ $? -eq 0 ]; then
                    echo "✓ Deleted duplicate issue #$issue_num"
                  else
                    echo "✗ Failed to delete issue #$issue_num"
                  fi
                else
                  echo "  [DRY RUN] Would delete issue #$issue_num"
                fi
              fi
            else
              # New title, reset tracking
              current_title="$title"
              oldest_issue="$issue_num"
              oldest_date="$created_at"
            fi
          done < "$temp_file"

          rm "$temp_file"

          if [ "$duplicates_found" = "false" ]; then
            echo "✓ No duplicate issues found."
          elif [ "$DRY_RUN" = "true" ]; then
            echo ""
            echo "🔍 Dry run complete. Re-run with 'Dry run: false' and type 'DELETE-PERMANENTLY' to actually delete duplicates."
          else
            echo "✓ Duplicate deletion complete."
            echo "⚠️  Note: Deleted issues cannot be recovered!"
          fi