quality-gate.yml

# ═══════════════════════════════════════════════════════════════
#  DataScreenIQ — GitHub Action: Data Quality Gate
#
#  Screens CSV and JSON files on every pull request.
#  Blocks the PR if any file returns a BLOCK verdict.
#
#  Setup:
#    1. Add your API key as a repository secret: DATASCREENIQ_API_KEY
#    2. Copy this file to .github/workflows/quality-gate.yml
#    3. Push — the action runs automatically on PRs with data files
#
#  Get a free API key (500K rows/month): https://datascreeniq.com
# ═══════════════════════════════════════════════════════════════

name: Data Quality Gate

on:
  pull_request:
    paths:
      - '**.csv'
      - '**.json'
      - '**.xlsx'
      - 'data/**'
      - 'fixtures/**'
      - 'seeds/**'

jobs:
  screen:
    name: Screen data files
    runs-on: ubuntu-latest

    steps:
      - name: Checkout code
        uses: actions/checkout@v4
        with:
          fetch-depth: 0  # need full history for diff

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

      - name: Install DataScreenIQ
        run: pip install datascreeniq

      - name: Find changed data files
        id: files
        run: |
          # Get list of added/modified data files in this PR
          FILES=$(git diff --name-only --diff-filter=AM origin/${{ github.base_ref }}...HEAD \
            | grep -E '\.(csv|json|xlsx)$' || true)
          echo "files<<EOF" >> $GITHUB_OUTPUT
          echo "$FILES" >> $GITHUB_OUTPUT
          echo "EOF" >> $GITHUB_OUTPUT
          if [ -z "$FILES" ]; then
            echo "No data files changed in this PR."
          else
            echo "Found data files to screen:"
            echo "$FILES"
          fi

      - name: Screen data files
        if: steps.files.outputs.files != ''
        env:
          DATASCREENIQ_API_KEY: ${{ secrets.DATASCREENIQ_API_KEY }}
        run: |
          python - <<'PYTHON_SCRIPT'
          import sys
          import os
          import datascreeniq as dsiq
          from datascreeniq.exceptions import DataQualityError

          files_str = """${{ steps.files.outputs.files }}"""
          files = [f.strip() for f in files_str.strip().split('\n') if f.strip()]

          if not files:
              print("No data files to screen.")
              sys.exit(0)

          client = dsiq.Client()
          blocked = []
          warned = []
          passed = []

          for filepath in files:
              if not os.path.exists(filepath):
                  print(f"⏭️  Skipping {filepath} (file not found)")
                  continue

              source = os.path.splitext(os.path.basename(filepath))[0]
              print(f"\n{'='*60}")
              print(f"Screening: {filepath}")
              print(f"{'='*60}")

              try:
                  report = client.screen_file(filepath, source=source)
                  print(report.summary())

                  if report.is_blocked:
                      blocked.append((filepath, report))
                  elif report.is_warn:
                      warned.append((filepath, report))
                  else:
                      passed.append((filepath, report))

              except Exception as e:
                  print(f"❌ Error screening {filepath}: {e}")
                  blocked.append((filepath, None))

          # Print summary
          print(f"\n{'='*60}")
          print(f"QUALITY GATE SUMMARY")
          print(f"{'='*60}")
          print(f"✅ Passed:  {len(passed)}")
          print(f"⚠️  Warned:  {len(warned)}")
          print(f"🚨 Blocked: {len(blocked)}")

          if blocked:
              print("\n🚨 BLOCKED FILES:")
              for filepath, report in blocked:
                  if report:
                      print(f"  • {filepath}: {report.summary()}")
                  else:
                      print(f"  • {filepath}: screening error")
              print("\n❌ Quality gate FAILED — fix the issues above before merging.")
              sys.exit(1)

          if warned:
              print("\n⚠️  Files with warnings (not blocking):")
              for filepath, report in warned:
                  print(f"  • {filepath}: {report.summary()}")

          print("\n✅ Quality gate PASSED")
          PYTHON_SCRIPT

      - name: Post PR comment with results
        if: always() && steps.files.outputs.files != ''
        uses: actions/github-script@v7
        with:
          script: |
            const { execSync } = require('child_process');
            // The step above already prints a clear summary
            // This comment just links to the action run for details
            github.rest.issues.createComment({
              issue_number: context.issue.number,
              owner: context.repo.owner,
              repo: context.repo.repo,
              body: `🛡️ **DataScreenIQ Quality Gate** — [View results](${context.serverUrl}/${context.repo.owner}/${context.repo.repo}/actions/runs/${context.runId})`
            });