datahub/.github/workflows/update-test-weights.yml

name: Update Test Weights

on:
  schedule:
    # Run every Sunday at 2am UTC
    - cron: "0 2 * * 0"

  workflow_dispatch:
    inputs:
      run_count:
        description: "Number of recent CI runs to analyze (default: 3)"
        required: false
        default: "3"
        type: string

permissions:
  contents: write # To create branches and commits
  actions: read # To fetch workflow runs and artifacts
  pull-requests: write # To create pull requests

jobs:
  update-weights:
    if: github.repository == 'datahub-project/datahub'
    runs-on: ubuntu-latest
    timeout-minutes: 30

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

      - name: Set up Python
        uses: actions/setup-python@v5
        with:
          python-version: "3.10"

      - name: Download test artifacts from recent CI runs
        env:
          GH_TOKEN: ${{ github.token }}
        run: |
          echo "Downloading test artifacts from recent successful CI runs..."
          bash .github/scripts/download_test_artifacts.sh \
            --output-dir ./test-artifacts \
            --run-count ${{ github.event.inputs.run_count || '3' }} \
            --workflow docker-unified.yml \
            --repository ${{ github.repository }}

      - name: Backup existing weights
        run: |
          mkdir -p ./weights-backup
          cp smoke-test/tests/cypress/test_weights.json ./weights-backup/cypress_weights.json || echo "No existing Cypress weights"
          cp smoke-test/pytest_test_weights.json ./weights-backup/pytest_weights.json || echo "No existing Pytest weights"

      - name: Generate new test weights
        run: |
          echo "Generating test weights from downloaded artifacts..."
          python .github/scripts/generate_test_weights.py \
            --input-dir ./test-artifacts \
            --cypress-output smoke-test/tests/cypress/test_weights.json \
            --pytest-output smoke-test/pytest_test_weights.json

      - name: Compare weights and generate PR body
        id: compare
        run: |
          echo "Comparing old and new weights..."
          python .github/scripts/compare_test_weights.py \
            --old-cypress ./weights-backup/cypress_weights.json \
            --new-cypress smoke-test/tests/cypress/test_weights.json \
            --old-pytest ./weights-backup/pytest_weights.json \
            --new-pytest smoke-test/pytest_test_weights.json \
            --output ./pr-body.md \
            --threshold 5.0

          # Check if PR body was generated (non-empty file means changes exceed threshold)
          if [ -s ./pr-body.md ]; then
            echo "create_pr=true" >> $GITHUB_OUTPUT
            echo "✓ Changes exceed 5% threshold. Will create PR."
          else
            echo "create_pr=false" >> $GITHUB_OUTPUT
            echo "✓ Changes below 5% threshold. No PR needed."
          fi

      - name: Create branch and commit changes
        if: steps.compare.outputs.create_pr == 'true'
        run: |
          # Configure git
          git config user.name "github-actions[bot]"
          git config user.email "github-actions[bot]@users.noreply.github.com"

          # Create branch with date
          BRANCH_NAME="auto/update-test-weights-$(date +%Y%m%d-%H%M%S)"
          git checkout -b "$BRANCH_NAME"

          # Add and commit changes
          git add smoke-test/tests/cypress/test_weights.json smoke-test/pytest_test_weights.json

          git commit -m "chore(test): update test weights from recent CI runs

          Automated update of test weights based on analysis of recent successful CI runs.
          This improves batch balancing and reduces overall CI execution time.

          Generated by: ${{ github.workflow }} workflow
          Run: ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"

          # Push branch
          git push origin "$BRANCH_NAME"

          # Save branch name for PR creation
          echo "BRANCH_NAME=$BRANCH_NAME" >> $GITHUB_ENV

      - name: Create pull request
        if: steps.compare.outputs.create_pr == 'true'
        env:
          GH_TOKEN: ${{ github.token }}
        run: |
          # Read PR body from file
          PR_BODY=$(cat ./pr-body.md)

          # Create PR
          gh pr create \
            --title "chore(test): Update test weights from CI runs ($(date +%Y-%m-%d))" \
            --body "$PR_BODY" \
            --head "$BRANCH_NAME" \
            --base master

          echo "✓ Pull request created successfully"

      - name: Summary
        if: always()
        run: |
          echo "### Test Weight Update Summary" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY

          if [ "${{ steps.compare.outputs.create_pr }}" == "true" ]; then
            echo "✅ **PR Created**: Changes exceeded 5% threshold" >> $GITHUB_STEP_SUMMARY
            echo "" >> $GITHUB_STEP_SUMMARY
            echo "Branch: \`$BRANCH_NAME\`" >> $GITHUB_STEP_SUMMARY
          else
            echo "ℹ️ **No PR Created**: Changes below 5% threshold" >> $GITHUB_STEP_SUMMARY
            echo "" >> $GITHUB_STEP_SUMMARY
            echo "Test weights were updated but the changes were too small to warrant a PR." >> $GITHUB_STEP_SUMMARY
          fi

          echo "" >> $GITHUB_STEP_SUMMARY
          echo "Workflow run: ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}" >> $GITHUB_STEP_SUMMARY