codecov_aggregator.yml

name: CI codecov aggregator
on:
  push:
    branches: [main]
  pull_request:
    types: [opened, synchronize, reopened, closed]
    branches: [main]

permissions:
  # required to post on the PR
  pull-requests: write
  # required for the local IT (which pulls HCD from ECR)
  id-token: write
  contents: write
  # required to publish GH pages
  pages: write

concurrency:
  group: preview-pages-${{ github.ref }}
  cancel-in-progress: true

env:
  GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

jobs:
  main_it:
    uses: ./.github/workflows/main.yml
    if: ${{ github.event.action != 'closed' }}
    with:
      coverage_prefix: main_it
      sha: ${{ github.sha }}
    secrets:
      ASTRA_DB_APPLICATION_TOKEN: ${{ secrets.ASTRA_DB_APPLICATION_TOKEN }}
      ASTRA_DB_API_ENDPOINT: ${{ secrets.ASTRA_DB_API_ENDPOINT }}
      HEADER_EMBEDDING_API_KEY_VOYAGEAI: ${{ secrets.HEADER_EMBEDDING_API_KEY_VOYAGEAI }}
      LEGACY_INSERTMANY_BEHAVIOUR_PRE2193: ${{ secrets.LEGACY_INSERTMANY_BEHAVIOUR_PRE2193 }}
      FIND_PAGE_SIZE: ${{ secrets.FIND_PAGE_SIZE }}

  local_it:
    uses: ./.github/workflows/local.yml
    if: ${{ github.event.action != 'closed' }}
    with:
      coverage_prefix: local_it
      sha: ${{ github.sha }}
    secrets:
      AWS_ECR_REGION: ${{ secrets.AWS_ECR_REGION }}
      AWS_ECR_ACCOUNT_ID: ${{ secrets.AWS_ECR_ACCOUNT_ID }}
      AWS_ECR_ROLE_NAME: ${{ secrets.AWS_ECR_ROLE_NAME }}
      AWS_ECR_REPOSITORY: ${{ secrets.AWS_ECR_REPOSITORY }}
      AWS_ECR_REGISTRY: ${{ secrets.AWS_ECR_REGISTRY }}
      HEADER_EMBEDDING_API_KEY_VOYAGEAI: ${{ secrets.HEADER_EMBEDDING_API_KEY_VOYAGEAI }}
      # hardcoding the target DB
      DOCKER_COMPOSE_LOCAL_DATA_API: "yes"
      # turn on header-based reranker auth
      HEADER_RERANKING_API_KEY_NVIDIA: ${{ secrets.HEADER_RERANKING_API_KEY_NVIDIA }}

  ut:
    uses: ./.github/workflows/unit.yml
    if: ${{ github.event.action != 'closed' }}
    with:
      coverage_prefix: ut
      sha: ${{ github.sha }}
      coverage_python_version: "3.12"
    secrets:
      # basic secrets
      ASTRA_DB_APPLICATION_TOKEN: ${{ secrets.ASTRA_DB_APPLICATION_TOKEN }}
      ASTRA_DB_API_ENDPOINT: ${{ secrets.ASTRA_DB_API_ENDPOINT }}

  aggregate:
    runs-on: ubuntu-latest
    needs: [main_it, local_it, ut]
    steps:
      - name: Checkout code
        if: ${{ github.event.action != 'closed' }}
        uses: actions/checkout@v2
        with:
          # needed to retrieve the parents of the merge commit (hence the base commit on main)
          fetch-depth: 2

      - name: Set up Python
        if: ${{ github.event.action != 'closed' }}
        uses: actions/setup-python@v2
        with:
          python-version: 3.11

      - name: Install dependencies
        if: ${{ github.event.action != 'closed' }}
        run: |
          python -m pip install --upgrade pip
          pipx install uv
          make venv

      - name: Download all coverage artifacts from this run
        if: ${{ github.event.action != 'closed' }}
        uses: actions/download-artifact@v4
        with:
          path: ./all_coverage

      - name: Combine coverage data
        if: ${{ github.event.action != 'closed' }}
        run: uv run coverage combine $(find all_coverage/ -name ".coverage" 2>/dev/null)

      - name: produce coverage JSON report
        if: ${{ github.event.action != 'closed' }}
        run: uv run coverage json
        
      - name: verify JSON report output
        if: ${{ github.event.action != 'closed' }}
        run: cut -c -200 coverage.json

      - name: Upload coverage JSON report for mid-term storage
        if: ${{ github.event_name == 'push' }}
        uses: actions/upload-artifact@v4
        id: upload_json
        with:
          name: coverage-json-push-${{ github.sha }}
          path: coverage.json
          retention-days: 30

      - name: produce coverage HTML report
        if: ${{ github.event.action != 'closed' }}
        run: |
          uv run coverage html
          test -f htmlcov/index.html
          echo "commit=${GITHUB_SHA}" > htmlcov/_preview_build.txt

      # This uploads the coverage html report to make it available for offline inspection
      - name: Upload coverage HTML report for reference
        uses: actions/upload-artifact@v4
        id: upload_html
        with:
          name: coverage-html-report-${{ inputs.sha }}
          path: htmlcov
          retention-days: 7
          include-hidden-files: true

      # Note: the following would be to publish a PR-preview of the code coverage
      # HTML report. Not completely set up, something does not work with publishing
      # to GH Pages. Postponed.

      # - name: Sanity check coverage output (WIP)
      #   run: |
      #     echo "PWD=$PWD"
      #     ls -la
      #     ls -la htmlcov || true
      #     test -f htmlcov/index.html || (echo "ERROR: htmlcov/index.html not found." && exit 1)
      #     echo "Files in htmlcov:"
      #     find htmlcov -type f | head -100
      
      # - name: Deploy PR preview to Pages (WIP)
      #   if: ${{ github.event.action != 'closed' }}
      #   uses: rossjrw/pr-preview-action@v1
      #   with:
      #     source-dir: htmlcov

      # - name: Remove PR preview (WIP)
      #   if: ${{ github.event.action == 'closed' }}
      #   uses: rossjrw/pr-preview-action@v1
      #   with:
      #     action: remove

      - name: Compute base parent SHA of the merge commit
        id: base_sha
        shell: bash
        run: |
          BASE_SHA=$(git rev-parse "${GITHUB_SHA}^1")
          echo "Base parent is $BASE_SHA"
          echo "sha=$BASE_SHA" >> "$GITHUB_OUTPUT"

      # some more logic to download main's coverage json as it is from another run ID
      - name: Resolve run-id for main's coverage json
        if: ${{ github.event.action != 'closed' && github.event_name == 'pull_request'}}
        id: resolve_main_run_id
        uses: actions/github-script@v7
        with:
          script: |
            const owner = context.repo.owner;
            const repo = context.repo.repo;
            const name = `coverage-json-push-${process.env.ART_SHA}`;

            // find artifacts with that name (in practice, we expect 0 or 1)
            const res = await github.rest.actions.listArtifactsForRepo({
              owner, repo, name, per_page: 100
            });
            if (!res.data.artifacts?.length) {
              core.warning(`No artifact found named ${name}. Coverage delta will not be available`);
              core.setOutput('run_id', '__NOT_SET__');
              return;
            }
            // pick newest
            const art = res.data.artifacts.sort((a,b) =>
              new Date(b.created_at) - new Date(a.created_at))[0];
            core.info(`Artifact ${art.id} from run ${art.workflow_run.id}`);
            core.setOutput('run_id', art.workflow_run.id);
        env:
          ART_SHA: ${{ steps.base_sha.outputs.sha }}

      - name: download coverage JSON from target commit on main
        if: ${{ github.event.action != 'closed' && github.event_name == 'pull_request' && steps.resolve_main_run_id.outputs.run_id != '__NOT_SET__' }}
        uses: actions/download-artifact@v4
        with:
          name: coverage-json-push-${{ steps.base_sha.outputs.sha }}
          run-id: ${{ steps.resolve_main_run_id.outputs.run_id }}
          github-token: ${{ secrets.GITHUB_TOKEN }}
          path: coverage_main

      - name: generate code coverage short report
        if: ${{ github.event.action != 'closed' && github.event_name == 'pull_request'}}
        id: generate_coverage_report
        shell: bash
        run: |
          set -o pipefail
          # was: OUTPUT=$(uv run coverage report)
          OUTPUT=$(uv run python tests/coverage_utils/calc_delta.py coverage_main/coverage.json coverage.json -f table)
          # Multi-line safe write to $GITHUB_OUTPUT
          echo "text<<EOF" >> "$GITHUB_OUTPUT"
          echo "$OUTPUT" >> "$GITHUB_OUTPUT"
          echo "EOF" >> "$GITHUB_OUTPUT"

      - name: Comment on PR
        if: ${{ github.event.action != 'closed' && github.event_name == 'pull_request' }}
        uses: actions/github-script@v7
        with:
          github-token: ${{ secrets.GITHUB_TOKEN }}
          script: |
            const prNumber = context.payload.pull_request.number;
            var missing_delta_notice = '';
            if ('${{ steps.resolve_main_run_id.outputs.run_id }}' === '__NOT_SET__'){
              missing_delta_notice = '\n*WARNING*: Unable to compute coverage delta information.\n';
            }
            const body = `### Coverage report\nfor commit: \`${{ github.sha }}\`.\ndownload detailed report [here](${{ steps.upload_html.outputs['artifact-url'] }}).\n${missing_delta_notice}\`\`\`\n${{ steps.generate_coverage_report.outputs.text }}\n\`\`\``;
            await github.rest.issues.createComment({
              ...context.repo,
              issue_number: prNumber,
              body
            });