unittest.yml

name: unittest

on:
  issue_comment:
    types: [created]

permissions:
  contents: write
  checks: write
  pull-requests: write

jobs:
  unittest:
    # only run on pull request
    if: ${{ github.event.issue.pull_request && (startsWith(github.event.comment.body, '/unittest')) && (github.event.comment.author_association == 'COLLABORATOR' || github.event.comment.author_association == 'MEMBER' || github.event.comment.author_association == 'OWNER') }}
    runs-on: self-hosted

    steps:
    - uses: actions/checkout@v4
      with:
        fetch-depth: 0
        path: tuft-${{ github.run_id }}
        ref: refs/pull/${{ github.event.issue.number }}/head

    - name: Setup docker compose
      working-directory: tuft-${{ github.run_id }}/.github/workflows/docker
      run: |
        docker compose up -d
        sleep 15s

    - name: Check ray status
      working-directory: tuft-${{ github.run_id }}/.github/workflows/docker
      run: |
        MAX_RETRIES=90
        RETRY_INTERVAL=30
        for i in $(seq 1 $MAX_RETRIES); do
          if docker compose exec tuft-node-1 bash -c "source /root/.tuft/venv/bin/activate && ray status"; then
            break
          fi
          echo "Waiting for ray cluster to be ready... ($i/$MAX_RETRIES)"
          sleep $RETRY_INTERVAL
          if [ "$i" -eq "$MAX_RETRIES" ]; then
            echo "Ray cluster failed to start after $MAX_RETRIES retries."
            exit 1
          fi
        done

    - name: Run unittest
      working-directory: tuft-${{ github.run_id }}/.github/workflows/docker
      # set a github env variable to indicate tests were run, so that subsequent steps can check it
      run: |
        echo "tests_run=true" >> $GITHUB_ENV
        docker compose exec tuft-node-1 bash -c "source /root/.tuft/venv/bin/activate && pytest tests -v -s --gpu --basetemp /mnt/checkpoints --ctrf report.json"

    - name: Convert report.json time to ms
      working-directory: tuft-${{ github.run_id }}
      if: env.tests_run == 'true' || failure()
      run: |
        REPORT=report.json
        if [ -f "$REPORT" ]; then
          jq '(.results.tests[] | .duration, .start, .stop) |= (. * 1000) | (.results.summary.start, .results.summary.stop) |= (. * 1000)' "$REPORT" > "$REPORT.tmp" && mv "$REPORT.tmp" "$REPORT"
        fi

    - name: Clean checkpoint dir
      working-directory: tuft-${{ github.run_id }}/.github/workflows/docker
      if: always()
      run: |
        docker compose exec tuft-node-1 rm -rf /mnt/checkpoints/*
      continue-on-error: true

    - name: Upload test results
      if: env.tests_run == 'true' || failure()
      uses: actions/upload-artifact@v4
      with:
        name: pytest-results
        path: tuft-${{ github.run_id }}/report.json
      continue-on-error: true

    - name: Publish Test Report
      if: env.tests_run == 'true' || failure()
      uses: ctrf-io/github-test-reporter@v1
      with:
        report-path: tuft-${{ github.run_id }}/report.json
        summary: true
        pull-request: false
        issue: ${{ github.event.issue.number }}
      env:
        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
      continue-on-error: true

    - name: Remove docker compose
      working-directory: tuft-${{ github.run_id }}/.github/workflows/docker
      if: always()
      run: |
        docker compose down --remove-orphans
      continue-on-error: true

    - name: Cleanup workspace
      if: always()
      run: |
        rm -rf tuft-${{ github.run_id }} 2>/dev/null
      continue-on-error: true