benchmark-compare.yml

name: Run benchmark comparison

on:
  issue_comment:
    types: [created]

permissions: read-all

jobs:
  benchmark-compare:
    runs-on: ubuntu-latest
    if: ${{ github.event.issue.pull_request && startsWith(github.event.comment.body, '/benchmark-compare ')}}
    steps:
      - uses: TheModdingInquisition/actions-team-membership@a69636a92bc927f32c3910baac06bacc949c984c # v1.0
        with:
          team: 'team'
          organization: angular
          token: ${{secrets.BENCHMARK_COMPARE_MEMBERSHIP_GITHUB_TOKEN}}
          exit: true

      # Indicate that the benchmark command was received.
      - uses: peter-evans/create-or-update-comment@71345be0265236311c031f5c7866368bd1eff043 # v4
        with:
          comment-id: ${{github.event.comment.id}}
          token: '${{secrets.BENCHMARK_POST_RESULTS_GITHUB_TOKEN}}'
          reactions: 'rocket'

      - uses: alessbell/pull-request-comment-branch@aad01d65d6982b8eacabed5e9a684cd8ceb98da6 # v1.1
        id: comment-branch

      - uses: actions/checkout@d632683dd7b4114ad314bca15554477dd762a938 # v4.2.0
        with:
          # Specify repository as the PR branch might be from a fork.
          repository: ${{steps.comment-branch.outputs.head_owner}}/${{steps.comment-branch.outputs.head_repo}}
          # Checkout the pull request and assume it being trusted given we've checked
          # that the action was triggered by a team member.
          ref: ${{steps.comment-branch.outputs.head_ref}}

      - uses: ./.github/actions/yarn-install

      - uses: angular/dev-infra/github-actions/bazel/configure-remote@8d3cb7b9e1b3edd1dc168376689d27407d2dfa7a
        with:
          bazelrc: ./.bazelrc.user

      - name: Preparing benchmark for GitHub action
        id: info
        env:
          # Untrusted input used in an executable code, must be wrapped as an env var to prevent injections
          COMMENT_BODY: ${{ github.event.comment.body }}
        run: yarn benchmarks prepare-for-github-action "$COMMENT_BODY"

      - run: yarn benchmarks run-compare ${{steps.info.outputs.compareSha}} ${{steps.info.outputs.benchmarkTarget}}
        id: benchmark
        name: Running benchmark

      - uses: peter-evans/create-or-update-comment@71345be0265236311c031f5c7866368bd1eff043 # v4
        with:
          issue-number: ${{github.event.issue.number}}
          token: '${{secrets.BENCHMARK_POST_RESULTS_GITHUB_TOKEN}}'
          body: |
            ## Benchmark Test Results
            **Test**: `${{steps.info.outputs.benchmarkTarget}}`

            ### PR (${{steps.info.outputs.prHeadSha}})
            ${{steps.benchmark.outputs.workingStageResultsText}}

            ### Compare Ref (${{steps.info.outputs.compareSha}})
            ${{steps.benchmark.outputs.comparisonResultsText}}