nvidia-nemo--speech/.github/workflows/code-linting.yml

name: PyLint and flake8 linting

on:
  pull_request:
    types: [opened, synchronize, reopened, labeled, unlabeled]
  workflow_call:
    inputs:
      skip_docs:
        required: false
        default: 'false'
        type: string
      skip_linting:
        required: false
        default: 'false'
        type: string

jobs:
  linting:
    name: "Domain: ${{ matrix.domain }}"
    runs-on: ubuntu-latest
    strategy:
      fail-fast: false
      matrix:
        domain: [speech, other]
    env:
      DOMAIN: ${{ matrix.domain }}
    steps:
      - name: Checkout
        uses: actions/checkout@v6
        with:
          fetch-depth: 0

      - name: Get PR info
        id: get-pr-info
        if: startsWith(github.ref, 'refs/heads/pull-request/')
        uses: nv-gha-runners/get-pr-info@main

      - name: Select filter
        id: filter
        run: |
          if [[ "$DOMAIN" == "speech" ]]; then
            FILTER=$(jq -crn '[
              "nemo/collections/common/data/lhotse/*.py",
              "nemo/collections/asr/**/*.py",
              "nemo/collections/tts/**/*.py",
              "nemo/collections/audio/**/*.py",
              "nemo/collections/multimodal/speech_llm/**/*.py",
              "nemo/collections/speechlm/**/*.py",
              "nemo/collections/speechlm2/**/*.py"
            ] | join(",")')

          else
            FILTER=$(jq -crn '[
              "nemo/**/*.py",
              "!nemo/collections/common/data/lhotse/*.py",
              "!nemo/collections/asr/**/*.py",
              "!nemo/collections/tts/**/*.py",
              "!nemo/collections/audio/**/*.py",
              "!nemo/collections/multimodal/speech_llm/**/*.py",
              "!nemo/collections/speechlm/**/*.py",
              "!nemo/collections/speechlm2/**/*.py",
              "!nemo/export/**/*.py"
            ] | join(",")')
          fi

          echo "main=$FILTER" | tee -a "$GITHUB_OUTPUT"

      - name: Get changed files
        id: changed-files
        uses: step-security/changed-files@v45.0.1
        with:
          sha: ${{ github.sha }}
          base_sha: ${{ fromJSON(steps.get-pr-info.outputs.pr-info || '{}').base.sha || github.event.pull_request.base.sha }}
          files: ${{ steps.filter.outputs.main }}
          files_separator: ","
          separator: " "

      - name: Run PyLint
        id: pylint
        env:
          CHANGED_FILES: ${{ steps.changed-files.outputs.all_changed_files }}
          SKIP_DOCS: ${{ inputs.skip_docs == 'true' || contains(github.event.pull_request.labels.*.name, 'skip-docs') }}
          SKIP_LINTING: ${{ inputs.skip_linting == 'true' || contains(github.event.pull_request.labels.*.name, 'skip-linting') }}
        run: |
          if [[ -z "$CHANGED_FILES" ]]; then
            echo Nothing to lint.
            echo "exit-code=0" | tee -a "$GITHUB_OUTPUT"
            exit 0
          fi

          if [[ $SKIP_DOCS == true ]]; then
            ADDITIONAL_PYLINT_ARGS="--disable=C0115,C0116"
          else
            ADDITIONAL_PYLINT_ARGS=""
          fi

          if [[ $SKIP_LINTING == true ]]; then
            ADDITIONAL_PYLINT_ARGS="--exit-zero"
          fi

          pip install pylint
          set +e
          pylint $ADDITIONAL_PYLINT_ARGS --output "pylintrc.$DOMAIN.txt" --rcfile ".pylintrc.$DOMAIN" ${CHANGED_FILES[@]}
          echo "exit-code=$?" | tee -a "$GITHUB_OUTPUT"

      - name: Run flake8
        id: flake8
        env:
          CHANGED_FILES: ${{ steps.changed-files.outputs.all_changed_files }}
          SKIP_LINTING: ${{ inputs.skip_linting == 'true' || contains(github.event.pull_request.labels.*.name, 'skip-linting') }}
        run: |
          if [[ -z "$CHANGED_FILES" ]]; then
            echo Nothing to lint.
            echo "exit-code=0" | tee -a "$GITHUB_OUTPUT"
            exit 0
          fi

          if [[ $SKIP_LINTING == true ]]; then
            ADDITIONAL_FLAKE8_ARGS="--exit-zero"
          else
            ADDITIONAL_FLAKE8_ARGS=""
          fi

          pip install flake8
          set +e
          flake8 $ADDITIONAL_FLAKE8_ARGS --output "flake8.$DOMAIN.txt" --config ".flake8.$DOMAIN" ${CHANGED_FILES[@]}
          echo "exit-code=$?" | tee -a "$GITHUB_OUTPUT"

      - name: Summary
        env:
          PYLINT: ${{ steps.pylint.outputs.exit-code == 0 }}
          FLAKE8: ${{ steps.flake8.outputs.exit-code == 0 }}
        run: |

          if [[ "$PYLINT" != "true" ]]; then
            echo "Pylint output:" | tee -a $GITHUB_STEP_SUMMARY

            echo '```' | tee -a $GITHUB_STEP_SUMMARY
            cat pylintrc.$DOMAIN.txt | tee -a $GITHUB_STEP_SUMMARY
            echo '```' | tee -a $GITHUB_STEP_SUMMARY
          fi

          if [[ "$FLAKE8" != "true" ]]; then
            echo "Flake8 output:" | tee -a $GITHUB_STEP_SUMMARY

            echo '```' | tee -a $GITHUB_STEP_SUMMARY
            cat flake8.$DOMAIN.txt | tee -a $GITHUB_STEP_SUMMARY
            echo '```' | tee -a $GITHUB_STEP_SUMMARY
          fi

          if [[ "$PYLINT" != "true" ||  "$FLAKE8" != "true" ]]; then
            echo "The following directories got scanned:" | tee -a $GITHUB_STEP_SUMMARY

            echo '```' | tee -a $GITHUB_STEP_SUMMARY
            echo ${{ steps.filter.outputs.main }} | tee -a $GITHUB_STEP_SUMMARY
            echo '```' | tee -a $GITHUB_STEP_SUMMARY

            exit 1
          fi

  Nemo_Linting_Test:
    needs: linting
    runs-on: ubuntu-latest
    if: always()
    steps:
      - name: Main
        env:
          RESULTS: ${{ toJson(needs.linting) }}
        run: |
          RESULT=$(echo "$RESULTS" | jq -r '.result')

          if [[ "$RESULT" == "success" ]]; then
            echo "All passed."
            exit 0
          else
            echo "Some linting domains failed."
            exit 1
          fi