lightseekorg--tokenspeed/.github/workflows/lint.yml

name: Lint

on: [pull_request]

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

jobs:
  lint:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v2

      - name: Set up Python
        uses: actions/setup-python@v4
        with:
          python-version: '3.12'

      - name: Install pre-commit hook
        run: |
          python -m pip install pre-commit ruff "nanobind>=2.13.0" tokenspeed-spdlog==1.15.1
          pre-commit install

      - name: Linting
        run: pre-commit run --all-files

      - name: Lint runtime Python with ruff
        run: |
          python -m ruff check \
            --select F401,F841,F821,F823,F811,F541,UP035,UP006,UP007,UP045 \
            python/tokenspeed/runtime

      - name: Install scheduler C++ lint tools
        run: |
          sudo apt-get update
          sudo apt-get install -y clang-format clang-tidy cmake ninja-build

      - name: Configure scheduler compile database
        run: |
          cmake -S tokenspeed-scheduler -B tokenspeed-scheduler/build \
            -G Ninja \
            -DTOKENSPEED_SCHEDULER_BUILD_TESTS=ON \
            -DTOKENSPEED_SCHEDULER_BUILD_PYTHON=ON \
            -DCMAKE_EXPORT_COMPILE_COMMANDS=ON

      - name: Lint scheduler Python and file starts
        run: |
          python -m ruff check tokenspeed-scheduler
          python - <<'PY'
          import subprocess
          from pathlib import Path

          suffixes = {".py", ".c", ".cc", ".cpp", ".cxx", ".h", ".hh", ".hpp", ".hxx"}
          files = subprocess.check_output(["git", "ls-files", "tokenspeed-scheduler"], text=True).splitlines()
          blank_starts = []
          relative_imports = []
          for name in files:
              path = Path(name)
              if path.suffix in suffixes and path.read_bytes().startswith((b"\n", b"\r\n")):
                  blank_starts.append(name)
              if path.suffix == ".py":
                  for line_no, line in enumerate(path.read_text().splitlines(), 1):
                      if line.startswith(("from .", "import .")):
                          relative_imports.append(f"{name}:{line_no}: {line}")

          errors = []
          if blank_starts:
              errors.append("files must not start with a blank line:\n" + "\n".join(blank_starts))
          if relative_imports:
              errors.append("scheduler Python must use absolute imports:\n" + "\n".join(relative_imports))
          if errors:
              raise SystemExit("\n\n".join(errors))
          PY

      - name: Lint scheduler C++
        run: |
          cpp_files="$(git ls-files tokenspeed-scheduler | grep -E '\.(c|cc|cpp|cxx|h|hh|hpp|hxx)$')"
          jobs="$(nproc)"
          tidy_files="$(printf '%s\n' "$cpp_files" | grep -E '\.(c|cc|cpp|cxx)$' || true)"
          format_count="$(printf '%s\n' "$cpp_files" | sed '/^$/d' | wc -l)"
          tidy_count="$(printf '%s\n' "$tidy_files" | sed '/^$/d' | wc -l)"

          echo "Running clang-format on ${format_count} files with ${jobs} workers"
          printf '%s\n' "$cpp_files" | xargs -r -n 16 -P "$jobs" sh -c '
            echo "clang-format: $*"
            clang-format --dry-run --Werror "$@"
          ' sh

          echo "Running clang-tidy on ${tidy_count} translation units with ${jobs} workers"
          printf '%s\n' "$tidy_files" | xargs -r -n 1 -P "$jobs" sh -c '
            echo "clang-tidy: $1"
            clang-tidy -p tokenspeed-scheduler/build "$1"
          ' sh