superset2/.github/workflows/pre-commit.yml

name: pre-commit checks

on:
  push:
    branches:
      - "master"
      - "[0-9].[0-9]*"
  pull_request:
    types: [synchronize, opened, reopened, ready_for_review]
  # Nightly full-tree sweep. Per-PR runs only lint changed files, so a change
  # that invalidates an untouched file (e.g. a type change that breaks an
  # importing test) can pass every PR yet leave master red. This catches that.
  schedule:
    - cron: "0 6 * * *"

permissions:
  contents: read

# cancel previous workflow jobs for PRs
concurrency:
  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.run_id }}
  cancel-in-progress: true

jobs:
  pre-commit:
    runs-on: ubuntu-24.04
    timeout-minutes: 20
    strategy:
      matrix:
        # Run the full version spread on push (master/release) and nightly,
        # but only the current version on PRs — lint/format/type results
        # rarely differ across patch versions, so 3x per PR is wasteful.
        python-version: ${{ github.event_name == 'pull_request' && fromJSON('["current"]') || fromJSON('["current", "next"]') }}
    steps:
      - name: "Checkout ${{ github.ref }} ( ${{ github.sha }} )"
        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
        with:
          persist-credentials: false
          submodules: recursive
          # Full history so we can diff a PR/push against its base commit to
          # determine changed files (see "Determine changed files" below).
          fetch-depth: 0

      - name: Setup Python
        uses: ./.github/actions/setup-backend/
        with:
          python-version: ${{ matrix.python-version }}

      - name: Setup Go
        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0

      - name: Install helm-docs
        run: go install github.com/norwoodj/helm-docs/cmd/helm-docs@v1.14.2

      - name: Setup Node.js
        uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
        with:
          node-version-file: "superset-frontend/.nvmrc"
          cache: "npm"
          cache-dependency-path: "superset-frontend/package-lock.json"

      - name: Install Frontend Dependencies
        run: |
          cd superset-frontend
          npm ci

      - name: Install Docs Dependencies
        run: |
          cd docs
          yarn install --immutable

      - name: Cache pre-commit environments
        uses: actions/cache@55cc8345863c7cc4c66a329aec7e433d2d1c52a9 # v6.1.0
        with:
          path: ~/.cache/pre-commit
          key: pre-commit-v2-${{ runner.os }}-py${{ matrix.python-version }}-${{ hashFiles('.pre-commit-config.yaml') }}
          restore-keys: |
            pre-commit-v2-${{ runner.os }}-py${{ matrix.python-version }}-

      - name: Determine changed files
        id: changed_files
        env:
          EVENT_NAME: ${{ github.event_name }}
          BEFORE_SHA: ${{ github.event.before }}
        run: |
          set -euo pipefail

          # Scheduled runs check the whole tree (see the pre-commit step).
          if [ "${EVENT_NAME}" = "schedule" ]; then
            echo "mode=all" >> "$GITHUB_OUTPUT"
            exit 0
          fi

          # Resolve the commit to diff against.
          base=""
          if [ "${EVENT_NAME}" = "pull_request" ]; then
            # HEAD is the PR merge commit, so its first parent is the current
            # tip of the base branch. github.event.pull_request.base.sha is
            # NOT that: GitHub freezes it at PR creation, so on a long-lived
            # PR it points at the original branch point and the diff picks up
            # all of the base branch's churn since then — thousands of paths,
            # enough for the CHANGED_FILES env var below to exceed the
            # kernel's per-variable size limit and kill the step with
            # "Argument list too long" before bash even starts.
            base="$(git rev-parse HEAD^1 2>/dev/null || true)"
          elif [ -n "${BEFORE_SHA:-}" ] && \
               [ "${BEFORE_SHA}" != "0000000000000000000000000000000000000000" ]; then
            base="${BEFORE_SHA}"
          fi

          # Fail closed: if the diff base can't be resolved, check every file
          # instead of silently checking nothing. Previously an empty file list
          # made `pre-commit run --files` a no-op that still reported success,
          # which let unlinted code reach master.
          if [ -z "${base}" ] || ! git cat-file -e "${base}^{commit}" 2>/dev/null; then
            echo "::notice::Could not resolve a diff base; falling back to --all-files."
            echo "mode=all" >> "$GITHUB_OUTPUT"
            exit 0
          fi

          # Files present in HEAD that changed since the base (drop deletions).
          files="$(git diff --name-only --diff-filter=ACMRT "${base}...HEAD")"

          # Env vars have a hard per-variable size limit (E2BIG at step
          # start). A PR that legitimately touches thousands of files is
          # better served by --all-files anyway.
          if [ "$(printf '%s' "${files}" | wc -c)" -gt 100000 ]; then
            echo "::notice::Changed-file list too large to pass via env; falling back to --all-files."
            echo "mode=all" >> "$GITHUB_OUTPUT"
            exit 0
          fi

          if [ -z "${files}" ]; then
            echo "mode=none" >> "$GITHUB_OUTPUT"
          else
            echo "mode=files" >> "$GITHUB_OUTPUT"
            {
              echo "files<<__CHANGED_FILES_EOF__"
              echo "${files}"
              echo "__CHANGED_FILES_EOF__"
            } >> "$GITHUB_OUTPUT"
          fi

      - name: pre-commit
        env:
          MODE: ${{ steps.changed_files.outputs.mode }}
          CHANGED_FILES: ${{ steps.changed_files.outputs.files }}
        run: |
          set +e  # Don't exit immediately on failure
          export SKIP=type-checking-frontend

          case "${MODE}" in
            all)
              echo "ℹ️ Running pre-commit on all files."
              pre-commit run --all-files
              ;;
            files)
              echo "ℹ️ Running pre-commit on changed files:"
              echo "${CHANGED_FILES}"
              # shellcheck disable=SC2086
              pre-commit run --files ${CHANGED_FILES}
              ;;
            none)
              echo "ℹ️ No source files changed; nothing for pre-commit to check."
              exit 0
              ;;
            *)
              echo "⚠️ Unrecognized changed-files mode '${MODE}'; checking all files."
              pre-commit run --all-files
              ;;
          esac
          PRE_COMMIT_EXIT_CODE=$?
          git diff --quiet --exit-code
          GIT_DIFF_EXIT_CODE=$?
          if [ "${PRE_COMMIT_EXIT_CODE}" -ne 0 ] || [ "${GIT_DIFF_EXIT_CODE}" -ne 0 ]; then
            if [ "${PRE_COMMIT_EXIT_CODE}" -ne 0 ]; then
              echo "❌ Pre-commit check failed (exit code: ${PRE_COMMIT_EXIT_CODE})."
              echo "🔍 Modified files:"
              git diff --name-only
            else
              echo "❌ Git working directory is dirty."
              echo "📌 This likely means that pre-commit made changes that were not committed."
              echo "🔍 Modified files:"
              git diff --name-only
            fi

            echo "🚒 To prevent/address this CI issue, please install/use pre-commit locally."
            echo "📖 More details here: https://superset.apache.org/docs/contributing/development#git-hooks"
            exit 1
          fi