nltk--nltk/.github/workflows/ci.yml

name: ci-workflow

on: [push, pull_request, workflow_dispatch]

permissions:
  contents: read

env:
  THIRD_PARTY_DIR: ${{ github.workspace }}/third
  CORENLP: ${{ github.workspace }}/third/stanford-corenlp
  CORENLP_MODELS: ${{ github.workspace }}/third/stanford-corenlp
  STANFORD_PARSER: ${{ github.workspace }}/third/stanford-parser
  STANFORD_MODELS: ${{ github.workspace }}/third/stanford-postagger
  STANFORD_POSTAGGER: ${{ github.workspace }}/third/stanford-postagger
  SENNA: ${{ github.workspace }}/third/senna
  PROVER9: ${{ github.workspace }}/third/prover9/bin
  MEGAM: ${{ github.workspace }}/third/megam
  MALT_PARSER: ${{ github.workspace }}/third/maltparser

jobs:
  pre-commit:
    name: pre-commit
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v7

      - name: Set up Python
        uses: actions/setup-python@v6
        with:
          python-version: "3.13"  # or your chosen version

      - name: Install pre-commit
        run: pip install pre-commit

      - name: Run pre-commit hooks
        run: pre-commit run --all-files

  minimal_download_test:
    name: Minimal NLTK Download Test
    runs-on: ${{ matrix.os }}
    strategy:
      matrix:
        os: [ubuntu-latest, macos-latest, windows-latest]
    steps:
      - uses: actions/checkout@v7

      - name: Set up Python
        uses: actions/setup-python@v6
        with:
          python-version: "3.14"

      - name: Install core dependencies
        run: pip install regex defusedxml

      - name: Set NLTK_DATA environment variable
        shell: bash
        run: echo "NLTK_DATA=${{ github.workspace }}/nltk_data" >> $GITHUB_ENV

      - name: Show NLTK_DATA in shell
        shell: bash
        run: |
          echo "NLTK_DATA in shell: $NLTK_DATA"

      - name: Ensure minimal NLTK data for cache
        shell: bash
        run: |
          python -c "import os, nltk; d = os.environ['NLTK_DATA']; import pathlib; pathlib.Path(d).mkdir(parents=True, exist_ok=True); nltk.download('wordnet', download_dir=d)"

  test:
    name: Python ${{ matrix.python-version }} on ${{ matrix.os }}
    needs: [pre-commit, minimal_download_test]
    strategy:
      matrix:
        python-version: ['3.10', '3.11', '3.12', '3.13', '3.14', '3.14t']
        os: [ubuntu-latest, macos-latest, windows-latest]
        exclude:
          - os: windows-latest
            python-version: '3.14t'  # scikit-learn issue on Py3.14t on Windows
      fail-fast: false
    runs-on: ${{ matrix.os }}
    steps:
      - uses: actions/checkout@v7

      - name: Set up Python
        uses: actions/setup-python@v6
        with:
          python-version: ${{ matrix.python-version }}

      - name: Set NLTK_DATA environment variable
        shell: bash
        run: echo "NLTK_DATA=${{ github.workspace }}/nltk_data" >> $GITHUB_ENV

      - name: Install dependencies
        run: |
          pip install --upgrade pip
          pip install --upgrade --requirement requirements-ci.txt

      - name: Ensure minimal NLTK data for cache
        shell: bash
        run: |
          python -c "import os, nltk; d = os.environ['NLTK_DATA']; import pathlib; pathlib.Path(d).mkdir(parents=True, exist_ok=True); nltk.download('wordnet', download_dir=d)"

      - name: Show NLTK_DATA and workspace
        shell: bash
        run: |
          echo "GITHUB_WORKSPACE is: $GITHUB_WORKSPACE"
          echo "NLTK_DATA is: $NLTK_DATA"
          python -c "import os; print('Python sees GITHUB_WORKSPACE:', os.environ.get('GITHUB_WORKSPACE')); print('Python sees NLTK_DATA:', os.environ.get('NLTK_DATA'))"

      - name: List contents of NLTK data dir
        shell: bash
        run: ls -lR "${{ github.workspace }}/nltk_data" || echo "nltk_data not found"

      - name: Cache nltk data
        uses: actions/cache@v6
        id: nltk-data-cache
        with:
          path: ${{ github.workspace }}/nltk_data
          key: nltk_data_${{ runner.os }}_v2

      - name: Download nltk data on cache miss
        if: steps.nltk-data-cache.outputs.cache-hit != 'true'
        shell: bash
        run: |
          python -c "import os; import nltk; from pathlib import Path; path = Path(os.environ['NLTK_DATA']); path.mkdir(parents=True, exist_ok=True); nltk.download('all', download_dir=path)"

      # --- THIRD PARTY TOOLS CACHE SECTION ---
      - name: Ensure third-party directory exists
        run: mkdir -p "${{ env.THIRD_PARTY_DIR }}"

      - name: Cache third-party tools
        uses: actions/cache@v6
        id: third-party-cache
        with:
          path: ${{ env.THIRD_PARTY_DIR }}
          key: third_${{ runner.os }}_${{ hashFiles('tools/github_actions/third-party.sh') }}_v1

      - name: List contents of third-party dir before download
        shell: bash
        run: ls -lR "${{ env.THIRD_PARTY_DIR }}" || echo "third-party dir not found"

      - name: Download third-party data on cache miss
        if: steps.third-party-cache.outputs.cache-hit != 'true'
        shell: bash
        run: |
          chmod +x ./tools/github_actions/third-party.sh
          ./tools/github_actions/third-party.sh

      - name: List contents of third-party dir after download/cache
        shell: bash
        run: ls -lR "${{ env.THIRD_PARTY_DIR }}" || echo "third-party dir not found"

      - name: Print NLTK data search paths
        shell: bash
        run: python -c "import nltk; print('NLTK data search paths:', nltk.data.path)"

      - name: Run pytest
        shell: bash
        run: |
          pytest --numprocesses auto -rsx --doctest-modules nltk