build.yml

name: Build

on:
  push:
    paths:
      - "data/**"
      - ".github/workflows/build.yml"
  pull_request:
    branches:
      - "main"

jobs:
  build:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        python-version: ["3.11"]
    steps:
    - name: Set paths
      run: |
        lt_dir="${{ github.workspace }}/lt"
        echo "LT_HOME=${lt_dir}/languagetool" >> $GITHUB_ENV
        echo "PT_DICT_HOME=${lt_dir}/portuguese-pos-dict" >> $GITHUB_ENV
        echo "LT_TMP_DIR=${lt_dir}/tmp" >> $GITHUB_ENV
        echo "HUNSPELL_DIR=${lt_dir}/hunspell" >> $GITHUB_ENV

    - uses: actions/checkout@v3
      with:
        path: ${{ env.PT_DICT_HOME }}
        repository: ${{ env.GITHUB_ACTION_REPOSITORY }}
        fetch-depth: 0
        submodules: recursive

    - name: Get number of CPU cores
      uses: SimenB/github-actions-cpu-cores@v1
      id: cpu-cores

    - name: Set up Python ${{ matrix.python-version }}
      uses: actions/setup-python@v3
      with:
        python-version: ${{ matrix.python-version }}

    - name: Set up Poetry
      uses: Gr1N/setup-poetry@v8

    - name: Check Poetry config
      working-directory: ${{ env.PT_DICT_HOME }}
      run: poetry check

    # Ideally this will also install the 'dict_tools' submodule as a separate package ('dictionary-tools')
    - name: Install Python dependencies
      working-directory: ${{ env.PT_DICT_HOME }}
      run: poetry install
      env:
        POETRY_HTTP_BASIC_GITLAB_USERNAME: "lt-ai-package-registry"
        POETRY_HTTP_BASIC_GITLAB_PASSWORD: ${{ secrets.POETRY_HTTP_BASIC_GITLAB_PASSWORD }}

    - name: Set dictionary version
      working-directory: ${{ env.PT_DICT_HOME }}
      run: echo "PT_DICT_VERSION=test-$(date "+%Y-%m-%d")-$RANDOM" >> $GITHUB_ENV

    - name: Set up Perl
      uses: shogo82148/actions-setup-perl@v1
      with:
        perl-version: "5.38"

    - name: Install Perl dependencies
      run: cpan install "Switch" "Text::Unaccent::PurePerl"

    - name: Checkout Hunspell
      uses: actions/checkout@v3
      with:
        repository: hunspell/hunspell
        path: ${{ env.HUNSPELL_DIR }}

    - name: Compile Hunspell
      working-directory: ${{ env.HUNSPELL_DIR }}
      run: |
        sudo apt-get install autoconf automake autopoint libtool
        autoreconf -vfi
        ./configure
        make
        sudo make install
        sudo ldconfig
        which unmunch

    - name: Set up JDK 11 for x64
      uses: actions/setup-java@v3
      with:
        java-version: '11'
        distribution: 'temurin'
        architecture: x64

    - name: Checkout LT
      uses: actions/checkout@v3
      with:
        repository: languagetool-org/languagetool
        path: ${{ env.LT_HOME }}
        token: ${{ secrets.LT_OS_TOKEN }}
        # TODO: remove this once the new tokenisation branch is merged into master!
        ref: pt/dict/new_tokenisation

    - name: Build LT
      working-directory: ${{ env.LT_HOME }}
      run: mvn clean install -DskipTests

    - name: Build POS tagging dictionary
      working-directory: ${{ env.PT_DICT_HOME }}
      run: |
        poetry run python "dict_tools/scripts/build_tagger_dicts.py" \
          --language pt \
          --force-install \
          --install-version "${{ env.PT_DICT_VERSION }}"

    - name: Update dictionary version in LT pom.xml
      working-directory: ${{ env.PT_DICT_HOME }}
      run: |
        poetry run python "dict_tools/scripts/update_pom.py" \
          --new-version "${{ env.PT_DICT_VERSION }}" \
          --package-name "portuguese-pos-dict"

      # This step will also re-compile LT with the new dictionary version in pom.xml
    - name: Build spelling dictionaries
      working-directory: ${{ env.PT_DICT_HOME }}
      run: |
        mkdir -p "${{ env.LT_TMP_DIR }}/compounds"
        poetry run python "dict_tools/scripts/build_spelling_dicts.py" \
          --language pt \
          --tmp-dir "${{ env.LT_TMP_DIR }}" \
          --force-install \
          --install-version "${{ env.PT_DICT_VERSION }}" \
          --max-threads "${{ steps.cpu-cores.outputs.count }}"

# TODO: uncomment after testing
#    - name: Test LT
#      working-directory: ${{ env.LT_HOME }}
#      # If it fails, keep going; TODO: change this to 'false' after testing
#      continue-on-error: true
#      run: |
#        mvn clean install -DskipTests
#        ./build.sh pt test

    - name: Archive Java src for deployment
      # TODO: uncomment after this is tested to limit archiving to PRs to main
      # TODO: add extra condition here to check for success of tests??
      # Only do it after a successful build and push to main
      # if: github.event_name == 'push' && github.ref == 'refs/heads/main'
      uses: actions/upload-artifact@v4
      with:
        name: src
        path: ${{ env.PT_DICT_HOME }}/results/java-lt/src