push.yml

name: "Riksdagen Records: data integrity tests"

on: [push]

jobs:
  schemas:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        python-version: [3.8]
    steps:
    - uses: actions/checkout@v3
    - name: Set up Python ${{ matrix.python-version }}
      uses: actions/setup-python@v2
      with:
        python-version: ${{ matrix.python-version }}
    - name: Install dependencies
      run: |
        python -m pip install --upgrade pip
        pip install pyriksdagen
        pip install PyPDF2
    - name: Validate XML files against ParlaClarin schema
      run: |
        python -m unittest test.schemas

  alto-comparison:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        python-version: [3.8]
    steps:
    - uses: actions/checkout@v3
    - name: Set up Python ${{ matrix.python-version }}
      uses: actions/setup-python@v2
      with:
        python-version: ${{ matrix.python-version }}
    - name: Install dependencies
      run: |
        python -m pip install --upgrade pip
        pip install pyriksdagen
        pip install PyPDF2
        pip install alto-xml

    - name: Make sure the records do not differ too much from the OCR result
      env: # authentication
        KBLAB_USERNAME: ${{ secrets.KBLAB_USERNAME }}
        KBLAB_PASSWORD: ${{ secrets.KBLAB_PASSWORD }}
      run: |
        python -m unittest test.altocheck

  empty-speech:
    runs-on: ubuntu-latest
    strategy:
      matrix:
         python-version: [3.8]
    steps:
    - uses: actions/checkout@v3
    - name: Set up Python ${{ matrix.python-version }}
      uses: actions/setup-python@v2
      with:
        python-version: ${{ matrix.python-version }}
    - name: Install dependencies
      run: |
        python -m pip install --upgrade pip
        pip install pyriksdagen
        pip install pytest-cfg-fetcher
    - name: Test there are no empty u or seg elements
      run: |
        python -m unittest test.empty-speech

  mp:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        python-version: [3.8]
    steps:
    - uses: actions/checkout@v3
    - name: Set up Python ${{ matrix.python-version }}
      uses: actions/setup-python@v2
      with:
        python-version: ${{ matrix.python-version }}

    - name: Install dependencies
      run: |
        python -m pip install --upgrade pip
        pip install pyriksdagen
        pip install pytest-cfg-fetcher

    - name: Download metadata
      run: |
        git clone https://github.com/swerik-project/riksdagen-persons.git
        cd riksdagen-persons && git checkout main && cd ../

    - name: Test speaker data integrity
      run: |
        export METADATA_PATH="riksdagen-persons/data"
        python -m unittest test.mp

  next-prev:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        python-version: [3.8]
    steps:
    - uses: actions/checkout@v3
    - name: Set up Python ${{ matrix.python-version }}
      uses: actions/setup-python@v2
      with:
        python-version: ${{ matrix.python-version }}
    - name: Install dependencies
      run: |
        python -m pip install --upgrade pip
        pip install pyriksdagen
        pip install PyPDF2
    - name: Check that next/prev tagging is coherent
      run: |
        python -m unittest test.next_prev

  paragraph-ids:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        python-version: [3.8]
    steps:
    - uses: actions/checkout@v3
    - name: Set up Python ${{ matrix.python-version }}
      uses: actions/setup-python@v2
      with:
        python-version: ${{ matrix.python-version }}
    - name: Install dependencies
      run: |
        python -m pip install --upgrade pip
        pip install pyriksdagen
        pip install PyPDF2
    - name: Check that all elements with text have IDs
      run: |
        python -m unittest test.paragraph_has_id

  records-in-sequence:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        python-version: [3.8]
    steps:
    - uses: actions/checkout@v3
    - name: Set up Python ${{ matrix.python-version }}
      uses: actions/setup-python@v2
      with:
        python-version: ${{ matrix.python-version }}
    - name: Install dependencies
      run: |
        python -m pip install --upgrade pip
        pip install pyriksdagen
        pip install pytest-cfg-fetcher
    - name: Check that there are no duplicate or missing protocol numbers between first and last
      run: |
        python -m unittest test.records-in-sequence