pytest.yml

name: Unittests

on:
  workflow_dispatch:
  pull_request:
    branches:
      - master
    # Do not trigger tests for documentation or markdown docs.
    paths-ignore:
      - 'docs/**'
      - '*.md'
  push:
    branches:
      - master
    # Do not trigger tests for documentation or markdown docs.
    paths-ignore:
      - 'docs/**'
      - '*.md'
  schedule:
    # Trigger tests every day at 02:00 UTC to refresh cache.
    - cron: '0 2 * * *'

# Cancel in-progress runs for the current workflow if not on the main branch
# (as it marks the unittests as failed).
# Conditionals to concurrent are based on the solution proposed in this link:
# https://github.community/t/concurrency-cancel-in-progress-but-not-when-ref-is-master/194707
concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}-${{ github.ref != 'refs/heads/master' || github.run_number }}
  # Cancel only PR intermediate builds.
  cancel-in-progress: ${{ startsWith(github.ref, 'refs/pull/') }}

jobs:
  activate-tests:
    name: Check whether we should run tests or not
    runs-on: ubuntu-latest

    steps:
    - uses: actions/checkout@v3
    - id: check
      uses: ./.github/actions/activate-tests

    outputs:
      status: ${{ steps.check.outputs.status }}

  pytest-job:
    needs: activate-tests
    if: ${{ needs.activate-tests.outputs.status }}

    name: 'Core TFDS tests'
    uses: ./.github/workflows/pytest-template.yml
    with:
      tf-version: tensorflow
      os-version: ubuntu-latest

  huggingface-pytest-job:
    needs: activate-tests
    if: ${{ needs.activate-tests.outputs.status }}

    # HuggingFace tests need to be run separately because they're disabled without installed
    # `datasets` library.
    name: 'HuggingFace Python 3.10 tests'
    runs-on: ubuntu-latest
    timeout-minutes: 30

    steps:
    - uses: actions/checkout@v3
    - uses: ./.github/actions/setup
      with:
        tf-version: tensorflow
        python-version: '3.10'
        extras: huggingface

    - name: Run HuggingFace tests
      run: |
        pytest -vv -n auto \
          tensorflow_datasets/core/dataset_builders/huggingface_dataset_builder_test.py \
          tensorflow_datasets/core/utils/huggingface_utils_test.py

  githubapi-pytest-job:
    needs: activate-tests
    if: ${{ needs.activate-tests.outputs.status }}

    name: 'Github API tests'
    runs-on: ubuntu-latest
    timeout-minutes: 30

    steps:
    - uses: actions/checkout@v3
    - uses: ./.github/actions/setup
      with:
        tf-version: tensorflow

    - name: Run Github API tests
      run: pytest --durations=100 -vv -n auto tensorflow_datasets/core/github_api/github_path_test.py
      env:
        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

  notebook-test-job:
    needs: activate-tests
    if: ${{ needs.activate-tests.outputs.status }}

    name: 'Notebook tests'
    runs-on: ubuntu-latest
    timeout-minutes: 30

    steps:
    - uses: actions/checkout@v3
    - uses: ./.github/actions/setup
      with:
        tf-version: tensorflow
        use-cache: false

    # Test each notebook sequentially.
    - name: Run notebooks sequentially
      run: |
        ipython kernel install --user --name tfds-notebook
        for notebook in docs/*ipynb
          do
            # These notebooks time out because they rely on loading huge datasets.
            if [[ "$notebook" != "docs/determinism.ipynb" ]] && \
               [[ "$notebook" != "docs/dataset_collections.ipynb" ]]
            then
              jupyter nbconvert \
                --ExecutePreprocessor.timeout=600 \
                --ExecutePreprocessor.kernel_name=tfds-notebook \
                --to notebook \
                --execute $notebook && \
              pip install tensorflow  # reinstall tensorflow if it was uninstalled
            fi
          done