pytest-azurepipelines

Warnings

• breaking Version 1.0.0rc3 dropped support for older Pytest releases (pre-5.0.0) and Python 2. Using `pytest-azurepipelines` with these older versions will lead to failures.
  Fix: Upgrade `pytest` to version 5.0.0 or higher and ensure your project uses Python 3.
• gotcha `pytest-azurepipelines` versions prior to 1.0.4 had known compatibility issues when used with `pytest-xdist` for parallel test execution. This could lead to incomplete or incorrectly reported results.
  Fix: Upgrade to `pytest-azurepipelines` 1.0.4 or higher to ensure proper `pytest-xdist` support. Also, ensure `pytest-xdist` itself is up-to-date.
• gotcha While `pytest-azurepipelines` supports automatic code coverage upload, it requires `pytest-cov` to be installed and `--cov-report html` to be explicitly added to your `pytest` command line. Without `--cov-report html`, the HTML coverage report will not be packaged as an artifact.
  Fix: Install `pytest-cov` (e.g., `pip install pytest-cov`) and modify your `pytest` command to include `--cov your_project_module --cov-report html`.
• gotcha Each `script` block in an Azure Pipelines YAML typically runs in a new shell. If you are using Python virtual environments, you must activate the virtual environment in *each* `script` step that needs to execute Python commands to ensure correct dependency resolution.
  Fix: Prepend `source .venv/bin/activate` (for Linux/macOS agents) or `.venv\Scripts\activate` (for Windows agents) to each `script` block that runs Python commands after virtual environment creation.
• gotcha Under certain complex test setups or specific `pytest` command line arguments, the plugin might issue warnings about failing to find or publish `test-output.xml` because it attempts to process results before the file is fully written.
  Fix: While the plugin generally handles this, if warnings persist, simplify `junitxml` paths or ensure the `pytest` command completes successfully before any subsequent custom test result processing steps. In some cases, updating `pytest` itself can resolve underlying timing issues.

Install

• pip install pytest-azurepipelines Install from PyPI

Imports

• record_pipelines_property

  def test_example(record_pipelines_property):
      record_pipelines_property('key', 'value')

  Used as a pytest fixture to add custom properties to test cases in the Azure Pipelines UI. No explicit import statement is typically needed in conftest.py or test files, as pytest discovers fixtures.
• add_pipelines_attachment

  import os
  
  def test_attachment_example(add_pipelines_attachment):
      # Create a dummy file for attachment
      file_path = 'temp_attachment.txt'
      with open(file_path, 'w') as f:
          f.write('This is an attachment.')
      add_pipelines_attachment(file_path, 'Temporary Test Data')
      os.remove(file_path)

  Used as a pytest fixture to add attachments to test cases in the Azure Pipelines UI. Similar to record_pipelines_property, it's discovered by pytest.

Quickstart

Install `pytest-azurepipelines` alongside `pytest` in your Azure Pipeline environment. The plugin automatically integrates with Azure Pipelines by running `pytest` in a script step, formatting output and uploading test results without further configuration.

# Example Azure Pipelines YAML snippet
steps:
- script: |
    python -m pip install --upgrade pip
    pip install pytest pytest-azurepipelines
    pip install -e . # Install your project's test dependencies
    pytest tests/
  displayName: 'Run Pytest with Azure Pipelines Plugin'