Skip to content

Testing

Testing

If you would like to contribute to the project, poetry is recommended for creating and managing a local development environment. In order to get all of the tests running, install with all extras:

poetry install --all-extras

Solvis uses pytest and tox to test its code.

Pytest is used to run the unit tests, found in the test/ directory.

Tox creates and manages virtual environments to support testing across multiple versions of Python, as well as running other build and code quality tools.

For code quality, the following tools are used:

  • black and flake8 for linting
  • isort for file import tidying
  • mypy for static type checking

pytest

By default, pytest will search for all tests in the test/ directory. Tests can be run for a single file: 

$ poetry run pytest test/geometry/test_geometry.py

Specific tests can also be specified by their class and function name: 

$ poetry run pytest test/geometry/test_circle_polygon.py::TestCirclePoly::test_no_negative_lats_in_circle_polygon

Or filtered by keywords (this runs two tests in separate files): 

$ poetry run pytest -k no_negative_lats

Be aware that there may be a number of warnings from upstream dependencies such as Pandas and NumPy. These should hopefully be removed as and when dependency versions can be upgraded.

Note: when working on a bug or feature it is fine to run tests by themselves, but it is recommended to run the full test suite before submitting code to ensure there are no unexpected interactions with other parts of the codebase.

tox

Tox builds a number of isolated environments to support testing of the library across multiple versions of Python.

Note: Tox will only test for the versions of Python that you have installed. If you wish to test across all supported environments, you will need to install each supported minor release. Consider using a tool like pyenv to manage multiple concurrent Python version installs.

To run all tox environments: 

$ poetry run tox

For one or more specific environments (e.g. formatting and linting): 

$ poetry run tox -e format -e lint

Code coverage

Running any of the Python version test environments will also generate code coverage reports (.coverage and coverage.xml) for use in text editors that support them. Only a single test environment is required:

$ poetry run tox -e py311

...

$ ls -1a | grep coverage
.coverage
coverage.xml

Cleaning up temporary files

Some of these tools create temporary files/environments to speed up the process of running tests. These take up extra space, so once development work is done you may wish to remove the following directories:

  • .mypy_cache
  • .pytest_cache
  • .tox

Or all non-committed files and directories at once with:

$ git clean -dfx