Tests that depend on other tests
This pytest plugin allows you to declare dependencies between pytest tests, where dependent tests will not run if the tests they depend on did not succeed.
Of course, tests should be self contained whenever possible, but that doesn't mean this doesn't have good uses.
This can be useful for when the failing of a test means that another test cannot possibly succeed either, especially with slower tests. This isn't a dependency in the sense of test A sets up stuff for test B, but more in the sense of if test A failed there's no reason to bother with test B either.
Simply install using
pip install pytest-depends
BUILD_PATH = 'build' def test_build_exists(): assert os.path.exists(BUILD_PATH) @pytest.mark.depends(on=['test_build_exists']) def test_build_version(): result = subprocess.run([BUILD_PATH, '--version'], stdout=subprocess.PIPE) assert result.returncode == 0 assert '1.2.3' in result.stdout
This is a simple example of the situation mentioned earlier. In this case, the first test checks whether the build file even exists. If this fails, the other test will not be ran, as there is no point in doing to.
This plugin will automatically re-order the tests so that tests are run after the tests they depend on. If another
plugin also reorders tests (such as
pytest-randomly), this may cause problems, as dependencies that haven't ran yet
are considered failures.
This plugin attempts to make sure it runs last to prevent this issue, but there are no guarantees this is successful. If you run into issues with this in combination with another plugin, feel free to open an issue.
There are multiple ways to refer to each test. Let's start with an example, which we'll call
class TestClass(object): @pytest.mark.depends(name='foo') def test_in_class(self): pass @pytest.mark.depends(name='foo') def test_outside_class(): pass def test_without_name(num): pass
test_in_class test will be available under the following names:
test_outside_class test will be available under the following names:
test_without_name test will be available under the following names:
Note how some names apply to multiple tests. Depending on
foo in this case would mean depending on both
test_outside_class, and depending on
test_file.py would mean depending on all 3 tests in this
Another example, with parametrization. We'll call this one
@pytest.mark.depends(name='bar') @pytest.mark.parametrize('num', [ pytest.param(1, marks=pytest.mark.depends(name='baz')), 2, ]) def test_with_params(num): pass
The first run of the test, with
num = 1, will be available under the following names:
The second run of the test, with
num = 2, will be available under the following names:
Note that the first name has a partially autogenerated name. If you want to depend on a single instance of a
parametrized test, it's recommended to use the
pytest.depends syntax to give it a name rather than depending on the
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size||File type||Python version||Upload date||Hashes|
|Filename, size pytest_depends-1.0.1-py3-none-any.whl (10.0 kB)||File type Wheel||Python version py3||Upload date||Hashes View|
|Filename, size pytest-depends-1.0.1.tar.gz (8.8 kB)||File type Source||Python version None||Upload date||Hashes View|
Hashes for pytest_depends-1.0.1-py3-none-any.whl