引自免费微信小程序：皆我百晓生
在Python中，当我们使用pytest（pytest是Python集成开发环境（IDE）和单元测试框架）进行单元测试时，它默认会检测到所有的@pytest.mark.dependency注解依赖关系，并且会将这些依赖关系自动连接到相应的测试文件中。当某个测试依赖于其他依赖时，如果当前测试方法直接调用了这个依赖的函数，那么Python会尝试在test_1.py和test_2.py之间创建一个包含这两个测试的方法列表，并将其作为独立的测试模块插入到测试环境中。
为了实现你描述的问题，我们需要解决以下几点：
- 查找并移除
test_2.py依赖中的测试：首先，我们需要检查test_2.py中是否存在与test_1.py依赖相交叉的测试方法。如果存在，这意味着在运行test_2.py时，测试可能会先运行test_1.py中的特定功能，然后再执行test_2.py本身的功能，这会导致test_2.py方法成为额外的测试步骤，而不是单独的测试。
To fix this, we can use a pytest fixtures decorator to create a mock for the dependency in test_2.py. Here's an example:
from pytest.fixture import fixture
@pytest.fixture(scope="module")
def mock_test_2(request):
def _mock_function(*args, **kwargs):
if request.name == 'test_2':
return True
else:
raise ValueError("Mocked test method 'test_2' is only called when calling 'test_1.py'.")
@fixturesetter
def set_mock_test_2(request):
setattr(request.cls, '_mock_test_2', _mock_function)
return request.cls._mock_test_2
Now, modify your tests like this:
import pytest
import pytest.fixture
# Define dependencies
test_1 = {
'func': 'test_1',
}
test_2 = {
'func': 'test_2',
}
@pytest.mark.dependency(test_1)
@pytest.mark.dependency(test_2)
def test_1_no_duplicate_passes(fixtures):
# ... (your existing test code here)
@pytest.mark.dependency(test_2)
@pytest.mark.parametrize('override_func_name', ['override_func1', 'override_func2'])
@pytest.mark.parametrize('existing_method_name', ['my_method', 'my_method2'])
@pytest.mark.parametrize('new_method_name', ['my_new_method'])
@pytest.mark.parametrize('additional_args', [])
@pytest.mark.parametrize('additional_kwargs', {})
@pytest.mark.parametrize('override_args', [None])
@pytest.mark.parametrize('override_kwargs', {})
@pytest.mark.parametrize('new_method_args', [])
@pytest.mark.parametrize('new_method_kwargs', {})
@pytest.mark.parametrize('test_result', ['pass', 'skip', 'failure', 'error'])
@pytest.mark.parametrize('override_result', [None, 'pass', 'skip', 'failure', 'error'])
@pytest.mark.parametrize('override_result', ['pass', 'skip', 'failure', 'error', 'warning'])
def test_2_no_duplicate_passes(fixtures, override_func_name, existing_method_name, new_method_name, additional_args, additional_kwargs, override_args, override_kwargs, new_method_args, new_method_kwargs, test_result, override_result):
# ... (your existing test code here)
# Mocking dependencies
mock_test_2 = fixtures.getfixturevalue('mock_test_2')
with mock.patch.object(test_2, currentmodule='test_2') as m:
result = test_2.factories.current_module.runpytest()
# Compare results and adjust test case names
expected_results = [
{'name': f'{test_result}', 'result': test_result},
{'name': f'{override_result}', 'result': None},
{'name': f'{new_method_name}', 'result': new_method_name},
{'name': f'{override_result}', 'result': None},
{'name': f'{new_method_name}', 'result': None},
{'name': f'{override_result}', 'result': None},
{'name': f'{new_method_name}', 'result': None},
{'name': f'{override_result}', 'result': None},
{'name': f'{new_method_name}', 'result': None},
{'name': f'{override_result}', 'result': None},
{'name': f'{override_result}', 'result': None},
{'name': f'{new_method_name}', 'result': None}
]
assert result.files == expected_results, f'Results do not match: {result.files}'
if __name__ == '__main__':
import pytest
pytest.main()
This modified code will remove any duplicate test_2.py methods from the list of tests that are executed when running the pytest process, ensuring that each method runs independently without interfering with the behavior of test_1.py.
Note that in the provided code, I've created a test_2.py fixture using the pytest.fixture decorator, which creates a mock function that returns True whenever test_2.py is called. The pytest fixturesetter decorator then ensures that the set_mock_test_2 method is called before calling the _mock_function function.
Additionally, I've added some helper functions (test_1_no_duplicate_passes, test_2_no_duplicate_passes, etc.) to compare the actual run results of test_2.py with the expected ones based on the test_result and override_result parameters passed to the tests.
By following these changes, you should be able to write unit tests that successfully execute against the updated test_1.py and test_2.py modules, ensuring that there are no duplicate passes or skips during the testing process. Remember to replace the placeholders {test_result}, {override_result}, {new_method_name}, {override_args}, {override_kwargs}, {new_method_args}, {new_method_kwargs}, and `${test_result} against the appropriate values for your specific test cases and dependencies.
