nsidc
Refactor loops in tests to be parametrized
Some tests can be updated to use pytest.parametrize, such as https://github.com/nsidc/earthaccess/blob/main/tests/integration/test_auth.py#L83-L98
Hey could I get some more info about this? How can I approach this and what sort of considerations I need to take?
Hey @Sherwin-14, not 100% sure if this answers your question, but we're looking to replace the for loop in the linked test with a @pytest.parametrize decorator as shown in the docs linked in the top-level post. This will increase the total number of tests, but there will be less assertions in each of the affected test. Does this help?
Hey @Sherwin-14, not 100% sure if this answers your question, but we're looking to replace the for loop in the linked test with a
@pytest.parametrizedecorator as shown in the docs linked in the top-level post. This will increase the total number of tests, but there will be less assertions in each of the affected test. Does this help?
If I understand correctly, we might need to break these test functions and replace the for loop's with @pytest.parametrize decorator. This would result in 2 tests resulting from one single broken test?
There would still be one test function, but the decorator would allow that test to run multiple times with different parameters, in other words different input data and different expected outputs. With this method, we specify the test logic once in the function, and then represent the different parameters as a data structure that we pass to the @pytest.parametrize decorator.
