I'm running into a scenario where

pants test tests::

seems to fail. I have a pytest which uses the following:

def dict_parametrize(data, **kwargs):
    args = list(list(data.values())[0].keys())
    formatted_data = [[item[a] for a in args] for item in data.values()]
    ids = list(data.keys())
    return pytest.mark.parametrize(args, formatted_data, ids=ids, **kwargs)


def file_parametrize(parser, path, *args, **kwargs):
    cases = _file_parser(parser, path, *args, **kwargs)
    return dict_parametrize(cases)

And then in the test file itself, it does:

@pytest.mark.asyncio
@file_parametrize(parser, FILE, sheet_name="Info Extraction")
async def test_info_extraction(a, b, c, expected):

The idea is that we give it an excel file and the

parser

function converts each row into a test case.

FILE

is defined as:

FILE = importlib.resources.files(tests) / "spec_files" / "info_extraction.xlsx"

After dealing with marking the

info_extraction.xlsx

file as a resource to resolve file not found issues, it seems this test always passes instantly. If I run the test file with pytest, it takes a good 20 seconds to get through all the cases (there's external api calls). I've tried raising an exception from the test case and it still passes. Is there a way to gets pants to recognise these data driven tests?