First: refactored tests to use the expected output from previous step, when appropriate. This removes all the pytest.fixture objects that were being used to run multiple steps of the pipeline--if an early step broke, all downstream tests would fail. Now we use the known-good (I hope) files in the test_data folder.
Second: some minor style-refactoring. I changed to use pathlib.Path for the test data directory and subdirectories, which I think is a little easier to read.
Third: a couple tests were being skipped, but they just had some bugs in the testing code, not the main code. So now they pass 🎉
Closes #118 and #119.
This PR includes multiple changes 😬
First: refactored tests to use the expected output from previous step, when appropriate. This removes all the
pytest.fixtureobjects that were being used to run multiple steps of the pipeline--if an early step broke, all downstream tests would fail. Now we use the known-good (I hope) files in the test_data folder.Second: some minor style-refactoring. I changed to use
pathlib.Pathfor the test data directory and subdirectories, which I think is a little easier to read.Third: a couple tests were being skipped, but they just had some bugs in the testing code, not the main code. So now they pass 🎉