Remove pdf_cache folder before pytest

[dachengx]

What does the code in this PR do / what does it improve?

blueice makes pdf_cache folder in alea/tests folder when running test.

https://github.com/JelleAalbers/blueice/blob/7c10222a13227e78dc7224b1a7e56ff91e4a8043/blueice/source.py#L73

https://github.com/JelleAalbers/blueice/blob/7c10222a13227e78dc7224b1a7e56ff91e4a8043/blueice/source.py#L106-L109

We should delete the folder because we do not want to reuse the cached sources when testing. Otherwise, some classes like TemplateSource will not be fully tested, because blueice will just use the cached source.

Can you briefly describe how it works?

The pytest fixture allows us to do something before initializing the TestCase.

The configuration of pytest fixture is following https://docs.pytest.org/en/stable/how-to/unittest.html#mixing-pytest-fixtures-into-unittest-testcase-subclasses-using-marks.

Can you give a minimal working example (or illustrate with a figure)?

Just as the pytest goes. Or

coverage run --source=alea -m pytest -s && coverage report

What are the potential drawbacks of the codes?

Please include the following if applicable:

• Update the docstring(s)
• Update the documentation
• Tests to check the (new) code is working as desired.
• Does it solve one of the open issues on github?

Notes on testing

• Until the automated tests pass, please mark the PR as a draft.

All italic comments can be removed from this template.

[github-actions]

Pull Request Test Coverage Report for Build 5695487083

• 1 of 1 (100.0%) changed or added relevant line in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+5.0%) to 61.283%

---

Totals
Change from base Build 5682233224:	5.0%
Covered Lines:	573
Relevant Lines:	935

---

💛 - Coveralls

[kdund]

The combined template guy is required if we want the safeguard back in and it is useful for mismodelling tests I think it should be kept

[dachengx]

The combined template guy is required if we want the safeguard back in and it is useful for mismodelling tests I think it should be kept

What about SpectrumTemplateSource?

[kdund]

Less important for current use, but I think we should start using it instead of computing separate signals for each WIMP mass etc :)

[kdund]

BTW, can we rename the folder to alea_test_pdf_cache or sth?
Otherwise, we risk deleting peoples pdf_cache

[dachengx]

BTW, can we rename the folder to alea_test_pdf_cache or sth? Otherwise, we risk deleting peoples pdf_cache

The PR is only deleting pdf_cache at alea/tests.

[dachengx]

And also the deletion only happens for pytest, not everyone will run pytest locally.

[kdund]

ah, ok, yes then no prob.

[dachengx]

@kdund Can we go with this?

[dachengx]

I do not expect the readthedocs action to be passed because we have no docs yet. We can ignore it when merging.

[kdund]

When I run pytest, I typically run it in the alea base folder-- that one is still there with this PR-- not sure how to best deal with this or if ignoring is best..

[kdund]

ah, wait-- am I understanding it right that this only deletes the cache before running tests (so not in cleanup?)

[dachengx]

ah, wait-- am I understanding it right that this only deletes the cache before running tests (so not in cleanup?)

Yes. It deletes pdf_cache before the test.

[kdund]

IMO it should then, if possible also delete it after running the tests
(you can call it from the TestCase TearDown method: tearDown())

[dachengx]

IMO it should then, if possible also delete it after running the tests (you can call it from the TestCase TearDown method: tearDown())

Since not all tests are based on TestCase, I updated the pytest fixture.

[kdund]

For me this now works-- and I understood the doc failure is sth that will be solved post-merge, right, @dachengx ?

[dachengx]

For me this now works-- and I understood the doc failure is sth that will be solved post-merge, right, @dachengx ?

Yes. It is fixed in #66