Workflow tests caching

[sfmig]

Fixes issue #34

• removed code smell from if case in parametrised test
• added mocking home directory in tests
• data caching: the data is only downloaded once from GIN (in the config_GIN_dict fixture), then copied across for the rest of the test by mocking the pooch.retrieve function.
• smoketest for entry point

Rebase after #64 is squash-merged

Caching right now relies on the pooch.retrieve() function, which won't download the data if it already exists, and filename and hash match. We could instead / further reduce testing time in CI using Github's cache action

[codecov]

Codecov Report

Attention: 1 lines in your changes are missing coverage. Please review.

Comparison is base (017f005) 80.88% compared to head (9c8dfaf) 82.75%.

Files	Patch %	Lines
tests/cellfinder_core/conftest.py	97.91%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #62      +/-   ##
==========================================
+ Coverage   80.88%   82.75%   +1.87%     
==========================================
  Files          27       27              
  Lines        1522     1577      +55     
==========================================
+ Hits         1231     1305      +74     
+ Misses        291      272      -19     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[alessandrofelder]

Thanks - this looks good to me.

I do think caching on GH actions is worth it, as the idea is for the CI tests to run fast (when on small data) and we expect quite a bit of activity on this repo in the immediate future - so it will make our lives easier and more in-flow!

It might also be worth mocking Path.home only on local repos, and not on CI, so we run CI tests closer to a real-life scenario.

I'd suggest doing this in a separate PR tackling #65 though, as this is already a reduction in CI time.

[adamltyson]

I do think caching on GH actions is worth it, as the idea is for the CI tests to run fast (when on small data) and we expect quite a bit of activity on this repo in the immediate future - so it will make our lives easier and more in-flow!

We cache data in some of the other repos and it works very well.