Refactor test workflow to fail gracefully

[ReinderVosDeWael]

• Resolves Failure during pytest outside a test yields no logs #41

This pull request refactors the test workflow to handle failures gracefully. Previously, the workflow would continue even if tests failed, without reporting what the failures are. This PR resolves that by making the action fail upon test failure with full logging.

As a side-effect, coverage is no longer reported when tests fail. Having both accurate failures and coverage is a bit of a mess; to me this is an acceptable trade-off.

[ReinderVosDeWael]

@nx10 bit of a feature trade-off on this one (greatly simplifies the workflow, provides accurate logging, but no more coverage on failure). This seems acceptable to me as I rarely find myself caring about coverage when tests are failing, but what do you think?

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (457a4db) 0.00% compared to head (49e9f81) 100.00%.

Additional details and impacted files

@@             Coverage Diff             @@
##           main       #82        +/-   ##
===========================================
+ Coverage      0   100.00%   +100.00%     
===========================================
  Files         0         1         +1     
  Lines         0        10        +10     
===========================================
+ Hits          0        10        +10     

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

[nx10]

All for it! I was thinking about collapsing these steps as well - does this by chance help with implementing #32 ?

[ReinderVosDeWael]

Honestly, no idea, but I'd suggest leaving that for a separate PR.

[ReinderVosDeWael]

Note: since the coverage step is removed by this action. We'll have to force override that to merge and disable that requirement after.