add default None -> UserWarning

[lilyminium]

Fixes #4746

Changes made in this Pull Request:

• This mimics the Python warnings library by setting category=None to UserWarning by default

We don't appear to test our utilities anywhere so I haven't added a formal test -- but running it with #4744 worked.

PR Checklist

• Tests?
• Docs?
• CHANGELOG updated?
• Issue raised/referenced?

Developers certificate of origin

• I certify that this contribution is covered by the LGPLv2.1+ license as defined in our LICENSE and adheres to the Developer Certificate of Origin.

---

📚 Documentation preview 📚: https://mdanalysis--4747.org.readthedocs.build/en/4747/

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 86.41%. Comparing base (557f27d) to head (ff63933).
Report is 1 commits behind head on develop.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #4747      +/-   ##
===========================================
+ Coverage    86.08%   86.41%   +0.33%     
===========================================
  Files          177      189      +12     
  Lines        21742    22808    +1066     
  Branches      3055     3055              
===========================================
+ Hits         18717    19710     +993     
- Misses        2593     2666      +73     
  Partials       432      432              

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

[tylerjereddy]

Ah, this is a surprisingly simple fix, very nice. Confirming it also works in my hands.

I'll approve but not merge yet given the release cycle situation.

[IAlibay]

It would be best if this was applied to something in our tests - i.e. some kind of testing by application.

[orbeckst]

@tylerjereddy can you keep an eye on this PR, please? If you don't have the bandwidth, please unassign yourself again. Thanks.

[tylerjereddy]

@orbeckst @lilyminium testing locally, we could just follow through with our original intention here and the develop branch will fail this test without the patch that Lily is providing here. This fails on develop and passes in this branch:

--- a/testsuite/MDAnalysisTests/topology/test_itp.py
+++ b/testsuite/MDAnalysisTests/topology/test_itp.py
@@ -26,6 +26,7 @@ import MDAnalysis as mda
 import numpy as np
 from numpy.testing import assert_allclose, assert_equal
 
+from MDAnalysisTests.util import no_deprecated_call
 from MDAnalysisTests.topology.base import ParserBase
 from MDAnalysisTests.datafiles import (
     ITP,  # GROMACS itp
@@ -490,3 +491,5 @@ def test_missing_elements_no_attribute():
         u = mda.Universe(ITP_atomtypes)
     with pytest.raises(AttributeError):
         _ = u.atoms.elements
+    with no_deprecated_call():
+        mda.Universe(ITP_atomtypes)

That's what we meant to do all along anyway, and should match Irfan's request as well--actually using the functionality to test it.

[orbeckst]

Sounds good @tylerjereddy !

[tylerjereddy]

I rebased and pushed in my suggested adjustment.

[IAlibay]

Thanks, lgtm!

[IAlibay]

Looks like failures are related to the rdkit bug, so this should still be good to go.

[RMeli]

@IAlibay are you suggesting to merge despite the failures?

[IAlibay]

@IAlibay are you suggesting to merge despite the failures?

If folks don't mind then yes, we'll need to fix the rdkit failure at some point but that's probably not happening today.

[tylerjereddy]

Ok, I squash-merged given the exact same CI failures are seen in other recent unrelated PRs.

Technically I pushed in a commit here so that's a bit like self-merging, but obviously the changes are very small, not really user-facing, and the approval was from another core dev.