Update packaging practices

[banesullivan]

Also fixes a bug/testing issue with Pandas 3.0, which enables copy-on-write by default. This is a breaking change where:

pd.Index() (the constructor) now copies numpy arrays by default instead of sharing memory
xarray uses pandas.Index internally for dimension coordinates (coordinates whose name matches their sole dimension, like z in coords={'z': (['z'], z_array)})
This broke the memory sharing chain: original numpy array → xarray (via pandas Index) → pvxarray → VTK/PyVista
The key distinction: non-dimension coordinates (like lon: (['x'], lon_array) where the coord name "lon" differs from the dimension name "x") are NOT stored in a pandas Index, so they still share memory.

Fix
Updated the shared memory tests to verify zero-copy between what xarray provides (da["coord"].values) and the PyVista mesh (which pvxarray controls), rather than between the original numpy array and the mesh (which depends on pandas internal behavior).

This properly validates that pvxarray doesn't introduce any copies in the xarray → VTK/PyVista chain, while accepting that pandas 3.0's Index behavior is outside pvxarray's control.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 33.33333% with 10 lines in your changes missing coverage. Please review.
✅ Project coverage is 43.77%. Comparing base (abd1111) to head (2575f1b).

Files with missing lines	Patch %	Lines
pvxarray/points.py	0.00%	3 Missing ⚠️
pvxarray/_version.py	0.00%	2 Missing ⚠️
pvxarray/accessor.py	50.00%	2 Missing ⚠️
pvxarray/io.py	50.00%	2 Missing ⚠️
pvxarray/vtk_source.py	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #83      +/-   ##
==========================================
+ Coverage   43.09%   43.77%   +0.68%     
==========================================
  Files          11       11              
  Lines         478      466      -12     
==========================================
- Hits          206      204       -2     
+ Misses        272      262      -10     

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.