Fix reset_index crash when obs index name matches existing column

[timtreis]

Summary

• _inner_join_spatialelement_table and _left_join_spatialelement_table call table.obs.reset_index() which raises ValueError: cannot insert <name>, already exists when the obs index name matches an existing column (e.g. EntityID in Merfish data)
• Adds a small helper _reset_index_preserving_existing_columns that drops the index when its name already exists as a column, avoiding the collision

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 91.93%. Comparing base (78f75ef) to head (ef5b015).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1100   +/-   ##
=======================================
  Coverage   91.93%   91.93%           
=======================================
  Files          51       51           
  Lines        7772     7776    +4     
=======================================
+ Hits         7145     7149    +4     
  Misses        627      627           

Files with missing lines	Coverage Δ
src/spatialdata/_core/query/relational_query.py	91.70% <100.00%> (+0.07%)	⬆️

🚀 New features to boost your workflow:

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

[LucaMarconato]

So here we are dropping the index instead of a column. This could be a problem if the column is different from the index.

What's the origin behind this problem? Is it that for the merfish data we should have had sdata['table'].index.name = None? That would be a better fix, while still raising an exception here.

[LucaMarconato]

I think it should crash in general to avoid a silent branching in the behavior. The ok case would be if both the index and the column have the same values.