Wrap span restoration in __exit__ in capture_internal_exceptions

[sentrivana]

Ref #4718

Does not solve the underlying issue and might leave things in an inconsistent state, but it's still preferable to letting an error bubble up to the user.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 84.89%. Comparing base (f702ec9) to head (44d568e).
⚠️ Report is 5 commits behind head on master.
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #4719      +/-   ##
==========================================
- Coverage   84.89%   84.89%   -0.01%     
==========================================
  Files         156      156              
  Lines       16115    16116       +1     
  Branches     2741     2741              
==========================================
  Hits        13681    13681              
- Misses       1645     1650       +5     
+ Partials      789      785       -4     

Files with missing lines	Coverage Δ
sentry_sdk/tracing.py	86.66% <100.00%> (+0.02%)	⬆️

... and 5 files with indirect coverage changes

[sentrivana]

Fixed properly via #4720

[cursor]

Bug: Tracing Context Corruption Due to Exception Handling

Wrapping the span's __exit__ cleanup in capture_internal_exceptions() can silently prevent the scope's active span from being restored, leaving the tracing context inconsistent. If an intermediate cleanup step fails, subsequent critical operations are skipped. Additionally, set_status is inconsistently outside this block, allowing its internal errors to surface.