fix(bundle-analysis): clean up temp files on load failure

[drazisil-codecov]

Summary

• BundleAnalysisReportLoader.load() called tempfile.mkstemp() but never deleted the temp file on the FileNotInStorageError path (returned None without cleanup) or on any unexpected exception path
• Workers running many BA tasks accumulated SQLite temp files on disk until hitting No space left on device (WORKER-WNZ)
• Fix restructures the try/except so both error paths call os.unlink(db_path) before returning/re-raising; the happy path still transfers ownership to the returned BundleAnalysisReport, which has its own cleanup() method

Test plan

• test_load_cleans_up_tempfile_on_file_not_found — patches mkstemp to capture the path, asserts file no longer exists after load() returns None
• test_load_cleans_up_tempfile_on_unexpected_error — same, but for an unexpected RuntimeError that should re-raise
• test_save_load_bundle_report — existing happy-path test still passes

🤖 Generated with Claude Code

---

Note

Low Risk
Low risk: narrow change to temporary-file lifecycle in BundleAnalysisReportLoader.load, plus targeted unit tests; only impacts failure paths and should reduce disk-leak incidents.

Overview
Fixes BundleAnalysisReportLoader.load() to always clean up the mkstemp-created SQLite temp file when the report is missing (FileNotInStorageError) or when any unexpected exception occurs, instead of leaking files on disk.

Adds unit tests that assert the temp file is deleted on both not found and unexpected error paths, while keeping the existing save/load happy-path behavior intact.

^{Reviewed by Cursor Bugbot for commit f2003b3. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 4adb001. Configure here.}

[codspeed-hq]

Merging this PR will not alter performance

✅ 9 untouched benchmarks

---

_{Comparing fix/bundle-analysis-tempfile-leak (f2003b3) with main (3d2b291)¹}

Footnotes

1. No successful run was found on main (9203b6d) during the generation of this report, so 3d2b291 was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

[codecov-notifications]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ All tests successful. No failed tests found.

📢 Thoughts on this report? Let us know!

[sentry]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 91.89%. Comparing base (9203b6d) to head (f2003b3).
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #917   +/-   ##
=======================================
  Coverage   91.89%   91.89%           
=======================================
  Files        1316     1316           
  Lines       50633    50639    +6     
  Branches     1625     1625           
=======================================
+ Hits        46531    46537    +6     
  Misses       3796     3796           
  Partials      306      306           

Flag	Coverage Δ
apiunit	94.95% <ø> (ø)
sharedintegration	36.96% <9.09%> (-0.01%)	⬇️
sharedunit	84.90% <100.00%> (+<0.01%)	⬆️
workerintegration	58.53% <ø> (ø)
workerunit	90.37% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[thomasrockhu-codecov]

Make sure os.close(fd) doesn't affect our ability to use db_path downstream

[drazisil-codecov]

Confirmed locally that unlinking fd does not affect db_path.