Fix 13 production error patterns across 15 files with 27 tests

[cursor]

Summary

Systematic code-driven investigation of the codebase to identify and fix error-prone patterns that cause production exceptions. This PR addresses 13 distinct error patterns across 15 source files, with 27 new tests.

Error Signatures Addressed

1. Sentry SDK Misuse (utils/sentry.py)

• log_error base_error scope bug: scope.set_extra("base_error", message) was recording the message string instead of the base_error — lost the original exception context in Sentry.
• log_request_error passing string to capture_exception: Sentry's capture_exception() expects an Exception, but received a plain string — resulted in malformed Sentry events or silent failures.

2. List Mutation During Iteration (utils/message.py)

• recipients.remove(recipient) inside for recipient in recipients — caused valid recipients to be skipped when consecutive invalid emails appeared.

3. IndexError on Empty Highlights (search/services/unified_search_service.py)

• highlights.paper_title[0] without checking if the fragment list is empty — OpenSearch can return empty fragment lists, causing IndexError.

4. KeyError on Malformed Institutions (search/serializers/person.py)

• institution["id"] and institution["name"] on OpenSearch AttrList data — None entries or missing keys caused KeyError during API serialization.

5. ZeroDivisionError (search/views/suggest.py)

• math.ceil(limit / len(indexes)) when indexes is empty — caused ZeroDivisionError.

6. ValueError on Query Params (research_ai/views/, note/views/note_view.py)

• int(request.query_params.get("limit", 10)) without try/except — non-numeric values like limit=abc caused unhandled ValueError → 500.
• Affected: ExpertSearchListView, EmailTemplateListView, GeneratedEmailListView, and Note invite_user.

7. KeyError on Missing paper_ids (paper/views/paper_views.py)

• request.query_params["paper_ids"] — missing query param caused KeyError → 500.

8. Dead None Guard (paper/views/paper_views.py)

• paper = Paper.objects.get(id=paper_id) followed by if not paper: — objects.get never returns falsy; it raises DoesNotExist. The guard was unreachable.

9. None.split() on Missing Work ID (paper/openalex_util.py)

• work.get("id").split("/") — when "id" key is present but None, calling .split() on None raised AttributeError.

10. None.lower() on Server Field (paper/ingestion/mappers/preprints/biorxiv_base.py)

• record.get("server", default).lower() — when key exists with explicit None value, default is not used, causing AttributeError on .lower().

11. IndexError on Empty Crossref Title (paper/paper_upload_tasks.py)

• results.get("title", [])[0] — empty title list from Crossref API caused IndexError.

12. Multiple Crash Paths in send_support_email (purchase/tasks.py)

• NameError: url referenced in context dict but never defined when content_type is "profile".
• AttributeError: paper.uploaded_by.full_name() when uploaded_by is None.
• AttributeError: paper.paper_type.split("_") when paper_type is None.
• DoesNotExist: Multiple unguarded Paper.objects.get() / ResearchhubPost.objects.get() calls.

13. Unguarded objects.get in Review Create (review/views/review_view.py)

• ResearchhubUnifiedDocument.objects.get(id=args[0]) and ContentType.objects.get(model=...) — both could raise DoesNotExist → 500.

Validation

• 27 new unit tests (all passing, no DB required)
• All fixes are minimal and defensive — they add guards/error handling without changing business logic
• No new dependencies introduced

Remaining Risk

• These fixes prevent crashes but some error conditions (e.g., missing Crossref title, missing OpenAlex work ID) may indicate upstream data quality issues worth monitoring
• PR Fix 9 production error patterns with tests (~100K+ events) #3225 (previous investigation) is still open with additional fixes
• External service errors (X API 403, OpenAlex 404 retries, ScrapingBee PDF failures) require credential/infrastructure changes, not code fixes

  

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[codecov]

Codecov Report

❌ Patch coverage is 33.05785% with 81 lines in your changes missing coverage. Please review.
✅ Project coverage is 79.08%. Comparing base (e3d8c1b) to head (df74a34).

Files with missing lines	Patch %	Lines
src/purchase/tasks.py	0.00%	45 Missing ⚠️
src/note/views/note_view.py	15.38%	11 Missing ⚠️
src/paper/views/paper_views.py	28.57%	5 Missing ⚠️
src/paper/openalex_util.py	0.00%	4 Missing ⚠️
src/paper/paper_upload_tasks.py	0.00%	4 Missing ⚠️
src/review/views/review_view.py	60.00%	4 Missing ⚠️
src/research_ai/views/email_views.py	60.00%	2 Missing ⚠️
src/research_ai/views/expert_finder_views.py	60.00%	2 Missing ⚠️
src/research_ai/views/template_views.py	60.00%	2 Missing ⚠️
src/search/serializers/person.py	77.77%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3257      +/-   ##
==========================================
- Coverage   79.09%   79.08%   -0.01%     
==========================================
  Files         620      620              
  Lines       35081    35130      +49     
==========================================
+ Hits        27746    27784      +38     
- Misses       7335     7346      +11     

☔ 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.