fix(ml): resolve Finding and dict type mismatch in deduplication #259

[prasiddhi-105]

Linked issue

Closes #259

What this PR does

Fixes a type annotation and attribute access mismatch in the deduplication pipeline where dict inputs could cause silent empty string extraction in embed_findings(). Also widens exception handling during SentenceTransformer initialization in embedder.py to prevent crashes caused by network errors, GPU memory exhaustion, or model load failures.

Type of change

• Bug fix
• New feature
• ML model / training pipeline
• Refactor (no behaviour change)
• Documentation
• Tests only

ML tier (if applicable)

• Tier 1 — Triage
• Tier 2 — Predictive
• Tier 3 — Autonomous
• Not ML-related

Stack affected

• Backend
• Frontend
• Both

---

Changes

Backend

• backend/app/ml/embedder.py:
  • Refactored _extract_text() to safely handle both Pydantic Finding objects and raw Python dict instances.
  • Widened exception handling around SentenceTransformer loading to catch all errors (e.g., OSError, ConnectionError, OutOfMemoryError) and fail gracefully.
  • Updated embed_findings() type hints to List[Union[Finding, Dict[str, Any]]].
• backend/app/ml/deduplicator.py:
  • Updated type signatures for deduplicate() and internal wrapper embed_findings() to explicitly allow Union[Finding, Dict[str, Any]].

Frontend

• No frontend changes.

New dependencies

• No new dependencies added.

Database / schema changes

• No database schema changes.

---

Testing

How did you test this?

• Ran unit tests locally with python -m pytest (109 passing tests).
• Verified _extract_text returns non-empty strings for both Finding model instances and standard dictionary inputs.

Checklist

• Tested locally end-to-end (upload ZIP or GitHub URL → scan → findings returned correctly)
• New ML model falls back gracefully when model file is absent
• No new console.error or unhandled Python exceptions introduced
• Added or updated tests where applicable
• requirements.txt / package.json updated if new dependencies added
• New model files (.pkl, .pt, etc.) are gitignored, not committed

---

Anything reviewers should focus on

Please check the dual Finding / dict handling in _extract_text() and the broadened Exception catch during SentenceTransformer initialization.

Screenshots (if UI changed)

N/A (Backend logic fix)

[github-actions]

🎉 Thank you @prasiddhi-105 for submitting a Pull Request!

We're excited to review your contribution.

✅ Before Review

• Ensure all CI checks pass.
• Complete the PR template.
• Link the related issue.

⚡ Want faster reviews and contributor support?

Join our Discord community:

🔗 https://discord.gg/FcXuyw2Rs

Maintainers and mentors are active there and can help resolve blockers quickly.

Happy Contributing! 🚀