test : added unit tests for finding_intelligence asset reference and signature helpers

[tmdeveloper007]

Closes #1246.

Summary of What Has Been Done:
Added unit tests for 4 pure helper functions in backend/secuscan/finding_intelligence.py: _extract_best_url, _guess_asset_ref, _issue_signature, and _finding_kind_for. Tests cover metadata/evidence/target URL extraction, asset ref resolution, signature generation, and finding kind classification.

Changes Made:

• Created testing/backend/unit/test_finding_intelligence_asset_ref.py with 33 test cases
• _extract_best_url: 10 tests (priority order, metadata keys, evidence, target fallback, edge cases)
• _guess_asset_ref: 5 tests (asset_refs priority, URL extraction, host metadata, target fallback)
• _issue_signature: 6 tests (CVE prefix, field-based, determinism, empty cases)
• _finding_kind_for: 12 tests (observation/suspected/validated classification, category and severity interactions)

Impact it Made:

• Guards finding deduplication and classification logic against regressions
• Documents expected behavior for asset reference extraction
• Increases overall test coverage of the finding_intelligence module

Note: This task is being handled by tmdeveloper007 -- please assign to that account when picking it up.