fix(matrixone): apply score threshold to vector search

[csurong]

Summary

Fixes #39783

• Normalize MatrixOne L2 distances into similarity scores with 1 / (1 + distance).
• Filter vector search results using score_threshold.
• Include the similarity score in returned document metadata.
• Add unit coverage for the threshold behavior.

Screenshots

Not applicable (backend-only change).

Checklist

• This change requires a documentation update, included: Dify Document
• I understand that this PR may be closed in case there was no previous discussion or issues. (This doesn't apply to typos!)
• I've added a test for each change that was introduced, and I tried as much as possible to make a single atomic change.
• I've updated the documentation accordingly.
• I ran make lint && make type-check (backend) and cd web && pnpm exec vp staged (frontend) to appease the lint gods

Tests

• uv run --project api pytest -q -o addopts='' api/providers/vdb/vdb-matrixone/tests/unit_tests/test_matrixone_vector.py (15 passed)
• Ruff check and format check for the changed files
• Pyrefly check for the changed backend file

[github-actions]

Pyrefly Type Coverage

Metric	Base	PR	Delta
Type coverage	56.78%	56.78%	+0.00%
Strict coverage	56.31%	56.31%	+0.00%
Typed symbols	37,092	37,092	0
Untyped symbols	28,468	28,467	-1
Modules	3091	3091	0