Feat: Add support for dynamic Milvus search parameters (radius, range_filter)

[ron-42]

Summary

Adds support for radius and range_filter parameters in Milvus vector database searches, allowing users to fine-tune search behavior at runtime by specifying similarity ranges.

Changes:

• Added optional search_params parameter to 4 search methods: search(), async_search(), hybrid_search(), async_hybrid_search()
• Supports range-based searches with radius (outer boundary) and range_filter (inner boundary)
• Full backward compatibility maintained
• Includes test coverage (25/25 tests passing)

Fixes #5173

Type of change

• Bug fix
• New feature
• Breaking change
• Improvement
• Model update
• Other:

---

Checklist

• Code complies with style guidelines
• Ran format/validation scripts (./scripts/format.sh and ./scripts/validate.sh)
• Self-review completed
• Documentation updated (comments, docstrings)
• Examples and guides: Relevant cookbook examples have been included or updated (if applicable)
• Tested in clean environment
• Tests added/updated (if applicable)

---

Changes Made

Modified libs/agno/agno/vectordb/milvus/milvus.py to accept search_params dictionary in:

• search() / async_search() - Parameters passed directly to Milvus client
• hybrid_search() / async_hybrid_search() - Parameters merged into dense/sparse search configs

Added tests in libs/agno/tests/unit/vectordb/test_milvusdb.py:

• test_search_with_radius_and_range_filter() - Tests sync search with parameters
• test_async_search_with_search_params() - Tests async search with parameters

Usage Example

from agno.vectordb.milvus import Milvus

# Initialize Milvus vector database
vector_db = Milvus(
    collection="my_collection",
    uri="http://localhost:19530",
)

# Search with radius only (outer boundary)
results = vector_db.search(
    query="your search query",
    limit=10,
    search_params={
        "radius": 0.8,  # Only return results with similarity >= 0.8
    }
)

# Search within a similarity range [0.5, 0.8]
results = vector_db.search(
    query="your search query",
    limit=10,
    search_params={
        "radius": 0.8,        # Outer boundary (maximum distance)
        "range_filter": 0.5,  # Inner boundary (minimum distance)
    }
)

# Works with async search too
results = await vector_db.async_search(
    query="your search query",
    limit=10,
    search_params={"radius": 0.9, "range_filter": 0.6}
)

Backward Compatibility

• ✅ search_params is optional (defaults to None)
• ✅ No breaking changes to existing code
• ✅ Works with all distance metrics (cosine, L2, IP)
• ✅ All 25 tests passing

Additional Notes

• Works with Milvus Lite, Milvus server, and Zilliz Cloud
• Range searches may return fewer results than limit if constraints are strict
• No performance overhead when search_params is not provided
• Related: Milvus Range Search Docs