fix: handle empty filters in redis vector search

[Exploreunive]

Summary

This PR fixes RedisDB.search() so it no longer crashes when filters are omitted or effectively empty.

Problem

The Redis vector store built a filter expression like this:

conditions = [Tag(key) == value for key, value in filters.items() if value is not None]
filter = reduce(lambda x, y: x & y, conditions)

When filters was:

• None
• {}
• or a dict whose values were all None

conditions became empty and reduce(...) raised an exception.

This makes Redis search less robust than other vector store backends for no-filter / empty-filter calls.

Fix

• Normalize filters with filters = filters or {}
• Only build the combined Redis filter expression when at least one non-None condition exists
• Otherwise pass None as filter_expression

Tests

Added regression coverage in tests/vector_stores/test_redis.py for:

• filters=None
• filters={}
• filters={"user_id": None}

Validation

.venv/bin/python -m pytest -q tests/vector_stores/test_redis.py

Result:

5 passed