feat: Support advanced search filters (time range, tags, type, importance)

[NTLx]

This PR implements advanced search filtering capabilities, addressing Issue #140.

🚀 Features

• Standardized Filter Parser: Introduced src/powermem/utils/filter_parser.py to translate high-level API filters into storage-compatible queries.
• Time Range Filtering: Supports start_time and end_time filters (mapped to created_at range queries).
• Tag Filtering: Supports tags filtering (list or single value).
• Type/Category Filtering: Maps type filter to internal category field.
• Importance Filtering: Supports importance threshold filtering.
• Integration: Integrated the parser into PySeekDBVectorStore, enabling these features for the new backend.

🛠 Implementation Details

1. Filter Logic: The parser converts user-friendly keys like start_time into operator-based queries (e.g., {'': ...}) compatible with vector databases like Chroma/SeekDB.
2. Extensibility: The parser is designed as a standalone utility, making it easy to apply consistent filtering logic across other storage backends (OceanBase, PGVector) in the future.
3. PySeekDB Support: Updated the PySeekDBVectorStore implementation to utilize this parser in both search() and list() methods.

✅ AI Contribution Details

• Analysis: Analyzed the existing SearchRequest model and identified that the service layer already supports passing filters, isolating the work to the storage/utility layer.
• Design: Designed a non-intrusive utility function to standardize filter parsing without modifying core Service logic.
• Implementation: Implemented the parser and integrated it into the storage adapter.

Ready for review! 🔍

[Teingi]

There are merge conflicts in this pull request. Please resolve them before we can proceed with the review.

[NTLx]

Hello @Teingi, I have resolved the merge conflicts in pyseekdb_store.py. The advanced filter logic has been correctly integrated with the latest base code. The PR is now ready for further review. Thanks!

[Teingi]

LGTM