DavidMStraub/sifts

Leverages native full-text search capabilities in SQLite and PostgreSQL through a unified Python API, with vector search via pgvector (PostgreSQL) or in-memory cosine similarity (SQLite). Automatically handles special character escaping for FTS5 and tsquery syntax, and integrates embedding functions for semantic search applications like RAG systems. Designed as a library for Flask and similar frameworks, enabling seamless database switching from SQLite development to PostgreSQL production without API changes.

4 stars and 1,588 monthly downloads. Available on PyPI.