anda_db_hnsw/lib.rs
1//! # Anda-DB HNSW Vector Search Library
2//!
3//! A high-performance implementation of Hierarchical Navigable Small World (HNSW) algorithm
4//! for approximate nearest neighbor search in high-dimensional spaces.
5//!
6//! HNSW is a graph-based indexing algorithm that creates a multi-layered structure
7//! to enable fast and accurate nearest neighbor search in high-dimensional spaces.
8//!
9//! ## Features
10//!
11//! - Fast approximate nearest neighbor search;
12//! - Multiple distance metrics (Euclidean, Cosine, Inner Product, Manhattan);
13//! - Configurable index parameters;
14//! - Thread-safe implementation with concurrent read/write operations;
15//! - Serialization and deserialization support;
16//! - Support for bf16 (brain floating point) vector storage for memory efficiency.
17//!
18
19mod distance;
20mod error;
21mod hnsw;
22
23pub use distance::*;
24pub use error::*;
25pub use hnsw::*;