Struct VectorLiteClient 

pub struct VectorLiteClient { /* private fields */ }

Main client for interacting with VectorLite

The VectorLiteClient provides a high-level interface for managing vector collections, performing searches, and handling embeddings. It’s designed to be thread-safe and efficient for AI agent workloads.

Thread Safety

The client uses Arc<RwLock<>> for collections and atomic counters for ID generation, making it safe to use across multiple threads.

Examples

use vectorlite::{VectorLiteClient, EmbeddingGenerator, IndexType};

let mut client = VectorLiteClient::new(Box::new(EmbeddingGenerator::new()?));
client.create_collection("docs", IndexType::HNSW)?;

Implementations

impl VectorLiteClient

pub fn new(embedding_function: Box<dyn EmbeddingFunction>) -> Self

pub fn create_collection( &mut self, name: &str, index_type: IndexType, ) -> VectorLiteResult<()>

pub fn get_collection(&self, name: &str) -> Option<&Arc<Collection>>

pub fn list_collections(&self) -> Vec<String>

pub fn delete_collection(&mut self, name: &str) -> VectorLiteResult<()>

pub fn has_collection(&self, name: &str) -> bool

pub fn add_text_to_collection( &self, collection_name: &str, text: &str, metadata: Option<Value>, ) -> VectorLiteResult<u64>

pub fn search_text_in_collection( &self, collection_name: &str, query_text: &str, k: usize, similarity_metric: SimilarityMetric, ) -> VectorLiteResult<Vec<SearchResult>>

pub fn delete_from_collection( &self, collection_name: &str, id: u64, ) -> VectorLiteResult<()>

pub fn get_vector_from_collection( &self, collection_name: &str, id: u64, ) -> VectorLiteResult<Option<Vector>>

pub fn get_collection_info( &self, collection_name: &str, ) -> VectorLiteResult<CollectionInfo>

pub fn add_collection(&mut self, collection: Collection) -> VectorLiteResult<()>

Add a collection directly (used for loading from files)