Expand description
§RuVector Graph Database
A high-performance graph database layer built on RuVector with Neo4j compatibility. Supports property graphs, hypergraphs, Cypher queries, ACID transactions, and distributed queries.
Re-exports§
pub use edge::Edge;pub use edge::EdgeBuilder;pub use error::GraphError;pub use error::Result;pub use graph::GraphDB;pub use hyperedge::Hyperedge;pub use hyperedge::HyperedgeBuilder;pub use hyperedge::HyperedgeId;pub use bm25::Bm25Index;pub use bm25::Bm25Params;pub use embed::Embedder;pub use embed::HashEmbedder;pub use node::Node;pub use node::NodeBuilder;pub use schema::reciprocal_rank_fusion;pub use schema::DistanceMetric;pub use schema::EdgeSchema;pub use schema::GraphSchema;pub use schema::NodeSchema;pub use schema::PropertySchema;pub use schema::PropertyType;pub use schema::VectorSchema;pub use typed_graph::Direction;pub use typed_graph::TraversalResult;pub use typed_graph::TraverseSpec;pub use typed_graph::TypedGraph;pub use storage::GraphStorage;pub use transaction::IsolationLevel;pub use transaction::Transaction;pub use transaction::TransactionManager;pub use types::EdgeId;pub use types::Label;pub use types::NodeId;pub use types::Properties;pub use types::PropertyValue;pub use types::RelationType;pub use hybrid::EmbeddingConfig;pub use hybrid::GnnConfig;pub use hybrid::GraphNeuralEngine;pub use hybrid::HybridIndex;pub use hybrid::RagConfig;pub use hybrid::RagEngine;pub use hybrid::SemanticSearch;pub use hybrid::VectorCypherParser;pub use distributed::Coordinator;pub use distributed::Federation;pub use distributed::GossipMembership;pub use distributed::GraphReplication;pub use distributed::GraphShard;pub use distributed::RpcClient;pub use distributed::RpcServer;pub use distributed::ShardCoordinator;pub use distributed::ShardStrategy;
Modules§
- bm25
- Compact BM25 keyword index over a node text property (ADR-252 P4).
- codegen
- Schema-driven typed client codegen (HelixDB-inspired, ADR-252 P6).
- cypher
- Cypher query language parser and execution engine
- distributed
- Distributed graph query capabilities
- edge
- Edge (relationship) implementation
- embed
- Pluggable text embedding for inline
embed()-at-insert/at-query (HelixDB-inspired, ADR-252 P3). - error
- Error types for graph database operations
- executor
- High-performance query execution engine for RuVector graph database
- graph
- Graph database implementation with concurrent access and indexing
- hybrid
- Vector-Graph Hybrid Query System
- hyperedge
- N-ary relationship support (hyperedges)
- index
- Index structures for fast node and edge lookups
- node
- Node implementation
- optimization
- Performance optimization modules for orders of magnitude speedup
- schema
- Optional, schema-first type layer for the graph (HelixDB-inspired, ADR-252 P1/P2).
- storage
- Persistent storage layer with redb and memory-mapped vectors
- transaction
- Transaction support for ACID guarantees with MVCC
- typed_
graph - Schema-validated graph wrapper with a fused vector-search-then-traverse operator (HelixDB-inspired, ADR-252 P2).
- types
- Core types for graph database