Module integration 

Cross-integration helpers for the ruvnet crate ecosystem.

Provides integration::FannAdapter for ruv-FANN layer-embedding storage and integration::SemanticSearchAdapter for sparc semantic file search. Cross-integration helpers for ruvnet crate ecosystem.

This module provides ergonomic adapters that make it straightforward to use ruvector-core as a dependency from other ruvnet crates:

• ruv-FANN: neural-network weights can be stored and retrieved via [FannAdapter] using cosine similarity search across layer embeddings.
• sparc / semantic file search: [SemanticSearchAdapter] wraps VectorDB with file-path metadata so sparc can locate relevant source files by embedding query strings.

Both adapters are thin, zero-overhead wrappers — they own no additional memory beyond what the underlying VectorDB already holds.

Structs

FannAdapter

Adapter that lets ruv-FANN store and retrieve layer-weight embeddings.

FileEntry

A file-path entry as indexed by SemanticSearchAdapter.

SemanticSearchAdapter

Adapter for sparc-style semantic file search.

Functions

cosine_similarity

Compute the cosine similarity in [−1, 1] between two vectors.

normalize

Normalise a vector to unit length for cosine-distance workloads.