Expand description
VectorExtractor trait for domain-specific document extraction.
This module defines the core abstraction that enables Fabryk to support
multiple knowledge domains for vector search. Each domain implements
VectorExtractor to control how content files are transformed into
VectorDocument instances for embedding.
§Design Philosophy
The extractor separates text composition from embedding. Domains control
what text gets embedded (title, description, body, etc.) by composing
the VectorDocument.text field. The embedding provider then handles
the actual vector generation.
Structs§
- Mock
Vector Extractor - A simple mock extractor for testing.
Traits§
- Vector
Extractor - Trait for extracting vector documents from domain-specific content.