Struct EnrichedRetriever 

pub struct EnrichedRetriever { /* private fields */ }

Metadata-enhanced retrieval strategies

Implementations

impl EnrichedRetriever

pub fn new() -> Self

Create a new enriched retriever

pub fn with_config(config: EnrichedRetrievalConfig) -> Self

Create with custom configuration

pub fn metadata_search( &self, query: &str, graph: &KnowledgeGraph, _analysis: &QueryAnalysis, base_results: &[SearchResult], ) -> Result<Vec<SearchResult>>

Search chunks using enriched metadata

This method boosts chunks that match:

1. Query keywords present in chunk keywords (TF-IDF extracted)
2. Chapter/Section mentioned in query
3. Summary content relevant to query

pub fn filter_by_structure( &self, query: &str, results: Vec<SearchResult>, graph: &KnowledgeGraph, ) -> Result<Vec<SearchResult>>

Filter chunks by chapter or section

Example: “What does Socrates say in Chapter 1?” -> filter to Chapter 1 chunks

pub fn boost_with_metadata( &self, results: Vec<SearchResult>, query: &str, graph: &KnowledgeGraph, ) -> Result<Vec<SearchResult>>

Boost results based on enriched metadata

pub fn get_chapter_chunks<'a>( &self, chapter_name: &str, graph: &'a KnowledgeGraph, ) -> Vec<&'a TextChunk>

Get chunks from a specific chapter

pub fn get_section_chunks<'a>( &self, section_name: &str, graph: &'a KnowledgeGraph, ) -> Vec<&'a TextChunk>

Get chunks from a specific section

pub fn search_by_keywords( &self, keywords: &[String], graph: &KnowledgeGraph, top_k: usize, ) -> Vec<SearchResult>

Search by keywords extracted from chunks

Trait Implementations

impl Default for EnrichedRetriever

fn default() -> Self

Returns the “default value” for a type.