Struct HybridSearchBuilder 

pub struct HybridSearchBuilder { /* private fields */ }

Builder for hybrid search queries that combine vector similarity with full-text search.

This generates queries that use both pgvector distance operators and PostgreSQL tsvector/tsquery for combined similarity scoring.

Examples

use prax_pgvector::query::HybridSearchBuilder;
use prax_pgvector::{Embedding, DistanceMetric};

let query = HybridSearchBuilder::new("documents")
    .vector_column("embedding")
    .text_column("content")
    .query_vector(Embedding::new(vec![0.1, 0.2, 0.3]))
    .query_text("machine learning")
    .metric(DistanceMetric::Cosine)
    .vector_weight(0.7)
    .text_weight(0.3)
    .limit(10)
    .build();

let sql = query.to_sql();

Implementations

impl HybridSearchBuilder

pub fn new(table: impl Into<String>) -> Self

Create a new hybrid search builder.

pub fn vector_column(self, column: impl Into<String>) -> Self

Set the vector column name.

pub fn text_column(self, column: impl Into<String>) -> Self

Set the text column name.

pub fn query_vector(self, embedding: Embedding) -> Self

Set the query vector.

pub fn query_text(self, text: impl Into<String>) -> Self

Set the text query.

pub fn metric(self, metric: DistanceMetric) -> Self

Set the vector distance metric.

pub fn vector_weight(self, weight: f64) -> Self

Set the weight for the vector similarity component (0.0 to 1.0).

pub fn text_weight(self, weight: f64) -> Self

Set the weight for the text relevance component (0.0 to 1.0).

pub fn limit(self, limit: usize) -> Self

Set the result limit.

pub fn language(self, language: impl Into<String>) -> Self

Set the text search language.

pub fn where_clause(self, condition: impl Into<String>) -> Self

Add a WHERE condition.

pub fn build(self) -> HybridSearchQuery

Build the hybrid search query.