Struct HybridSearchQuery 

pub struct HybridSearchQuery {
    pub table: String,
    pub vector_column: String,
    pub text_column: String,
    pub query_vector: Option<Embedding>,
    pub query_text: Option<String>,
    pub metric: DistanceMetric,
    pub vector_weight: f64,
    pub text_weight: f64,
    pub limit: usize,
    pub language: String,
    pub where_clauses: Vec<String>,
}

A hybrid search query combining vector similarity and full-text search.

Result shape: the SQL from to_sql returns two columns — an anonymous composite named coalesce (the merged matched row) and rrf_score. The composite’s fields cannot be expanded or selected in SQL; see to_sql for how to consume it.

Fields

table: String

Table name.

vector_column: String

Vector column.

text_column: String

Text column.

query_vector: Option<Embedding>

Query vector.

query_text: Option<String>

Text query.

metric: DistanceMetric

Distance metric.

vector_weight: f64

Weight for vector similarity (0.0-1.0).

text_weight: f64

Weight for text relevance (0.0-1.0).

limit: usize

Result limit.

language: String

Text search language.

where_clauses: Vec<String>

Additional WHERE conditions.

Implementations

impl HybridSearchQuery

pub fn to_sql(&self) -> String

Generate the SQL query using Reciprocal Rank Fusion (RRF).

RRF combines rankings from multiple retrieval methods: score = sum(1 / (k + rank_i)) where k is a constant (typically 60).

The query vector should be $1 and the text query should be $2.

Result shape

The final SELECT returns two columns: an anonymous composite named coalesce — the whole matched row from vector_results (when both sides match, the vector side wins) or text_results — plus rrf_score. The composite’s record type is never registered, so PostgreSQL rejects both (coalesce).* expansion and per-field access like (coalesce).id with “record type has not been registered”. Consumers must either decode the composite value client-side or wrap the query and project row_to_json(coalesce) (or to_json(coalesce)). The composite also carries the matching side’s rank column (vec_rank or text_rank) in addition to the base table columns, so its shape is not fixed.

Trait Implementations

impl Clone for HybridSearchQuery

fn clone(&self) -> HybridSearchQuery

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for HybridSearchQuery

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for HybridSearchQuery

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for HybridSearchQuery

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.