Struct RelevanceQuery 

pub struct RelevanceQuery { /* private fields */ }

Parsed query reused while scoring all candidate memories.

Implementations

impl RelevanceQuery

pub fn new(query: &str) -> Self

pub fn is_empty(&self) -> bool

pub fn tokens(&self) -> &[String]

The normalized query tokens (length > 2, deduplicated). Used by the inverted text index to prefilter blocks before lexical scoring.

pub fn lexical_score(&self, text: &str) -> f32

Lexical relevance in [0, 1].

Exact query-token coverage is dominant. A conservative prefix match helps inflected English and Hungarian words without allowing short, noisy fragments to dominate. Contiguous phrase matches break ties.

pub fn rank_distance( &self, text: &str, spatial_dist_sq: f32, keyword_boost: f32, importance: u8, ) -> f32

Continuous lower-is-better rank distance.

Lexical coverage owns most of the rank, spatial distance resolves semantically comparable candidates, and importance is deliberately a small prior so it cannot rescue an unrelated memory.

Trait Implementations

impl Clone for RelevanceQuery

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for RelevanceQuery

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

Formats the value using the given formatter.