searchengines-web/MeiliSearch
searchengines-web/MeiliSearch
1
0
Fork 0
mirror of https://github.com/meilisearch/MeiliSearch synced 2024-11-23 05:14:27 +01:00
Code Issues Releases Wiki Activity

Use new term matching strategy logic in resolve_maximally_reduced_...

Browse Source
This commit is contained in:
Loïc Lecrenier 2023-03-30 13:12:51 +02:00
parent 2a5997fb20
commit 2997d1f186
1 changed files with 11 additions and 27 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
milli/src/search/new/mod.rs
View File

@ -16,7 +16,7 @@ mod sort;
// TODO: documentation + comments // TODO: documentation + comments
mod words; mod words;
use std::collections::{BTreeSet, HashSet}; use std::collections::HashSet;
use charabia::TokenizerBuilder; use charabia::TokenizerBuilder;
use db_cache::DatabaseCache; use db_cache::DatabaseCache;
@ -25,7 +25,7 @@ use heed::RoTxn;
use interner::DedupInterner; use interner::DedupInterner;
pub use logger::detailed::DetailedSearchLogger; pub use logger::detailed::DetailedSearchLogger;
pub use logger::{DefaultSearchLogger, SearchLogger}; pub use logger::{DefaultSearchLogger, SearchLogger};
use query_graph::{QueryGraph, QueryNode, QueryNodeData}; use query_graph::{QueryGraph, QueryNode};
use query_term::{located_query_terms_from_string, Phrase, QueryTerm}; use query_term::{located_query_terms_from_string, Phrase, QueryTerm};
use ranking_rules::{bucket_sort, PlaceholderQuery, RankingRuleOutput, RankingRuleQueryTrait}; use ranking_rules::{bucket_sort, PlaceholderQuery, RankingRuleOutput, RankingRuleQueryTrait};
use resolve_query_graph::PhraseDocIdsCache; use resolve_query_graph::PhraseDocIdsCache;
@ -75,33 +75,17 @@ fn resolve_maximally_reduced_query_graph(
logger: &mut dyn SearchLogger<QueryGraph>, logger: &mut dyn SearchLogger<QueryGraph>,
) -> Result<RoaringBitmap> { ) -> Result<RoaringBitmap> {
let mut graph = query_graph.clone(); let mut graph = query_graph.clone();
let mut positions_to_remove = match matching_strategy {
TermsMatchingStrategy::Last => { let nodes_to_remove = match matching_strategy {
let mut all_positions = BTreeSet::new(); TermsMatchingStrategy::Last => query_graph
for (_, n) in query_graph.nodes.iter() { .removal_order_for_terms_matching_strategy_last()
match &n.data { .iter()
QueryNodeData::Term(term) => { .flat_map(|x| x.iter())
all_positions.extend(term.positions.clone()); .collect(),
}
QueryNodeData::Deleted | QueryNodeData::Start | QueryNodeData::End => {}
}
}
all_positions.into_iter().collect()
}
TermsMatchingStrategy::All => vec![], TermsMatchingStrategy::All => vec![],
}; };
// don't remove the first term graph.remove_nodes(&nodes_to_remove);
if !positions_to_remove.is_empty() {
positions_to_remove.remove(0);
}
loop {
if positions_to_remove.is_empty() {
break;
} else {
let position_to_remove = positions_to_remove.pop().unwrap();
let _ = graph.remove_words_starting_at_position(position_to_remove);
}
}
logger.query_for_universe(&graph); logger.query_for_universe(&graph);
let docids = compute_query_graph_docids(ctx, &graph, universe)?; let docids = compute_query_graph_docids(ctx, &graph, universe)?;