snowykami/meilisearch
1
0
Fork 0
mirror of https://github.com/meilisearch/meilisearch.git synced 2025-07-28 01:01:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

Introduce an initial candidates set that makes the difference between an exhaustive count and an estimation

Browse Source
This commit is contained in:
ManyTheFish
2022-12-07 18:29:25 +01:00
parent 6d50ea0830
commit 55724f2412
11 changed files with 180 additions and 101 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
milli/src/search/mod.rs
View File

@ -23,6 +23,7 @@ pub use self::matches::{
use self::query_tree::QueryTreeBuilder;
use crate::error::UserError;
use crate::search::criteria::r#final::{Final, FinalResult};
use crate::search::criteria::InitialCandidates;
use crate::{AscDesc, Criterion, DocumentId, Index, Member, Result};
// Building these factories is not free.
@ -235,11 +236,11 @@ impl<'a> Search<'a> {
mut criteria: Final,
) -> Result<SearchResult> {
let mut offset = self.offset;
let mut initial_candidates = RoaringBitmap::new();
let mut initial_candidates = InitialCandidates::Estimated(RoaringBitmap::new());
let mut excluded_candidates = self.index.soft_deleted_documents_ids(self.rtxn)?;
let mut documents_ids = Vec::new();
while let Some(FinalResult { candidates, bucket_candidates, .. }) =
while let Some(FinalResult { candidates, initial_candidates: ic, .. }) =
criteria.next(&excluded_candidates)?
{
debug!("Number of candidates found {}", candidates.len());
@ -247,7 +248,7 @@ impl<'a> Search<'a> {
let excluded = take(&mut excluded_candidates);
let mut candidates = distinct.distinct(candidates, excluded);
initial_candidates |= bucket_candidates;
initial_candidates |= ic;
if offset != 0 {
let discarded = candidates.by_ref().take(offset).count();
@ -265,9 +266,11 @@ impl<'a> Search<'a> {
}
}
initial_candidates.map_inplace(|c| c - excluded_candidates);
Ok(SearchResult {
matching_words,
candidates: initial_candidates - excluded_candidates,
candidates: initial_candidates.into_inner(),
documents_ids,
})
}