snowykami/meilisearch
1
0
Fork 0
mirror of https://github.com/meilisearch/meilisearch.git synced 2025-01-18 08:48:32 +08:00
Code Issues Actions 3 Packages Projects Releases Wiki Activity

Replace old tokenizer by charabia

Browse Source
This commit is contained in:
ManyTheFish 2022-06-06 15:53:28 +02:00
parent 8d09772334
commit 173eea06e1
1 changed files with 7 additions and 23 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

30
meilisearch-lib/src/index/search.rs
View File

@ -4,7 +4,7 @@ use std::str::FromStr;
use std::time::Instant; use std::time::Instant;
use either::Either; use either::Either;
use milli::tokenizer::{Analyzer, AnalyzerConfig}; use milli::tokenizer::TokenizerBuilder;
use milli::{ use milli::{
AscDesc, FieldId, FieldsIdsMap, Filter, FormatOptions, MatchBounds, MatcherBuilder, SortError, AscDesc, FieldId, FieldsIdsMap, Filter, FormatOptions, MatchBounds, MatcherBuilder, SortError,
}; };
@ -175,12 +175,9 @@ impl Index {
&displayed_ids, &displayed_ids,
); );
let stop_words = fst::Set::default(); let tokenizer = TokenizerBuilder::default().build();
let mut config = AnalyzerConfig::default();
config.stop_words(&stop_words);
let analyzer = Analyzer::new(config);
let mut formatter_builder = MatcherBuilder::from_matching_words(matching_words); let mut formatter_builder = MatcherBuilder::new(matching_words, tokenizer);
formatter_builder.crop_marker(query.crop_marker); formatter_builder.crop_marker(query.crop_marker);
formatter_builder.highlight_prefix(query.highlight_pre_tag); formatter_builder.highlight_prefix(query.highlight_pre_tag);
formatter_builder.highlight_suffix(query.highlight_post_tag); formatter_builder.highlight_suffix(query.highlight_post_tag);
@ -204,7 +201,6 @@ impl Index {
&displayed_document, &displayed_document,
&fields_ids_map, &fields_ids_map,
&formatter_builder, &formatter_builder,
&analyzer,
&formatted_options, &formatted_options,
query.show_matches_position, query.show_matches_position,
&displayed_ids, &displayed_ids,
@ -414,8 +410,7 @@ fn make_document(
fn format_fields<'a, A: AsRef<[u8]>>( fn format_fields<'a, A: AsRef<[u8]>>(
document: &Document, document: &Document,
field_ids_map: &FieldsIdsMap, field_ids_map: &FieldsIdsMap,
builder: &MatcherBuilder, builder: &MatcherBuilder<'a, A>,
analyzer: &'a Analyzer<'a, A>,
formatted_options: &BTreeMap<FieldId, FormatOptions>, formatted_options: &BTreeMap<FieldId, FormatOptions>,
compute_matches: bool, compute_matches: bool,
displayable_ids: &BTreeSet<FieldId>, displayable_ids: &BTreeSet<FieldId>,
@ -446,7 +441,6 @@ fn format_fields<'a, A: AsRef<[u8]>>(
std::mem::take(value), std::mem::take(value),
builder, builder,
format, format,
analyzer,
&mut infos, &mut infos,
compute_matches, compute_matches,
); );
@ -470,19 +464,14 @@ fn format_fields<'a, A: AsRef<[u8]>>(
fn format_value<'a, A: AsRef<[u8]>>( fn format_value<'a, A: AsRef<[u8]>>(
value: Value, value: Value,
builder: &MatcherBuilder, builder: &MatcherBuilder<'a, A>,
format_options: Option<FormatOptions>, format_options: Option<FormatOptions>,
analyzer: &'a Analyzer<'a, A>,
infos: &mut Vec<MatchBounds>, infos: &mut Vec<MatchBounds>,
compute_matches: bool, compute_matches: bool,
) -> Value { ) -> Value {
match value { match value {
Value::String(old_string) => { Value::String(old_string) => {
// this will be removed with charabia let mut matcher = builder.build(&old_string);
let analyzed = analyzer.analyze(&old_string);
let tokens: Vec<_> = analyzed.tokens().collect();
let mut matcher = builder.build(&tokens[..], &old_string);
if compute_matches { if compute_matches {
let matches = matcher.matches(); let matches = matcher.matches();
infos.extend_from_slice(&matches[..]); infos.extend_from_slice(&matches[..]);
@ -507,7 +496,6 @@ fn format_value<'a, A: AsRef<[u8]>>(
highlight: format_options.highlight, highlight: format_options.highlight,
crop: None, crop: None,
}), }),
analyzer,
infos, infos,
compute_matches, compute_matches,
) )
@ -527,7 +515,6 @@ fn format_value<'a, A: AsRef<[u8]>>(
highlight: format_options.highlight, highlight: format_options.highlight,
crop: None, crop: None,
}), }),
analyzer,
infos, infos,
compute_matches, compute_matches,
), ),
@ -536,12 +523,9 @@ fn format_value<'a, A: AsRef<[u8]>>(
.collect(), .collect(),
), ),
Value::Number(number) => { Value::Number(number) => {
// this will be removed with charabia
let s = number.to_string(); let s = number.to_string();
let analyzed = analyzer.analyze(&s);
let tokens: Vec<_> = analyzed.tokens().collect();
let mut matcher = builder.build(&tokens[..], &s); let mut matcher = builder.build(&s);
if compute_matches { if compute_matches {
let matches = matcher.matches(); let matches = matcher.matches();
infos.extend_from_slice(&matches[..]); infos.extend_from_slice(&matches[..]);