snowykami/meilisearch
1
0
Fork 0
mirror of https://github.com/meilisearch/meilisearch.git synced 2024-11-23 02:27:40 +08:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

simple crop before

Browse Source
This commit is contained in:
Marin Postma 2021-05-11 17:27:31 +02:00 committed by Clémentine Urquizar
parent 93002e734c
commit 56c9633c53
No known key found for this signature in database
GPG Key ID: D8E7CC7422E77E1A
1 changed files with 21 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
meilisearch-http/src/index/search.rs
View File

@ -7,7 +7,7 @@ use either::Either;
use heed::RoTxn;
use indexmap::IndexMap;
use itertools::Itertools;
use meilisearch_tokenizer::{Analyzer, AnalyzerConfig};
use meilisearch_tokenizer::{Analyzer, AnalyzerConfig, Token};
use milli::{FilterCondition, FieldId, FieldsIdsMap, MatchingWords};
use serde::{Deserialize, Serialize};
use serde_json::Value;
@ -303,7 +303,7 @@ impl<'a, A: AsRef<[u8]>> Highlighter<'a, A> {
&self,
value: Value,
matcher: &impl Matcher,
need_to_crop: Option<u32>,
need_to_crop: Option<usize>,
need_to_highlight: bool,
) -> Value {
match value {
@ -326,30 +326,34 @@ impl<'a, A: AsRef<[u8]>> Highlighter<'a, A> {
value => value,
}
}
fn format_string(&self, s: String, matcher: &impl Matcher, need_to_crop: Option<u32>, need_to_highlight: bool) -> String {
fn format_string(&self, s: String, matcher: &impl Matcher, need_to_crop: Option<usize>, need_to_highlight: bool) -> String {
let analyzed = self.analyzer.analyze(&s);
let word_iter: Box<dyn Iterator<Item = (String, bool)>> = if let Some(_crop_len) = need_to_crop {
// cropping iterator
todo!()
} else {
Box::new(analyzed.reconstruct().map(|(word, token)| {
if token.is_word() && matcher.matches(token.text()){
(word.to_string(), true)
} else {
(word.to_string(), false)
}
}))
let tokens: Box<dyn Iterator<Item=(&str, Token)>> = match need_to_crop {
Some(crop_len) => {
let mut taken = 0;
let iter = analyzed
.reconstruct()
.skip_while(|(_, token)| !matcher.matches(token.text()))
.take_while(move |(word, _)| {
let take = taken < crop_len;
taken += word.chars().count();
take
});
Box::new(iter)
},
None => Box::new(analyzed.reconstruct()),
};
word_iter.map(|(word, is_match)| {
if need_to_highlight && is_match {
tokens.map(|(word, token)| {
if need_to_highlight && token.is_word() && matcher.matches(token.text()){
let mut new_word = String::new();
new_word.push_str(&self.marks.0);
new_word.push_str(&word);
new_word.push_str(&self.marks.1);
new_word
} else {
word
word.to_string()
}
})
.collect::<String>()