searchengines-web/MeiliSearch
searchengines-web/MeiliSearch
1
0
Fork 0
mirror of https://github.com/meilisearch/MeiliSearch synced 2025-07-04 20:37:15 +02:00
Code Issues Releases Wiki Activity

Replace the regex highlighting by a simple algorithm

Browse source
This commit is contained in:
Clément Renault 2020-08-05 13:52:27 +02:00
parent 4873abe145
commit 6508d497ce
No known key found for this signature in database
GPG key ID: 92ADA4E935E71FA4
3 changed files with 16 additions and 37 deletions
Download patch file Download diff file Expand all files Collapse all files

26
src/bin/serve.rs
View file

@ -1,4 +1,5 @@
use std::borrow::Cow;
use std::collections::HashSet;
use std::fs::File;
use std::net::SocketAddr;
use std::path::PathBuf;
@ -7,8 +8,8 @@ use std::time::Instant;
use askama_warp::Template;
use heed::EnvOpenOptions;
use regex::Regex;
use serde::Deserialize;
use slice_group_by::StrGroupBy;
use structopt::StructOpt;
use warp::{Filter, http::Response};
@ -44,6 +45,18 @@ struct Opt {
http_listen_addr: String,
}
fn highlight_string(string: &str, words: &HashSet<String>) -> String {
let mut output = String::new();
for token in string.linear_group_by_key(|c| c.is_alphanumeric()) {
let lowercase_token = token.to_lowercase();
let to_highlight = words.contains(&lowercase_token);
if to_highlight { output.push_str("<mark>") }
output.push_str(token);
if to_highlight { output.push_str("</mark>") }
}
output
}
#[derive(Template)]
#[template(path = "index.html")]
struct IndexTemplate {
@ -173,15 +186,6 @@ async fn main() -> anyhow::Result<()> {
// We write the headers
body.extend_from_slice(headers);
let mut regex = format!(r"(?i)\b(");
let number_of_words = words.len();
words.into_iter().enumerate().for_each(|(i, w)| {
regex.push_str(&w);
if i != number_of_words - 1 { regex.push('|') }
});
regex.push_str(r")\b");
let re = Regex::new(&regex).unwrap();
for id in documents_ids {
let content = index.documents.get(&rtxn, &BEU32::new(id)).unwrap();
let content = content.expect(&format!("could not find document {}", id));
@ -190,7 +194,7 @@ async fn main() -> anyhow::Result<()> {
let content = if disable_highlighting {
Cow::from(content)
} else {
re.replace_all(content, "<mark>$1</mark>")
Cow::from(highlight_string(content, &words))
};
body.extend_from_slice(content.as_bytes());