mirror of
https://gitlab.com/news-flash/article_scraper.git
synced 2025-07-07 16:15:32 +02:00
fmt
This commit is contained in:
parent
ded7cf5adb
commit
92b4427a9f
2 changed files with 13 additions and 6 deletions
|
@ -63,11 +63,12 @@ pub static NEGATIVE: Lazy<Regex> = Lazy::new(|| {
|
|||
|
||||
pub static TITLE_SEPARATOR: Lazy<Regex> =
|
||||
Lazy::new(|| Regex::new(r#" [-|—\\/>»] "#).expect("TITLE_SEPARATOR regex"));
|
||||
pub static TITLE_CUT_END: Lazy<Regex> = Lazy::new(||
|
||||
pub static TITLE_CUT_END: Lazy<Regex> = Lazy::new(|| {
|
||||
RegexBuilder::new(r#"(.*)[-|—\\/>»] .*"#)
|
||||
.case_insensitive(true)
|
||||
.build()
|
||||
.expect("TITLE_CUT_END regex"));
|
||||
.case_insensitive(true)
|
||||
.build()
|
||||
.expect("TITLE_CUT_END regex")
|
||||
});
|
||||
pub static WORD_COUNT: Lazy<Regex> = Lazy::new(|| Regex::new(r#"\s+"#).expect("WORD_COUNT regex"));
|
||||
pub static TITLE_CUT_FRONT: Lazy<Regex> = Lazy::new(|| {
|
||||
RegexBuilder::new(r#"[^-|\\/>»]*[-|\\/>»](.*)"#)
|
||||
|
|
10
src/util.rs
10
src/util.rs
|
@ -317,8 +317,14 @@ impl Util {
|
|||
pub fn text_similarity(a: &str, b: &str) -> f64 {
|
||||
let a = a.to_lowercase();
|
||||
let b = b.to_lowercase();
|
||||
let tokens_a = constants::TOKENIZE.split(&a).filter(|token| !token.is_empty()).collect::<Vec<_>>();
|
||||
let tokens_b = constants::TOKENIZE.split(&b).filter(|token| !token.is_empty()).collect::<Vec<_>>();
|
||||
let tokens_a = constants::TOKENIZE
|
||||
.split(&a)
|
||||
.filter(|token| !token.is_empty())
|
||||
.collect::<Vec<_>>();
|
||||
let tokens_b = constants::TOKENIZE
|
||||
.split(&b)
|
||||
.filter(|token| !token.is_empty())
|
||||
.collect::<Vec<_>>();
|
||||
if tokens_a.is_empty() || tokens_b.is_empty() {
|
||||
return 0.0;
|
||||
}
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue