tanner/obsidian-tannersearch
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

perf: move regexp outside func to avoid repeated construction (#430)

Browse Source
This commit is contained in:
YuNing Chen
2025-01-03 19:25:49 +08:00
committed by GitHub
parent e272aeabef
commit c4fa419aa0
1 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
src/tools/utils.ts
View File

@@ -106,6 +106,11 @@ export function getTagsFromMetadata(metadata: CachedMetadata | null): string[] {
return tags
}
// Define cached diacritics regex once outside the function
const japaneseDiacritics = ['\\u30FC', '\\u309A', '\\u3099']
const regexpExclude = japaneseDiacritics.join('|')
const diacriticsRegex = new RegExp(`(?!${regexpExclude})\\p{Diacritic}`, 'gu')
/**
* https://stackoverflow.com/a/37511463
*/
@@ -114,11 +119,6 @@ export function removeDiacritics(str: string, arabic = false): string {
return ''
}
// Japanese diacritics that should be distinguished
const japaneseDiacritics: string[] = ['\\u30FC', '\\u309A', '\\u3099']
const regexpExclude: string = japaneseDiacritics.join('|')
const regexp: RegExp = new RegExp(`(?!${regexpExclude})\\p{Diacritic}`, 'gu')
if (arabic) {
// Arabic diacritics
// https://stackoverflow.com/a/40959537
@@ -139,7 +139,7 @@ export function removeDiacritics(str: string, arabic = false): string {
// Keep caret same as above
str = str.replaceAll('^', '[__omnisearch__caret__]')
// To keep right form of Korean character, NFC normalization is necessary
str = str.normalize('NFD').replace(regexp, '').normalize('NFC')
str = str.normalize('NFD').replace(diacriticsRegex, '').normalize('NFC')
str = str.replaceAll('[__omnisearch__backtick__]', '`')
str = str.replaceAll('[__omnisearch__caret__]', '^')
return str