fix: handle full CJK Unified Ideographs in search terms, closes #989 (#1002)

apps/readest-app/src/app/reader/components/sidebar/SearchBar.tsx

@@ -9,6 +9,7 @@ import { useReaderStore } from '@/store/readerStore';
 import { useTranslation } from '@/hooks/useTranslation';
 import { BookSearchConfig, BookSearchResult } from '@/types/book';
 import { useResponsiveSize } from '@/hooks/useResponsiveSize';
+import { isCJKStr } from '@/utils/lang';
 import Dropdown from '@/components/Dropdown';
 import SearchOptions from './SearchOptions';
 
@@ -104,8 +105,9 @@ const SearchBar: React.FC<SearchBarProps> = ({
   };
 
   const exceedMinSearchTermLength = (searchTerm: string) => {
-    const isCJK = /[\u4e00-\u9fa5\u3040-\u30ff\uac00-\ud7af]/.test(searchTerm);
-    const minLength = isCJK ? MINIMUM_SEARCH_TERM_LENGTH_CJK : MINIMUM_SEARCH_TERM_LENGTH_DEFAULT;
+    const minLength = isCJKStr(searchTerm)
+      ? MINIMUM_SEARCH_TERM_LENGTH_CJK
+      : MINIMUM_SEARCH_TERM_LENGTH_DEFAULT;
 
     return searchTerm.length >= minLength;
   };

apps/readest-app/src/app/reader/components/sidebar/SearchResults.tsx

@@ -32,7 +32,7 @@ const SearchResultItem: React.FC<SearchResultItemProps> = ({
     >
       <div className='line-clamp-3'>
         <span className=''>{excerpt.pre}</span>
-        <span className='font-semibold'>{excerpt.match}</span>
+        <span className='font-bold text-red-500'>{excerpt.match}</span>
         <span className=''>{excerpt.post}</span>
       </div>
     </li>

apps/readest-app/src/services/tts/WebSpeechClient.ts

@@ -2,6 +2,7 @@ import { getUserLocale } from '@/utils/misc';
 import { TTSClient, TTSMessageEvent, TTSVoice } from './TTSClient';
 import { AsyncQueue } from '@/utils/queue';
 import { findSSMLMark, parseSSMLLang, parseSSMLMarks } from '@/utils/ssml';
+import { isCJKStr } from '@/utils/lang';
 import { TTSGranularity } from '@/types/view';
 import { TTSUtils } from './TTSUtils';
 
@@ -123,7 +124,7 @@ async function* speakWithMarks(
   const isCJK = (lang: string | null) => {
     const cjkLangs = ['zh', 'ja', 'kr'];
     if (lang && cjkLangs.some((cjk) => lang.startsWith(cjk))) return true;
-    return /[\u4e00-\u9fff\u3040-\u30ff\uac00-\ud7af]/.test(plainText);
+    return isCJKStr(plainText);
   };
 
   if (!isCJK(lang)) {

apps/readest-app/src/utils/lang.ts

@@ -0,0 +1,3 @@
+export const isCJKStr = (str: string) => {
+  return /[\p{Script=Han}\p{Script=Hiragana}\p{Script=Katakana}\p{Script=Hangul}]/u.test(str ?? '');
+};