MCPcopy Index your code
hub / github.com/ChatLab/ChatLab / cleanText

Function cleanText

packages/core/src/nlp/text-utils.ts:105–119  ·  view source on GitHub ↗
(text: string)

Source from the content-addressed store, hash-verified

103 * 清理文本:移除 URL、@提及、表情、标点等
104 */
105export function cleanText(text: string): string {
106 return text
107 .replace(URL_REGEX, ' ')
108 .replace(MENTION_REGEX, ' ')
109 .replace(SYSTEM_PLACEHOLDER_REGEX, ' ')
110 .replace(BRACKET_EMOJI_PLACEHOLDER_REGEX, (match, name: string) => {
111 if (BRACKET_EMOJI_MAP[name]) return BRACKET_EMOJI_MAP[name]
112 return CJK_TEXT_REGEX.test(name) ? ' ' : match
113 })
114 .replace(EMOJI_REGEX, ' ')
115 .replace(EMOJI_VARIATION_SELECTOR_REGEX, ' ')
116 .replace(PUNCTUATION_REGEX, ' ')
117 .replace(/\s+/g, ' ')
118 .trim()
119}
120
121/**
122 * 判断是否为有效词语

Callers 6

text-utils.test.tsFile · 0.90
segmentChineseFunction · 0.90
segmentEnglishFunction · 0.90
segmentJapaneseFunction · 0.90
collectPosTagStatsFunction · 0.90

Calls

no outgoing calls

Tested by

no test coverage detected