feat(emphasis): add chinese punctuations as word's boundary

Tencent · Jun 8, 2022 · 6d8a769 · 6d8a769
1 parent 5a74b0c
commit 6d8a769
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 5 deletions.
diff --git a/src/core/hooks/Emphasis.js b/src/core/hooks/Emphasis.js
@@ -18,7 +18,7 @@ import {
   compileRegExp,
   DO_NOT_STARTS_AND_END_WITH_SPACES_MULTILINE_ALLOW_EMPTY,
   ALLOW_WHITESPACE_MULTILINE,
-  UNDERSCORE_EMPHASIS_BORDER,
+  UNDERSCORE_EMPHASIS_BOUNDARY,
 } from '@/utils/regexp';
 
 export default class Emphasis extends SyntaxBase {
@@ -97,9 +97,9 @@ export default class Emphasis extends SyntaxBase {
 
     // UNDERSCORE_EMPHASIS_BORDER：允许除下划线以外的「标点符号」和空格出现，使用[^\w\S \t]或[\W\s]会有性能问题
     const underscore = {
-      begin: `(^|${UNDERSCORE_EMPHASIS_BORDER})(_+)`, // ?<leading>, ?<symbol>
+      begin: `(^|${UNDERSCORE_EMPHASIS_BOUNDARY})(_+)`, // ?<leading>, ?<symbol>
       content: `(${REGEX})`, // ?<text>
-      end: `\\2(?=${UNDERSCORE_EMPHASIS_BORDER}|$)`,
+      end: `\\2(?=${UNDERSCORE_EMPHASIS_BOUNDARY}|$)`,
     };
 
     asterisk.reg = compileRegExp(asterisk, 'g');

diff --git a/src/utils/regexp.js b/src/utils/regexp.js
@@ -43,16 +43,25 @@ export const NOT_ALL_WHITE_SPACES_INLINE = '(?:[^\\n]*?\\S[^\\n]*?)';
 
 export const NORMAL_INDENT = '[ ]{0, 3}|\\t';
 export const NO_BACKSLASH_BEFORE_CAPTURE = '[^\\\\]';
+
 // https://spec.commonmark.org/0.29/#ascii-punctuation-character
 // !, ", #, $, %, &, ', (, ), *, +, ,, -, ., / (U+0021–2F),
 // :, ;, <, =, >, ?, @ (U+003A–0040),
 // [, \, ], ^, _, ` (U+005B–0060),
 // {, |, }, or ~ (U+007B–007E).
 export const PUNCTUATION = '[\\u0021-\\u002F\\u003a-\\u0040\\u005b-\\u0060\\u007b-\\u007e]';
 
+// extra punctuations
+export const CHINESE_PUNCTUATION = '[！“”¥‘’（），。—：；《》？【】「」·～｜]';
+
 // 下划线强调语法允许的边界符号
-export const UNDERSCORE_EMPHASIS_BORDER =
-  '[\\u0021-\\u002F\\u003a-\\u0040\\u005b\\u005d\\u005e\\u0060\\u007b-\\u007e \\t\\n]';
+export const UNDERSCORE_EMPHASIS_BOUNDARY =
+  '[' +
+  '\\u0021-\\u002F\\u003a-\\u0040\\u005b\\u005d\\u005e\\u0060\\u007b-\\u007e' + // punctuations defined in commonmark
+  ' ' +
+  '\\t\\n' +
+  '！“”¥‘’（），。—：；《》？【】「」·～｜' + // chinese punctuations
+  ']';
 
 // https://html.spec.whatwg.org/multipage/input.html#e-mail-state-(type%3Demail)
 export const EMAIL_INLINE = new RegExp(