Switch PUNCTUATION_REGEXP to use RegExp with u directly (#15621)

package.json

@@ -90,7 +90,6 @@
     "postcss-scss": "4.0.9",
     "postcss-selector-parser": "2.2.3",
     "postcss-values-parser": "2.0.1",
-    "regexp-util": "2.0.0",
     "remark-footnotes": "2.0.0",
     "remark-math": "3.0.1",
     "remark-parse": "8.0.3",

src/language-markdown/constants.evaluate.js

@@ -24,32 +24,61 @@ const CJK_REGEXP = new RegExp(
   `(?:${cjkCharset.toString()})(?:${variationSelectorsCharset.toString()})?`,
 );
 
-// http://spec.commonmark.org/0.25/#ascii-punctuation-character
-const asciiPunctuationCharset =
-  /* prettier-ignore */ new Charset(
-  "!", '"', "#",  "$", "%", "&", "'", "(", ")", "*",
-  "+", ",", "-",  ".", "/", ":", ";", "<", "=", ">",
-  "?", "@", "[", "\\", "]", "^", "_", "`", "{", "|",
-  "}", "~"
+const asciiPunctuationCharacters = [
+  "!",
+  '"',
+  "#",
+  "$",
+  "%",
+  "&",
+  "'",
+  "(",
+  ")",
+  "*",
+  "+",
+  ",",
+  "-",
+  ".",
+  "/",
+  ":",
+  ";",
+  "<",
+  "=",
+  ">",
+  "?",
+  "@",
+  "[",
+  "\\",
+  "]",
+  "^",
+  "_",
+  "`",
+  "{",
+  "|",
+  "}",
+  "~",
+];
+
+// https://spec.commonmark.org/0.25/#punctuation-character
+// https://unicode.org/Public/5.1.0/ucd/UCD.html#General_Category_Values
+const unicodePunctuationClasses = [
+  /* Pc */ "Connector_Punctuation",
+  /* Pd */ "Dash_Punctuation",
+  /* Pe */ "Close_Punctuation",
+  /* Pf */ "Final_Punctuation",
+  /* Pi */ "Initial_Punctuation",
+  /* Po */ "Other_Punctuation",
+  /* Ps */ "Open_Punctuation",
+];
+
+const PUNCTUATION_REGEXP = new RegExp(
+  `(?:${[
+    new Charset(...asciiPunctuationCharacters).toRegExp().source,
+    ...unicodePunctuationClasses.map(
+      (charset) => `\\p{General_Category=${charset}}`,
+    ),
+  ].join("|")})`,
+  "u",
 );
 
-// http://spec.commonmark.org/0.25/#punctuation-character
-const unicodePunctuationCharset = unicodeRegex({
-  // http://unicode.org/Public/5.1.0/ucd/UCD.html#General_Category_Values
-  General_Category: [
-    /* Pc */ "Connector_Punctuation",
-    /* Pd */ "Dash_Punctuation",
-    /* Pe */ "Close_Punctuation",
-    /* Pf */ "Final_Punctuation",
-    /* Pi */ "Initial_Punctuation",
-    /* Po */ "Other_Punctuation",
-    /* Ps */ "Open_Punctuation",
-  ],
-});
-
-const PUNCTUATION_REGEXP = new Charset(
-  asciiPunctuationCharset,
-  unicodePunctuationCharset,
-).toRegExp();
-
 export { CJK_REGEXP, PUNCTUATION_REGEXP };

src/language-markdown/printer-markdown.js

@@ -95,7 +95,7 @@ function genericPrint(path, options, print) {
               `(^|${PUNCTUATION_REGEXP.source})(_+)`,
               `(_+)(${PUNCTUATION_REGEXP.source}|$)`,
             ].join("|"),
-            "g",
+            "gu",
           ),
           (_, text1, underscore1, underscore2, text2) =>
             (underscore1

yarn.lock

@@ -7423,7 +7423,6 @@ __metadata:
     postcss-values-parser: "npm:2.0.1"
     prettier: "npm:3.3.2"
     pretty-bytes: "npm:6.1.1"
-    regexp-util: "npm:2.0.0"
     remark-footnotes: "npm:2.0.0"
     remark-math: "npm:3.0.1"
     remark-parse: "npm:8.0.3"
@@ -7631,7 +7630,7 @@ __metadata:
   languageName: node
   linkType: hard
 
-"regexp-util@npm:2.0.0, regexp-util@npm:^2.0.0":
+"regexp-util@npm:^2.0.0":
   version: 2.0.0
   resolution: "regexp-util@npm:2.0.0"
   dependencies: