fix: use emoji regex to correctly mask pair emojis

Author: mhyassin

package-lock.json

@@ -60,6 +60,7 @@
     "commonmark": "0.31.2",
     "cross-env": "^10.1.0",
     "dts-bundle-generator": "^9.5.1",
+    "emoji-regex": "^10.6.0",
     "esbuild": "^0.27.0",
     "esbuild-plugin-umd-wrapper": "^3.0.0",
     "eslint": "^9.39.1",

package.json

@@ -1,3 +1,4 @@
+import emojiRegex from 'emoji-regex';
 import { _Tokenizer } from './Tokenizer.ts';
 import { _defaults } from './defaults.ts';
 import { other, block, inline } from './rules.ts';
@@ -330,6 +331,9 @@ export class _Lexer<ParserOutput = string, RendererOutput = string> {
     // Mask out blocks from extensions
     maskedSrc = this.options.hooks?.emStrongMask?.call({ lexer: this }, maskedSrc) ?? maskedSrc;
 
+    const emojiMatcher = emojiRegex();
+    maskedSrc = maskedSrc.replace(emojiMatcher, (m) => 'a'.repeat(m.length));
+
     let keepPrevChar = false;
     let prevChar = '';
     while (src) {