fix: fix strikethrough flanking rules (#3882)

UziTech · web-flow · commit 1a5b124c0ce8 · 2026-02-11T09:37:41.000-07:00
diff --git a/src/Lexer.ts b/src/Lexer.ts
@@ -408,7 +408,7 @@ export class _Lexer<ParserOutput = string, RendererOutput = string> {
       }
 
       // del (gfm)
-      if (token = this.tokenizer.del(src)) {
+      if (token = this.tokenizer.del(src, maskedSrc, prevChar)) {
         src = src.substring(token.raw.length);
         tokens.push(token);
         continue;
diff --git a/src/Tokenizer.ts b/src/Tokenizer.ts
@@ -818,15 +818,56 @@ export class _Tokenizer<ParserOutput = string, RendererOutput = string> {
     }
   }
 
-  del(src: string): Tokens.Del | undefined {
-    const cap = this.rules.inline.del.exec(src);
-    if (cap) {
-      return {
-        type: 'del',
-        raw: cap[0],
-        text: cap[2],
-        tokens: this.lexer.inlineTokens(cap[2]),
-      };
+  del(src: string, maskedSrc: string, prevChar = ''): Tokens.Del | undefined {
+    let match = this.rules.inline.delLDelim.exec(src);
+    if (!match) return;
+
+    const nextChar = match[1] || '';
+
+    if (!nextChar || !prevChar || this.rules.inline.punctuation.exec(prevChar)) {
+      // unicode Regex counts emoji as 1 char; spread into array for proper count
+      const lLength = [...match[0]].length - 1;
+      let rDelim, rLength, delimTotal = lLength;
+
+      const endReg = this.rules.inline.delRDelim;
+      endReg.lastIndex = 0;
+
+      // Clip maskedSrc to same section of string as src
+      maskedSrc = maskedSrc.slice(-1 * src.length + lLength);
+
+      while ((match = endReg.exec(maskedSrc)) != null) {
+        rDelim = match[1] || match[2] || match[3] || match[4] || match[5] || match[6];
+
+        if (!rDelim) continue;
+
+        rLength = [...rDelim].length;
+
+        if (rLength !== lLength) continue;
+
+        if (match[3] || match[4]) { // found another Left Delim
+          delimTotal += rLength;
+          continue;
+        }
+
+        delimTotal -= rLength;
+
+        if (delimTotal > 0) continue; // Haven't found enough closing delimiters
+
+        // Remove extra characters
+        rLength = Math.min(rLength, rLength + delimTotal);
+        // char length can be >1 for unicode characters
+        const lastCharLength = [...match[0]][0].length;
+        const raw = src.slice(0, lLength + match.index + lastCharLength + rLength);
+
+        // Create del token - only single ~ or double ~~ supported
+        const text = raw.slice(lLength, -lLength);
+        return {
+          type: 'del',
+          raw,
+          text,
+          tokens: this.lexer.inlineTokens(text),
+        };
+      }
     }
   }
 
diff --git a/src/rules.ts b/src/rules.ts
@@ -278,6 +278,11 @@ const _punctuationGfmStrongEm = /(?!~)[\p{P}\p{S}]/u;
 const _punctuationOrSpaceGfmStrongEm = /(?!~)[\s\p{P}\p{S}]/u;
 const _notPunctuationOrSpaceGfmStrongEm = /(?:[^\s\p{P}\p{S}]|~)/u;
 
+// GFM allows * and _ inside strikethrough
+const _punctuationGfmDel = /(?![*_])[\p{P}\p{S}]/u;
+const _punctuationOrSpaceGfmDel = /(?![*_])[\s\p{P}\p{S}]/u;
+const _notPunctuationOrSpaceGfmDel = /(?:[^\s\p{P}\p{S}]|[*_])/u;
+
 // sequences em should skip over [title](link), `code`, <html>
 const blockSkip = edit(/link|precode-code|html/, 'g')
   .replace('link', /\[(?:[^\[\]`]|(?<a>`+)[^`]+\k<a>(?!`))*?\]\((?:\\[\s\S]|[^\\\(\)]|\((?:\\[\s\S]|[^\\\(\)])*\))*\)/)
@@ -332,6 +337,27 @@ const emStrongRDelimUnd = edit(
   .replace(/punct/g, _punctuation)
   .getRegex();
 
+// Tilde left delimiter for strikethrough (similar to emStrongLDelim for asterisk)
+const delLDelim = edit(/^~~?(?:((?!~)punct)|[^\s~])/, 'u')
+  .replace(/punct/g, _punctuationGfmDel)
+  .getRegex();
+
+// Tilde delimiter patterns for strikethrough (similar to asterisk)
+const delRDelimCore =
+  '^[^~]+(?=[^~])' // Consume to delim
++ '|(?!~)punct(~~?)(?=[\\s]|$)' // (1) #~~ can only be a Right Delimiter
++ '|notPunctSpace(~~?)(?!~)(?=punctSpace|$)' // (2) a~~#, a~~ can only be a Right Delimiter
++ '|(?!~)punctSpace(~~?)(?=notPunctSpace)' // (3) #~~a, ~~a can only be Left Delimiter
++ '|[\\s](~~?)(?!~)(?=punct)' // (4) ~~# can only be Left Delimiter
++ '|(?!~)punct(~~?)(?!~)(?=punct)' // (5) #~~# can be either Left or Right Delimiter
++ '|notPunctSpace(~~?)(?=notPunctSpace)'; // (6) a~~a can be either Left or Right Delimiter
+
+const delRDelim = edit(delRDelimCore, 'gu')
+  .replace(/notPunctSpace/g, _notPunctuationOrSpaceGfmDel)
+  .replace(/punctSpace/g, _punctuationOrSpaceGfmDel)
+  .replace(/punct/g, _punctuationGfmDel)
+  .getRegex();
+
 const anyPunctuation = edit(/\\(punct)/, 'gu')
   .replace(/punct/g, _punctuation)
   .getRegex();
@@ -389,6 +415,8 @@ const inlineNormal = {
   br,
   code: inlineCode,
   del: noopTest,
+  delLDelim: noopTest,
+  delRDelim: noopTest,
   emStrongLDelim,
   emStrongRDelimAst,
   emStrongRDelimUnd,
@@ -427,6 +455,8 @@ const inlineGfm: Record<InlineKeys, RegExp> = {
   ...inlineNormal,
   emStrongRDelimAst: emStrongRDelimAstGfm,
   emStrongLDelim: emStrongLDelimGfm,
+  delLDelim,
+  delRDelim,
   url: edit(/^((?:protocol):\/\/|www\.)(?:[a-zA-Z0-9\-]+\.?)+[^\s<]*|^email/)
     .replace('protocol', _caseInsensitiveProtocol)
     .replace('email', /[A-Za-z0-9._+-]+(@)[a-zA-Z0-9-_]+(?:\.[a-zA-Z0-9-_]*[a-zA-Z0-9])+(?![-_])/)
diff --git a/test/specs/new/del_flanking.html b/test/specs/new/del_flanking.html
@@ -0,0 +1,5 @@
+<p>~1,~2</p>
+<p>Between ~1.34 and ~5.4</p>
+<p>Between (~1.34) and (~5.4)</p>
+<p>~ test~</p>
+<p>~test ~</p>
diff --git a/test/specs/new/del_flanking.md b/test/specs/new/del_flanking.md
@@ -0,0 +1,9 @@
+~1,~2
+
+Between ~1.34 and ~5.4
+
+Between (~1.34) and (~5.4)
+
+~ test~
+
+~test ~
diff --git a/test/specs/new/emoji_strikethrough.html b/test/specs/new/emoji_strikethrough.html
@@ -0,0 +1,16 @@
+<p><del>test 💁</del></p>
+<p><del>💁 test</del></p>
+<p><del>🤓 test</del></p>
+<p><del>🏖️ test</del></p>
+<p><del>🏖️🤓💁 test</del></p>
+<p><del>💁 test</del> test</p>
+<p>test <del>💁 test</del></p>
+<p>test <del>💁 test</del> test</p>
+<p>~~💁 ~~</p>
+<p>~~ 💁~~</p>
+<p><del>⚠️ test</del></p>
+<p><del>💁 test</del></p>
+<p><del>t💁t</del> test</p>
+<p><del>t💁t</del> test</p>
+<p>~1 😁~2</p>
+<p><del>🏴‍☠️</del>  test</p>
diff --git a/test/specs/new/emoji_strikethrough.md b/test/specs/new/emoji_strikethrough.md
@@ -0,0 +1,31 @@
+~~test 💁~~
+
+~~💁 test~~
+
+~~🤓 test~~
+
+~~🏖️ test~~
+
+~~🏖️🤓💁 test~~
+
+~~💁 test~~ test
+
+test ~~💁 test~~
+
+test ~~💁 test~~ test
+
+~~💁 ~~
+
+~~ 💁~~
+
+~~⚠️ test~~
+
+~💁 test~
+
+~t💁t~ test
+
+~~t💁t~~ test
+
+~1 😁~2
+
+~~🏴‍☠️~~ test

Original file line number	Diff line number	Diff line change
`@@ -408,7 +408,7 @@ export class _Lexer<ParserOutput = string, RendererOutput = string> {`
`408`	`408`	`}`
`409`	`409`
`410`	`410`	`// del (gfm)`
`411`		`- if (token = this.tokenizer.del(src)) {`
	`411`	`+ if (token = this.tokenizer.del(src, maskedSrc, prevChar)) {`
`412`	`412`	`src = src.substring(token.raw.length);`
`413`	`413`	`tokens.push(token);`
`414`	`414`	`continue;`
-Original file line number
+Diff line change
@@ @@ -0,0 +1,9 @@ @@
 +~1,~2
++
 +Between ~1.34 and ~5.4
++
 +Between (~1.34) and (~5.4)
++
 +~ test~
++
 +~test ~