From 7421ecce679867af4e1345d53fc5a6dffca9311e Mon Sep 17 00:00:00 2001
From: Alex Kocharin <alex@kocharin.ru>
Date: Fri, 5 Jul 2019 17:31:53 +0300
Subject: [PATCH] Improve normalization for reference label matching

Simple toUpperCase() does not work for various ligatures and legacy
characters that are already in a different upper case form.

See also:
https://github.com/commonmark/commonmark-spec/commit/278ea515fe5ff3e0dd652ee2127ffcd5cbc91527
---
 lib/common/utils.js                           | 26 ++++++++++++++++---
 .../markdown-it/commonmark_extras.txt         | 20 ++++++++++++++
 2 files changed, 42 insertions(+), 4 deletions(-)

diff --git a/lib/common/utils.js b/lib/common/utils.js
index a2f9540..17dd817 100644
--- a/lib/common/utils.js
+++ b/lib/common/utils.js
@@ -241,10 +241,28 @@ function isMdAsciiPunct(ch) {
 // Hepler to unify [reference labels].
 //
 function normalizeReference(str) {
-  // use .toUpperCase() instead of .toLowerCase()
-  // here to avoid a conflict with Object.prototype
-  // members (most notably, `__proto__`)
-  return str.trim().replace(/\s+/g, ' ').toUpperCase();
+  // Trim and collapse whitespace
+  //
+  str = str.trim().replace(/\s+/g, ' ');
+
+  // In node v10 'ẞ'.toLowerCase() === 'Ṿ', which is presumed to be a bug
+  // fixed in v12 (couldn't find any details).
+  //
+  // So treat this one as a special case
+  // (remove this when node v10 is no longer supported).
+  //
+  if ('ẞ'.toLowerCase() === 'Ṿ') {
+    str = str.replace(/ẞ/g, 'ß');
+  }
+
+  // .toLowerCase().toUpperCase() should get rid of all differences
+  // between letter variants.
+  //
+  // Final result should be uppercased, because it's later stored in an object
+  // (this avoid a conflict with Object.prototype members,
+  // most notably, `__proto__`)
+  //
+  return str.toLowerCase().toUpperCase();
 }
 
 ////////////////////////////////////////////////////////////////////////////////
diff --git a/test/fixtures/markdown-it/commonmark_extras.txt b/test/fixtures/markdown-it/commonmark_extras.txt
index dcb86a9..369977d 100644
--- a/test/fixtures/markdown-it/commonmark_extras.txt
+++ b/test/fixtures/markdown-it/commonmark_extras.txt
@@ -23,6 +23,26 @@ Issue #55:
 .
 
 
+Reference labels: 'i̇θωkå'.toUpperCase() is 'İΘΩKÅ', but these should still be equivalent
+.
+[İϴΩKÅ]
+
+[i̇θωkå]: /url
+.
+<p><a href="/url">İϴΩKÅ</a></p>
+.
+
+
+Reference labels: support ligatures (equivalent according to unicode case folding)
+.
+[fffifl]
+
+[fffifl]: /url
+.
+<p><a href="/url">fffifl</a></p>
+.
+
+
 Issue #35. `<` should work as punctuation
 .
 an **(:**<br>