From 46f60b5967b9a4f93db7c43f82becec56ed60664 Mon Sep 17 00:00:00 2001
From: Mike Dalessio <mike.dalessio@gmail.com>
Date: Wed, 19 Jan 2022 23:38:44 -0500
Subject: [PATCH] test: failing multibyte unicode test

From #2410
---
 test/xml/test_c14n.rb | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/test/xml/test_c14n.rb b/test/xml/test_c14n.rb
index ccab61cedd3..379e60f0aff 100644
--- a/test/xml/test_c14n.rb
+++ b/test/xml/test_c14n.rb
@@ -1,3 +1,4 @@
+# coding: utf-8
 # frozen_string_literal: true
 
 require "helper"
@@ -237,6 +238,21 @@ def test_wrong_params
         assert_raises(TypeError) { doc.canonicalize(nil, :wrong_type) }
         doc.canonicalize(nil, nil, :wrong_type)
       end
+
+      def test_multibyte_unicode
+        # https://github.com/sparklemotion/nokogiri/issues/2410
+        doc = Nokogiri.XML(%{<foo>𡏅</foo>}, nil, "EUC-JP")
+
+        # I do not understand what's going on here
+        expected = if Nokogiri.jruby?
+          %{<foo>𡏅</foo>}
+        else
+          %{<foo>陝</foo>}
+        end
+
+        result = doc.canonicalize
+        assert_equal(expected, result)
+      end
     end
   end
 end