allow some more chars like qm etc.

skroutz · Apr 2, 2013 · 31de1e8 · 31de1e8
1 parent 99f10e6
commit 31de1e8
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 3 deletions.
diff --git a/lib/truncate_html/html_string.rb b/lib/truncate_html/html_string.rb
@@ -3,7 +3,7 @@ module TruncateHtml
   class HtmlString < String
 
     UNPAIRED_TAGS = %w(br hr img).freeze
-    REGEX = /(?:<script.*>.*<\/script>)+|<\/?[^>]+>|[[[:alpha:]][0-9]\|`~!@#\$%^&*\(\)\-_\+=\[\]{}:;'",\.\/?]+|\s+|[[:punct:]]/.freeze
+    REGEX = /(?:<script.*>.*<\/script>)+|<\/?[^>]+>|[[[:alpha:]][0-9]\|`~!@#\$%^&*\(\)\-_\+=\[\]{}:;'²³§",\.\/?]+|\s+|[[:punct:]]/.freeze
 
     def initialize(original_html)
       super(original_html)

diff --git a/spec/truncate_html/html_string_spec.rb b/spec/truncate_html/html_string_spec.rb
@@ -1,3 +1,4 @@
+#encoding: utf-8
 require File.join(File.dirname(__FILE__), '..', 'spec_helper')
 
 describe TruncateHtml::HtmlString do
@@ -8,8 +9,9 @@ def html_string(original_string)
 
   describe '#html_tokens' do
     it 'returns each token in the string as an array element removing any consecutive whitespace from the string' do
-      html = '<h1>Hi there</h1> <p>This          is sweet!</p>'
-      html_string(html).html_tokens.should == ['<h1>', 'Hi', ' ', 'there', '</h1>', ' ', '<p>', 'This', ' ', 'is', ' ', 'sweet!', '</p>']
+      html = '<h1>Hi there</h1> <p>This          is sweet!</p> <p> squaremeter m² </p>'
+      html_string(html).html_tokens.should == ['<h1>', 'Hi', ' ', 'there', '</h1>', ' ', '<p>', 'This', ' ', 'is', ' ', 'sweet!', '</p>',
+        ' ', '<p>', ' ', 'squaremeter', ' ', 'm²', ' ', '</p>']
     end
   end