Merge pull request #41347 from ricardotk002/use-string-grapheme-clusters

Use String#grapheme_clusters instead of String#scan
rails · Feb 5, 2021 · 7f2d44f · 7f2d44f
2 parents c8a86d3 + 95f8bfe
commit 7f2d44f
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 8 deletions.
diff --git a/activesupport/lib/active_support/core_ext/string/filters.rb b/activesupport/lib/active_support/core_ext/string/filters.rb
@@ -106,7 +106,7 @@ def truncate_bytes(truncate_at, omission: "…")
       self.class.new.tap do |cut|
         cut_at = truncate_at - omission.bytesize
 
-        scan(/\X/) do |grapheme|
+        each_grapheme_cluster do |grapheme|
           if cut.bytesize + grapheme.bytesize <= cut_at
             cut << grapheme
           else

diff --git a/activesupport/lib/active_support/multibyte/chars.rb b/activesupport/lib/active_support/multibyte/chars.rb
@@ -103,7 +103,7 @@ def slice!(*args)
       #
       #   'Café'.mb_chars.reverse.to_s # => 'éfaC'
       def reverse
-        chars(@wrapped_string.scan(/\X/).reverse.join)
+        chars(@wrapped_string.grapheme_clusters.reverse.join)
       end
 
       # Limits the byte size of the string to a number of bytes without breaking
@@ -126,26 +126,26 @@ def titleize
 
       # Performs canonical decomposition on all the characters.
       #
-      #   'é'.length                         # => 2
-      #   'é'.mb_chars.decompose.to_s.length # => 3
+      #   'é'.length                         # => 1
+      #   'é'.mb_chars.decompose.to_s.length # => 2
       def decompose
         chars(Unicode.decompose(:canonical, @wrapped_string.codepoints.to_a).pack("U*"))
       end
 
       # Performs composition on all the characters.
       #
-      #   'é'.length                       # => 3
-      #   'é'.mb_chars.compose.to_s.length # => 2
+      #   'é'.length                       # => 1
+      #   'é'.mb_chars.compose.to_s.length # => 1
       def compose
         chars(Unicode.compose(@wrapped_string.codepoints.to_a).pack("U*"))
       end
 
       # Returns the number of grapheme clusters in the string.
       #
       #   'क्षि'.mb_chars.length   # => 4
-      #   'क्षि'.mb_chars.grapheme_length # => 3
+      #   'क्षि'.mb_chars.grapheme_length # => 2
       def grapheme_length
-        @wrapped_string.scan(/\X/).length
+        @wrapped_string.grapheme_clusters.length
       end
 
       # Replaces all ISO-8859-1 or CP1252 characters by their UTF-8 equivalent