Liquid has UTF8 support.

lib/liquid.rb

@@ -20,14 +20,15 @@
 # WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
 module Liquid
+  WordRegex                   = RUBY_VERSION < "1.9" ? '\w' : '[[:word:]]'
   FilterSeparator             = /\|/
   ArgumentSeparator           = ','
   FilterArgumentSeparator     = ':'
   VariableAttributeSeparator  = '.'
   TagStart                    = /\{\%/
   TagEnd                      = /\%\}/
-  VariableSignature           = /\(?[\w\-\.\[\]]\)?/
-  VariableSegment             = /[\w\-]/
+  VariableSignature           = /\(?[#{WordRegex}\-\.\[\]]\)?/o
+  VariableSegment             = /[#{WordRegex}\-]/o
   VariableStart               = /\{\{/
   VariableEnd                 = /\}\}/
   VariableIncompleteEnd       = /\}\}?/
@@ -38,7 +39,7 @@ module Liquid
   OtherFilterArgument         = /#{ArgumentSeparator}(?:#{StrictQuotedFragment})/o
   SpacelessFilter             = /^(?:'[^']+'|"[^"]+"|[^'"])*#{FilterSeparator}(?:#{StrictQuotedFragment})(?:#{FirstFilterArgument}(?:#{OtherFilterArgument})*)?/o
   Expression                  = /(?:#{QuotedFragment}(?:#{SpacelessFilter})*)/o
-  TagAttributes               = /(\w+)\s*\:\s*(#{QuotedFragment})/o
+  TagAttributes               = /(#{WordRegex}+)\s*\:\s*(#{QuotedFragment})/o
   AnyStartingTag              = /\{\{|\{\%/
   PartialTemplateParser       = /#{TagStart}.*?#{TagEnd}|#{VariableStart}.*?#{VariableIncompleteEnd}/o
   TemplateParser              = /(#{PartialTemplateParser}|#{AnyStartingTag})/o

lib/liquid/block.rb

@@ -3,7 +3,7 @@ module Liquid
   class Block < Tag
     IsTag             = /^#{TagStart}/o
     IsVariable        = /^#{VariableStart}/o
-    FullToken         = /^#{TagStart}\s*(\w+)\s*(.*)?#{TagEnd}$/o
+    FullToken         = /^#{TagStart}\s*(#{WordRegex}+)\s*(.*)?#{TagEnd}$/o
     ContentOfVariable = /^#{VariableStart}(.*)#{VariableEnd}$/o
 
     def parse(tokens)

lib/liquid/htmltags.rb

@@ -1,6 +1,6 @@
 module Liquid
   class TableRow < Block
-    Syntax = /(\w+)\s+in\s+(#{QuotedFragment}+)/o
+    Syntax = /(#{WordRegex}+)\s+in\s+(#{QuotedFragment}+)/o
 
     def initialize(tag_name, markup, tokens)
       if markup =~ Syntax

lib/liquid/tags/capture.rb

@@ -12,7 +12,7 @@ module Liquid
   # in a sidebar or footer.
   #
   class Capture < Block
-    Syntax = /(\w+)/
+    Syntax = /(#{WordRegex}+)/o
 
     def initialize(tag_name, markup, tokens)
       if markup =~ Syntax

lib/liquid/tags/for.rb

@@ -44,7 +44,7 @@ module Liquid
   # forloop.last:: Returns true if the item is the last item.
   #
   class For < Block                                             
-    Syntax = /(\w+)\s+in\s+(#{QuotedFragment}+)\s*(reversed)?/o
+    Syntax = /(#{WordRegex}+)\s+in\s+(#{QuotedFragment}+)\s*(reversed)?/ou
   
     def initialize(tag_name, markup, tokens)
       if markup =~ Syntax

lib/liquid/variable.rb

@@ -23,7 +23,7 @@ module Liquid
         if match[2].match(/#{FilterSeparator}\s*(.*)/o)
           filters = Regexp.last_match(1).scan(FilterParser)
           filters.each do |f|
-            if matches = f.match(/\s*(\w+)/)
+            if matches = f.match(/\s*(#{WordRegex}+)/o)
               filtername = matches[1]
               filterargs = f.scan(/(?:#{FilterArgumentSeparator}|#{ArgumentSeparator})\s*(#{QuotedFragment})/o).flatten
               @filters << [filtername.to_sym, filterargs]