Ruby的Unicode支持确实是一个问题，尽管$KCODE=’u'可以解决一些字符串打印之类的问题，但是进行Unicode字符串的截取之类稍微复杂点操作就不行了，以前我的办法是，不论什么编码，先用iconv转换成UTF-16，之后截取，再转换回来，能在一定范围内解决一些小问题，现在activerecord (>=1.15.1)提供了 ActiveSupport::Multibyte::Chars类，很好的解决了Unicode的问题,下面是ActiveSupport::Multibyte::Chars用法的一点小例子:

  1. require "rubygems"
  2. require_gem "activesupport"
  3. require "active_support"
  4. $KCODE='u'
  5. s="中文测试"
  6. us=s.chars
  7. # => #<ActiveSupport::Multibyte::Chars:0xb79eb148 @string="中文测试">
  8. s.size 
  9. #=>12
  10. us.size
  11. #=> 4
  12. us.class
  13. #=> ActiveSupport::Multibyte::Chars
  14. us.string
  15. #=> "中文测试"
  16. us[0]
  17. #=> 20013
  18. us[0..0]
  19. #=> #<ActiveSupport::Multibyte::Chars:0xb79de0d8 @string="中">
  20. us[0..2]
  21. #=> #<ActiveSupport::Multibyte::Chars:0xb79e0fa4 @string="中文测">
  22. us.methods.sort
  23. #=> ["<", "<=", "<=>", "==", "===", "=~", ">", ">=", "__id__", "__send__", "`", "between?", "blank?", "class", "clone", "copy_instance_variables_from", "daemonize", "dclone", "display", "dup", "enable_warnings", "eql?", "equal?", "extend", "extend_with_included_modules_from", "extended_by", "freeze", "frozen?", "gem", "gsub", "handler", "hash", "id", "inspect", "instance_eval", "instance_exec", "instance_of?", "instance_values", "instance_variable_get", "instance_variable_set", "instance_variables", "is_a?", "kind_of?", "load", "method", "method_missing", "methods", "nil?", "object_id", "private_methods", "protected_methods", "public_methods", "remove_subclasses_of", "require", "require_gem", "require_library_or_gem", "respond_to?", "returning", "send", "silence_stderr", "silence_stream", "silence_warnings", "singleton_methods", "split", "string", "subclasses_of", "suppress", "taguri", "taguri=", "taint", "tainted?", "to_a", "to_json", "to_s", "to_str", "to_yaml", "to_yaml_properties", "to_yaml_style", "type", "unloadable", "untaint", "with_options"]