unicode-friendly ruby whitespace removal

safe_strip.rb

class String
   
   # performance hack - memoize these regexps
   SPACE_8_RE = Regexp.new('[[:space:]]'.encode('UTF-8'))
   SPACE_16_RE = Regexp.new('[[:space:]]'.encode('UTF-16LE'))

   def safe_strip
     if self.encoding=='UTF-8'
       self.gsub SPACE_8_RE, ''
     elsif self.encoding=='UTF-16'
       self.gsub SPACE_16_RE, ''
     else
       self.strip
     end
   end
   
end