utf8_string_scan.rb

def scan_for(matcher, text)
  scanner = StringScanner.new(text)
  last_byte, last_char = 0, 0
  while scanner.scan_until(matcher)
    byte    = scanner.pos - scanner.matched.length
    char    = last_char + text[last_byte..byte].unpack('U*').length - 1
    length  = scanner.matched.unpack('U*').length
    last_byte, last_char = byte, char
    # Do something with the "char" offset and "length"...
  end
end