gistfile1.rb

# -*- encoding: utf-8 -*-

require 'benchmark'
require 'rubygems'
require 'active_support'
require 'oniguruma'
include Oniguruma

SKIP_INVALID_UTF8 = Iconv.new('UTF-8//IGNORE', 'UTF-8')
SIZE_R = ORegexp.new('.', '', 'utf8')
SLICE_R = ORegexp.new('.{,10}', '', 'utf8')

class String
  def ord_u_all
    # Ingore invalid UTF-8 fix from http://po-ru.com/diary/fixing-invalid-utf-8-in-ruby-revisited/
    SKIP_INVALID_UTF8.iconv(self + ' ')[0..-2].unpack('U*')
  end
end

n = 100_000
text = '北朝鮮、寧辺で核再処理開始か　韓国政府が分析'

puts "\n> Size"

Benchmark.bm(10) do  |x|
  x.report('mb_chars') { n.times { text.mb_chars.size } }
  x.report('ord_u_all') { n.times { text.ord_u_all.size } }
  x.report('regexp') { n.times { SIZE_R.scan(text).size } }
end

puts "\n> Index"

Benchmark.bm(10) do  |x|
  x.report('mb_chars') { n.times { text.mb_chars[0..10] } }
  x.report('ord_u_all') { n.times { text.ord_u_all[0..10] } }
  x.report('regexp') { n.times { SLICE_R.match(text)[0] } }
end