#! /usr/bin/env ruby

require 'rubygems'
gem 'nokogiri', '>=1.0.6'
gem 'hpricot', '>=0.6.170'

require 'open-uri'
require 'benchmark'
require 'nokogiri'
require 'hpricot'

[
 [1000, "#{File.dirname(__FILE__)}/sample_post.html"],
 [10, "http://slashdot.com/"],
].each do |ntimes, uri|
  
  html = open(uri).read
  summary = []

  puts "For an html snippet #{html.size} bytes long ..."
  Benchmark.bm(20) do |x|
    x.report("regex * #{ntimes}") do
      ntimes.times do |j|
        html.gsub(/<a\s+(.*)>(.*)<\/a>/i, '<a \1><span>\2</span></a>') # broken regex
        html.gsub(/<(script|noscript|object|embed|style|frameset|frame|iframe)[>\s\S]*<\/\1>/, '')
        html
      end
    end
    
    stime = Time.now
    x.report("nokogiri * #{ntimes}") do
      ntimes.times do
        doc = Nokogiri::HTML(html)
        doc.search("a/text()").wrap("<span></span>")
        doc.search("script","noscript","object","embed","style","frameset","frame","iframe").unlink
        doc.inner_html
      end
    end
    etime = Time.now
    summary << ("it took an average of %.4f seconds for Nokogiri to parse and operate on an HTML snippet #{html.size} bytes long" % ((etime - stime) / ntimes))

    stime = Time.now
    x.report("hpricot * #{ntimes}") do
      ntimes.times do
        doc = Hpricot(html)
        doc.search("a/text()").wrap("<span></span>")
        doc.search(["script","noscript","object","embed","style","frameset","frame","iframe"]).remove
        doc.inner_html
      end
    end
    etime = Time.now
    summary << ("it took an average of %.4f seconds for Hpricot to parse and operate on an HTML snippet #{html.size} bytes long" % ((etime - stime) / ntimes))
  end

  puts
  puts summary
  puts
end