Cleaning HTML by using Chrome via Selenium

clean_html.rb

#!/usr/bin/env ruby
require "selenium-webdriver"
require "open-uri"
require "uri"
require "tempfile"

url = nil
if STDIN.tty?  
  url = ARGV.shift
  unless url
    usage = <<-EOS
    Usage:
    ./clean_html http://example.com/
    ./clean_html index.html
    cat index.html | ./clean_html
    EOS
    STDERR.puts usage.gsub(/^ +/, "")
    exit 1
  end
  if File.exists?(url)
    url = "file://" + File.expand_path(url)
  end
else
  file = Tempfile.new("clean_html")
  file.write STDIN.read
  file.close
  url = "file://#{file.path}"
end

HTML_REBUILDER_URL = "https://raw.github.com/labocho/html_rebuilder/master/js/html_rebuilder.min.js"
html_rebuilder_src = open(HTML_REBUILDER_URL){|f| f.read }

driver = Selenium::WebDriver.for :chrome

begin
  driver.navigate.to url
  driver.script(html_rebuilder_src)
  puts driver.script("return (new HtmlRebuilder(document)).html();")
ensure
  driver.close
end