scrubyt ajax scraping

require 'rubygems'
require 'scrubyt'

# I installed firewatir on my firefox 3.0
# And running "ruby merb_readme.rb" lauches firefox and then go to the right page.
property_data = Scrubyt::Extractor.define :agent => :firefox do
  fetch 'http://merbivore.com/documentation/current/doc/rdoc/stack/'

  # There's also another 'files' link in the page.
  # So I had to write exact location of the link.
  # But this line does not work as I expected.
  # I did googled for this but in vain.
  # click_link "files"
  click_link "//ul#groupType/li[3]"

  # Actually, I want scrap whole readme in haml format.
  # But nokogiri can not pass ajax like links.
  doc "/div#rdocContent'" do
    header "/div[@id='fileHeader']"
    content "/div[@id='bodyContent']"
  end
end
puts property_data.to_xml