google search by image

eikon.rb

require 'rubygems'
require 'net/http'
require 'open-uri'
require 'cgi'
require 'nokogiri'



def search_by_image_results_page(image)
  sbi_url = "http://www.google.com/searchbyimage?sbisrc=cr_1_0_0&image_url=#{CGI.escape(image)}"
  response = Net::HTTP.get_response(URI.parse(sbi_url))
  response['location'] #google always makes a 302 redirect for the results
end

def parse_results(url)
  #Fake a browser because google redirects you to google search if there is no user agent
  chrome_user_agent = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_6_8) AppleWebKit/535.1 (KHTML, like Gecko) Chrome/14.0.803.0 Safari/535.1"
  Nokogiri::HTML(open(url, "User-Agent" => chrome_user_agent))
end

def report(set)
  puts "\n"*3
  unless set.size == 0
    puts "The image [#{@image_url}] is used on the following urls:"
    set.each_with_index do |link, idx| 
      puts "   ##{idx+1} - #{link.text}: #{link['href']}"
    end
  end
  puts "\n"*3
end



# The image we want to check
@image_url = "http://www.ikbrunel.org.uk/userFiles/GreatBritainWeb.jpg"

results_page = search_by_image_results_page(@image_url)
parsed_page  = parse_results(results_page)
links        = parsed_page.search('//li[@class="g"]//h3[@class="r"]//a') # quick, dirty and fragile

report(links)

Example output.txt

[aitor@banba:~/workshop/github/aitor/eikon 12:45:34] ruby eikon.rb



The image [http://www.ikbrunel.org.uk/userFiles/GreatBritainWeb.jpg] is used on the following urls:
   #1 - Brunel's ss Great Britain - a great day out in Bristol: http://www.ssgreatbritain.org/
   #2 - SS Great Britain - Wikipedia, the free encyclopedia: http://en.wikipedia.org/wiki/SS_Great_Britain
   #3 - Visually similar images: /search?hl=en&tbs=simg:CAESVhpUCxCwjKcIGjgKNggBEhAyO6IBpALWAXFalgHHA8ICGiC7Z4CRSqBwRbPycayBw-pY1paEp0-Eh4efms4VtdVSZgwLEI6u_1ggaCgoICAESBKuhSWcM&q=ss+great+britain&tbm=isch&sa=X&ei=efAaTtW7EMSu8gPkhpwO&ved=0CC8Qsw4
   #4 - MyBrunel.co.uk :: Great Britain :: © 2011: http://www.mybrunel.co.uk/ships/britain/index.php
   #5 - SS (Steam Ship) Great Britain: http://www.crwflags.com/fotw/flags/gb~ssgb.html
   #6 - Brunel 200 Legacy: The Brunel Banquet: http://www.brunel200.com/legacy/brunel_200_events/banquet.htm
   #7 - victorian bristol | ss great britian | Event view: http://www.xtimeline.com/evt/view.aspx?id=73970
   #8 - SS Great Britain hakkında ansiklopedik bilgi: http://www.turkcebilgi.com/ss_great_britain/ansiklopedi