Google Search Term > Page positions finder

google_terms_position.rb

require 'rubygems'
require 'nokogiri'
require 'open-uri'
require 'cgi'

searches = [['alf', 'www.hedweb.com/alffaq.htm'], ['bold food', 'http://dockets.justia.com/docket/new-york/nysdce/1:2009cv00440/338836/']]

searches.each do |term, url|

  def search(term, url, start = 0)
    doc   = Nokogiri::HTML(open("http://www.google.co.uk/search?q=#{CGI::escape(term)}&start=#{start}"))
    count = 1
    found = false
    doc.xpath('//h3/a[@class="l"]').each do |link|
      count = start + count
      if "#{link[:href]}" =~ /#{url}/
        found = true
        break
      end
    end    
    return (found) ? count : false
  end

  attempts = 0 
  found    = false
  
  while attempts < 10
    res = search(term, url, 10 * attempts)
    if res != false
      found = res
      break
    else
      attempts = attempts + 1
    end
  end
  
  if found != false
    puts "#{term},#{url},#{found}"
  else
    puts "#{term},#{url},-1"    
  end

end