print URLs of google search results to stdout

xgoogle.rb

#!/usr/bin/env ruby
#
# print URLs of google search results to stdout
# the name 'xgoogle' is a homage to Peteris Krumins' xgoogle
#
# dependencies:
#    gem install nokogiri
#
# usage:
#    xgoogle <searchquery> [<number-of-results> [<language>]]
#
# this may be useful for downloading top 10 javascript pdf files:
#    xgoogle "javascript filetype:pdf" 10 | xargs wget -qc
#

require 'open-uri'
require 'nokogiri'

q   = ARGV[0].gsub(/ /, '%20')
num = ARGV[1] || '100'
hl  = ARGV[2] || 'en'

serp = Nokogiri::HTML(open("http://www.google.com/search?num=#{num}&hl=#{hl}&q=#{q}", :proxy => "http://your-proxy:8080"))

serp.css("li.g h3 a").each do |ser|
  puts ser['href'].match(/url\?q=(.*)&sa=/)[1]
end

# vim:ft=ruby