Sotheby's full-size image ripper

sothebys.rb

#!/usr/bin/env ruby

# Sothebys full-size image ripper.
# Gets the full-size images from the lots listed after __END__.
# By Henrik Nyh <http://henrik.nyh.se> 2010-12-23 under the MIT license.

require "open-uri"

# We get an error page if we do not specify a user agent.
USER_AGENT = "Mozilla/5.0 (compatible; MSIE 8.0; Windows NT 5.2; Trident/4.0; Media Center PC 4.0; SLCC1; .NET CLR 3.0.04320)"

def read_url(url)
  open(url, "User-Agent" => USER_AGENT).read
end

DATA.read.each_line do |url|
  html = read_url(url)
  image_base_url = html[%r{(http://s7d2\.scene7\.com/is/image/Sothebys/.+?)\?\$}, 1]
  local_file = "/tmp/sothebys-#{File.basename(image_base_url)}.jpg"
  # We're not allowed to request larger images than 4000 * 4000 px.
  image_url = "#{image_base_url}?req=tile&rect=0,0,4000,4000"

  File.open(local_file, "w") { |f| f.write read_url(image_url) }
  # ImageMagick - trim background color around image
  system("convert", local_file, "-trim", local_file)

  puts "Done: #{local_file}"
end

__END__
http://www.sothebys.com/app/live/lot/LotDetail.jsp?lot_id=159428750
http://www.sothebys.com/app/live/lot/LotDetail.jsp?lot_id=159377074