File: spider.rb

package info (click to toggle)
libwww-mechanize-ruby 1.0.0-1
  • links: PTS, VCS
  • area: main
  • in suites: squeeze
  • size: 956 kB
  • ctags: 883
  • sloc: ruby: 6,621; makefile: 4
file content (11 lines) | stat: -rw-r--r-- 302 bytes parent folder | download 
1
2
3
4
5
6
7
8
9
10
11
 
$:.unshift File.join(File.dirname(__FILE__), "..", "lib")

require 'rubygems'
require 'mechanize'

agent = Mechanize.new
stack = agent.get(ARGV[0]).links
while l = stack.pop
  next unless l.uri.host == agent.history.first.uri.host
  stack.push(*(agent.click(l).links)) unless agent.visited? l.href
end