behoimi.org ぶっこ抜き

bucko.rb

#!/usr/bin/env ruby
# -*- coding: utf-8 -*-

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

class Bucko

  def initialize
    @root = 'http://behoimi.org'
    @ua = 'Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; ja-JP-mac; rv:1.9.2.7) Gecko/20100713 Firefox/3.6.7'
  end

  def page_open (uri)
    open(URI.encode(uri), 'User-Agent' => @ua).read
  end

  def get_image (uri, ref)
    system "wget --user-agent=\"#{@ua}\" --referer=\"#{ref}\"  \"#{uri}\""
    sleep 1
  end

  def exec (tags, page=1)
    (1..page).each do |p|
      page = Nokogiri::XML(page_open("#{@root}/post/index?page=#{p}&tags=#{tags}"))
      page.xpath("//span[@class=\"thumb blacklisted\"]/a/@href").each do |l|
        get_image(Nokogiri::XML(page_open("#{@root}#{l}")).xpath("//a[@id=\"highres\"]/@href"), "#{@root}#{l}")
      end
    end
  end

end

# example
# Bucko.new().exec('lenfried', 108)