Gelbooru 一括ダウンローダー

gelbooru_downloader.rb

#!/usr/bin/ruby

require 'rubygems'
require 'active_support'
require 'mechanize'
require 'optparse'
require 'fileutils'

Version = "1.0.0"
options = { :tag => nil, :dir => 'downloads' }
parser = OptionParser.new do |ps|
  ps.on("-t TAG", "--tag TAG", "fetch tag") { |v| options[:tag] = v  }
  ps.on("-d DIR", "--dir DIR", "save dir") { |v| options[:dir] = v }
  ps.on("-p NUM", "--pid NUM", "start page id (gelbooru style: 25,50,75...)") { |v| options[:page] = v }
  ps.parse!(ARGV)
end

if options[:tag].blank?
  puts parser.help
  exit
end

class Gelbooru
  attr_reader :download_dir
  def initialize(opts = {})
    @agent = Mechanize.new do |a|
      a.user_agent_alias = 'Windows IE 7'
      a.max_history = 1
    end
    @base_url = 'http://gelbooru.com/index.php'
    @download_dir = (opts[:download_dir] or '.')
  end
  def fetch_by_tags(tag, pid = 0)
    search_url = URI.join(@base_url, "?page=post&s=list&tags=#{CGI.escape(tag)}&pid=#{pid}")
    while true
      puts "Open: #{search_url}"
      page = @agent.get(search_url)
      page.root.search('//span[@class="thumb"]/a').each do |elem|
        image_url = URI.join(@base_url, elem['href'])
        puts "Open: #{image_url}"
        image_page = @agent.get(image_url)
        if link = image_page.links.find {|l| l.text.match(/^original image$/i) }
          save_path = File.join(@download_dir, File.basename(link.href))
          if File.exist? save_path
            puts "File already exist: #{save_path}"
          else
            puts "Download: #{link.href}"
            orig_image = @agent.get_file(link.href)
            puts "Save to: #{save_path}"
            File.open(save_path, 'w') do |fh|
              fh << orig_image
            end
          end
        end
      end
      if next_link = page.links.find {|l| 'next' == l.attributes[:alt] }
        search_url = URI.join(@base_url, next_link.href)
      else
        break
      end
    end
  end
end

if not File.exist? options[:dir]
  FileUtils.mkpath options[:dir]
end

g = Gelbooru.new(:download_dir => options[:dir])
g.fetch_by_tags(options[:tag], options[:page])