aozoraget

#!/usr/bin/env ruby
# coding: utf-8
# usage: download book series from aozora-bunko.
# author: eric
# [todo] - opt-parse
require 'nokogiri'
require 'open-uri'

# variables
author_url = 'http://www.aozora.gr.jp/index_pages/person52.html'
author_name = "紫式部.与謝野晶子訳"
book_title = "源氏物語"
# [todo] - default to '/tmp/aozoraget'
save_dir = '/home/eric/tmp/genji/' # trailing '/' is a must.

page = Nokogiri::HTML open(author_url)
lst = page.css("ol:first-of-type > li")

lst.each do |item|
  t = item.text
  if t =~ /#{book_title}　\d\d.+新字新仮名/
    item_name = t.sub(/^.*(\d\d) (.+)（.*$/) {
                  "#{author_name}.#{book_title}.#$1#$2.zip" }
    print item_name + "\t"
    # get download url
    item_url = item.css("a:first-of-type").attr("href").to_s
                   .sub(/^\.\./, 'http://www.aozora.gr.jp')
    item_dlink_base = item_url.sub(/\/[^\/]+?$/, '')
    item_page = Nokogiri.HTML open(item_url)
    item_dlink = item_page.css('table.download a[href$="zip"]')
                          .attr("href").text.sub(/^\./, item_dlink_base)
    puts item_dlink
    # download
    system %[wget -O "#{save_dir + item_name}" '#{item_dlink}']
  end
end