新潮文庫の夏の１００冊公式サイトから、対象書籍の読書メーターのページURLを取得するスクリプト。

get_shincho100_titles.rb

# -*- coding: utf-8 -*-

require 'rubygems'
require 'open-uri'

id_list = []
bookmeter_100satsu_url = Hash.new

open("http://100satsu.com") do |f|
  f.each_line do |line|
    if /class=\"book(\d\d\d\d\d\d)\"/ =~ line
      id_list << $1.to_i
    end
  end
  
  p id_list
  p id_list.size

  title = nil 
  asin = nil
  
  id_list.each do |id|
    list_url = "http://www.shinchosha.co.jp/order/" + "#{id}"
    open(list_url) do |f|
      f.each_line do |line|
        if /<title>(.+)｜新潮社<\/title>/ =~ line.force_encoding("UTF-8")
          title = $1
        end
        if /amazon\.co\.jp\/gp\/product\/(\w\w\w\w\w\w\w\w\w\w)/ =~ line
          asin = $1
        end
        bookmeter_url = "http://book.akahoshitakuya.com/b/" + "#{asin}"
        bookmeter_100satsu_url[title] = bookmeter_url
      end
    end
  end
end

bookmeter_100satsu_url.each do |key, value|
  if key != nil
    puts "#{key}, #{value}"
  end
end