a script for fetching and parsing hostel list from yhachina.com then process it into a serialized json file

hostel_list_china.rb

require 'nokogiri'
require 'open-uri'

base = 'http://www.yhachina.com/3g/'

def nopen(page)
  Nokogiri::HTML(open(page))
end

doc = nopen(base + 'allthehostel.html')

uri_list = doc.xpath('//table//a/@href').map do |href|
  href.value
end

hostels = uri_list.map do |l|
  h = nopen(l)

  h_pic_url = h.xpath('//img[contains(@src, "imgls")]/@src').first.value
  h_title = h.xpath('//td[@class="hostel_t"]/text()').first.value

  h_meta = h.xpath('//img[contains(@src, "images/hostel")]/following-sibling::text()').map do |t|
    t.text
  end.tap {|x| x[0].chomp!(' (');x.delete_at(1)}

  h_desc = h.xpath('//*[@class="dotted"]/text()').first.text
  h_more_urls = h.xpath('//table[3]//a/@href').map {|h| base + h.value}

  h_facilities = nopen(h_more_urls[0]).xpath('//*[@class="administrative"]/text()').first.text
  h_roomtypes_urls = nopen(h_more_urls[0]).xpath('//table[3]//a/@href').map {|h| base + h.value}.uniq!
end