require 'rubygems'
require 'hpricot'
require 'net/http'
Net::HTTP.version_1_2

# const
HttpString = 'http://'
TwitterAddress = 'twitter.com'
PathOfFavorite = '/favourites?page='
PathOfStatus = '/statuses/'

# arg
if ARGV.length < 2 then
  printf("usage: ruby favorite2.rb username password [max_page=1000]\n")
  abort
end
$username = ARGV[0]
$password = ARGV[1]
if ARGV.length < 3 then
  $max_page = 1000
else
  $max_page = ARGV[2].to_i
end
if ARGV.length < 4 then
  $target = ARGV[0]
else
  $target = ARGV[3]
end

# parse
i = 1
while i<=$max_page
  begin
    path = '/'+$target+PathOfFavorite+i.to_s
    req = Net::HTTP::Get.new(path)
    req.basic_auth($username, $password)
    Net::HTTP.start(TwitterAddress){|http|
      body = http.request(req).body
      doc = Hpricot(body)
      table = doc.at('table')
      begin
        table.search('tr').each{|tr|
          begin
            content = tr.at("td[@class~='content']")
            screen_name = content.at('strong').at('a').inner_text
            entry_content = content.at("span[@class~='entry-content']").inner_text
            entry_content.gsub!(/\n/, '')
            entry_content.gsub!(/\t/, '')
            entry_date = content.at("a[@class~='entry-date']")
            url = entry_date['href']
            abbr = content.at('abbr')['title']
            printf("\"%s\",\"%s\",\"%s\",\"%s\"\n", screen_name, entry_content, abbr, url)
          rescue
            p $!
          end
        }
      rescue
        p $!
      end
    }
  rescue
    p $!
  end
  i+=1
end