require 'rubygems'
#needs version >= 0.5.00 of hpricot
#works with version 6.0 ruby
require 'hpricot'
require 'open-uri'
require 'uri'
require 'pp'
require "rexml/document"
require "date"
require "yaml"

#http://thepiratebay.org/s/?q=cow&searchTitle=on&page=0&orderby=99

#read about my suggestion @ http://code.whytheluckystiff.net/hpricot/ticket/37
#this will just run next node X times
#node_at just did not float my boat.
module Hpricot
	module Traverse
		# Returns the node neighboring this node to the south: just below it.
		# This method includes text nodes and comments and such.
		def next_node(loop=1)
			sib = parent.children
			sib[ sib.index(self) + loop ] if parent
		end
	end
end

module PirateBay

	BASE_URL = URI::parse("http://thepiratebay.org/")

	def self.find(search_term,type)
	end

	def self.get_browse_categories(yaml_file = nil)
		file = yaml_file ||File.join(File.dirname(__FILE__),"config","pirate_categories.yml")
		raise "No categories file" unless File.exists?(file)
		file_socket = File.open(file)
		YAML.load(file_socket).delete_if{|k,v| k =~ /^_/}
	end

	def self.top_100

	end

	def self.recent(page=1)
		html = ""
		page -= 1
		uri = BASE_URL.clone
		uri.path = "/recent/#{page}"

		html = Hpricot(open(uri.to_s))
		torrents = Array.new
		html.search("table[@id='searchResult']/tr").each{|tr|
			if  tr.search("img[@alt='Next']").first
				next
			else
				torrents.push(Torrent.process_from_table(tr))
			end
		}
		torrents
	end

	def self.tag_cloud
	end

	def self.search_cloud
	end

	def self.legal_threats
	end

	class Torrent
		def self.process_from_table(html)
			torrent_hash = {}
			url_to_category = html.search("td/a[@title='More from this category']").first.attributes["href"]
			torrent_hash[:category] = PirateBay::Categories::CATEGORIES_BY_NUMBER[url_to_category.split("/").last]
			url_to_torrent = html.search("td/a[@title='Download this torrent']").first.attributes["href"]
			torrent_hash[:link] = url_to_torrent
			url_to_details = html.search("td/a[@class='detLink']").first
			torrent_hash[:details_link] = url_to_details.attributes["href"]
			torrent_hash[:display_name] = url_to_details.inner_text

			last_three = html.search("td[@align='right']")
			torrent_hash[:size] = last_three.first.inner_html.sub(" "," ")
			#down size everything to megabytes. eazy to keep track of.
			# TODO add support for B and KiB?
			if torrent_hash[:size].downcase.include?("gib")
				torrent_hash[:size] = "#{torrent_hash[:size].to_f * 1000} MiB"
				#not sure if they have TIB
			elsif torrent_hash[:size].downcase.include?("tib")
				torrent_hash[:size] = "#{torrent_hash[:size].to_f * 1000000} MiB"
			end
			torrent_hash[:seeders] = last_three[1].inner_html.to_i
			torrent_hash[:leechers] = last_three[2].inner_html.to_i
			date = html.search("td/a[@title='Download this torrent']").first.parent.previous_sibling.inner_text
			# we get 0?mins?ago bs not sure why
			torrent_hash[:uploaded_time] = date_string_to_date_object(date.gsub("?"," "))
			#TODO file_name comments
			Torrent.new(torrent_hash)
		end

		def self.process_from_details_page(html)
		end

		attr_accessor :link, :uploaded_time, :category, :size, :seeders, :leechers
		attr_accessor :comments, :file_name, :display_name, :details_link

		def initialize(torrent_hash)
			@link = torrent_hash[:link]
			@uploaded_time = torrent_hash[:uploaded_time]
			@category = torrent_hash[:category]
			@size = torrent_hash[:size]
			@seeders = torrent_hash[:seeders]
			@leechers = torrent_hash[:leechers]
			@comments = torrent_hash[:comments]
			@file_name = torrent_hash[:file_name]
			@display_name = torrent_hash[:display_name]
			@details_link = torrent_hash[:details_link]
		end

		#sorts by MiB
		def <=>(other)
			self.size.to_f <=> other.size.to_f
		end

		private

		def self.date_string_to_date_object(string)
			offset = Time.new.gmt_offset/3600
			case string
			when /^(\d{1,2}) mins ago/i
				mins  = $1.to_i
				#grrr this is fed.TBD
				Time.at(Time.now.to_i + offset*60*60 - mins*60)
			# today at time gmt
			when /^today (\d\d):(\d\d)/i
				hour =  sprintf("%02d",($1.to_i + offset))
				min = $2
				Time.parse("#{hour}:#{min}")
			when /^y-day (\d\d):(\d\d)/i
				hour =  sprintf("%02d",($1+offset))
				min = $2
				today = Time.parse("#{hour}:#{min}")
				Time.at(today.to_i-(24*60*60))
			when /^(\d\d)-(\d\d) (\d\d):(\d\d)/
				months = Time::MonthValue.invert
				Time.gm(Time.now.year,$1,months[$2],$3,$4,0,0)
			when /^(\d\d)-(\d\d) (\d\d\d\d)/
				months = Time::MonthValue.invert
				Time.gm($3,$1,months[$2],0,0,0,0)
			end
		end
	end

	class Categories
		file = File.join(File.dirname(__FILE__),"config","pirate_categories.yml")
		if File.exists?(file)

			file_socket = File.open(file)
			categories = YAML.load(file_socket).delete_if{|k,v| k =~ /^_/}

			TOP_LEVEL_CATEGORIES = categories.keys

			#I know I can do this in one line
			all_categories = Hash.new
			categories.values.each{|hash| all_categories.merge!(hash)}

			ALL_CATEGORIES = all_categories
			CATEGORIES_BY_NUMBER = ALL_CATEGORIES.invert
		end
	end

end

if $0==__FILE__
	pp PirateBay::get_browse_categories
	pp PirateBay::Categories::ALL_CATEGORIES
	pp PirateBay::Categories::TOP_LEVEL_CATEGORIES
	pp PirateBay::Categories::CATEGORIES_BY_NUMBER
	recents = PirateBay::recent
	pp recents
	pp recents.sort
	pp recents.sort_by{|x| x.seeders/x.leechers rescue 0}
end