动漫花园资源获取.coffee

fs = require 'fs'
jsdom = require 'jsdom'
flowless = require 'flowless'
mysql = require 'mysql'
jquery = fs.readFileSync('../jquery.js').toString()
_ = require 'underscore'
strip_html = require 'strip_html'
serialize = require 'serialize'
cur_page = 1

start = (callback)->
  connection = mysql.createConnection
  	host:'127.0.0.1'
		user:'root'
		password:''
		database:'moefou'
	flowless.runSeq [
		#stert fetch
		(cb)->
			jsdom.env
				html:"http://share.dmhy.org/topics/list/page/#{cur_page}"
				src:[jquery]
				done:cb
		,(window,cb)->
			$ = window.$
			rows = []
			$('#topic_list td.title').each (i,v)->
				v = $ v
				title = v.children('a').text().replace(/\r\n\t\t\t\t/,'')
				tags = _.compact get_tags title
				url = 'http://share.dmhy.org' + v.children('a').attr('href')
				hash = require('crypto').createHash('md5').update(url).digest("hex")
				rows.push {
					title
					tags
					url
					hash
				}
			connection.connect((err)->
				cb err,rows
				)
		,flowless.map (element,cb)->
			flowless.runSeq [
				#check if the resource is already exists in the db
				(cb)->
					query = "SELECT `rs_id` FROM `mp_resources` WHERE `rs_encode` = ?"
					connection.query query,[element.hash],cb
				,(results,q,cb)->
					unless results.length is 0  
						cb null,false
						return 0
					jsdom.env
						html:element.url
						src:[jquery]
						proxy:'http://192.168.11.2:8888'
						done:cb
					console.log 'Send.'
				,(window,cb)->
					unless window?
						cb(null,false)
						return 0
					$ = window.$
					magnet = $('#tabs-1').children('p').eq(1).children('a').attr('href')
					torrent = 'http://share.dmhy.org' + $('#tabs-1').children('p').eq(0).children('a').attr('href')
					rs_data = serialize {magnet,torrent}
					rs_about = strip_html($('.topic-nfo').children('strong').remove().end().html()).trim()
					rs_size = get_size $('.resource-info').children().children().eq(4).children('span').text()
					res = {
						rs_data
						rs_about
						rs_size
						rs_encode:element.hash
						rs_site:'dmhy'
						rs_type:'bt'
						rs_title:element.title
						rs_source:element.url
						rs_date:Math.round new Date().getTime()/1000
						rs_img:''
					}
					console.log(JSON.stringify(res));
					cb()
			],cb
	],(err)->
		console.log err
exports.start = start
start()



get_tags = (raw)->
	return '' if raw is ''
	return '' if !raw?
	raw = raw.replace /☪|★|☆/g,''
	pattern = /\[(.+?)\]|【(.+?)\】|\((.+?)\)|『(.+?)\』|～(.+?)\～|◆(.+?)◇/g
	tmp = raw.match(pattern)
	tmp = _.compact tmp
	if tmp? then tmp = for val in tmp
		val[1...-1]
	else tmp = []
	_raw = raw.replace(pattern,'').split(/\s+/g)
	if _raw? and _raw.length >0 then for val in _raw
		if val.replace /\s/g,'' is not val then tmp.push val
	tmp = for tag in tmp
		tag = tag.replace(/\s|&|&|＆|\/|／|_|\|/g,' ')
		tag = tag.split(/\s/g)
	tmp = _.flatten tmp
	patterns = [
		/(第(\d+)话)/
		/(第(\d+)話)/
		/(第(\d+)集)/
		/(CH(\d+))/i
		/(VOL\.?(\d+-?\d*))/i
	]
	tmp = for tag in tmp
		for pattern in patterns
			sp = tag.split pattern
			continue if sp.length is 1
			tag = [sp[0],sp[1],sp[3]]
			break
		tag
	_.flatten tmp

get_size = (str,target = 'KB')->
	str = str.toUpperCase()
	[str,size,format] = str.match(/([0-9]+\.?[0-9]*)(TB|GB|MB|KB|B)/)
	size = switch format
		when 'TB' then size*1024*1024*1024*1024
		when 'GB' then size*1024*1024*1024
		when 'MB' then size*1024*1024
		when 'KB' then size*1024
	ret = switch target
		when 'TB' then size/1024/1024/1024/1024
		when 'GB' then size/1024/1024/1024
		when 'MB' then size/1024/1024
		when 'KB' then size/1024
	Math.round ret