#!/usr/bin/env python
# -*- coding: utf-8 -*-

import re
import os
import pycurl
import lxml.html
from lxml import etree
import StringIO
import html5lib
from html5lib import treebuilders


class vk_dwl():

    def __init__(self, email, password):
        self.email = email
        self.password = password
        self.cookie = ''
        self.crl = pycurl.Curl()
        self.errlog = open('err.log', 'a')
        #self.filt = re.compile('[\x00\x03\x10\x13\x20-\xD7FF\xE000-\xFFFD\x10000]')
        self.filt = re.compile('(?u)[\x00\x03\x10\x13]')
        self.fn1 = re.compile('(?u)  *')
        self.fn2 = re.compile('(?u)^ *|  *$')
        self.fn3 = re.compile('(?u) -$')
        self.fn4 = re.compile('(?u)/')

    def __del__(self):
        self.errlog.close()

    def get_url(self, url):
        if url:
            t = StringIO.StringIO()
            self.crl.setopt(pycurl.URL, url)
            self.crl.setopt(pycurl.HTTPHEADER, ["Accept:"])
            self.crl.setopt(pycurl.WRITEFUNCTION, t.write)
            self.crl.setopt(pycurl.FOLLOWLOCATION, 1)
            self.crl.setopt(pycurl.MAXREDIRS, 5)
            self.crl.setopt(pycurl.COOKIEFILE, self.cookie)
            self.crl.perform()
            t = self.filt.sub('', t.getvalue())
            return html5lib.parse(t, treebuilder = "lxml", namespaceHTMLElements = False), t
        else:
            return None

    def get_dump(self, url, fname):
        if url and fname:
            f = open(fname,'w')
            self.crl.setopt(pycurl.URL, url)
            self.crl.setopt(pycurl.HTTPHEADER, ["Accept:"])
            self.crl.setopt(pycurl.WRITEFUNCTION, f.write)
            self.crl.setopt(pycurl.FOLLOWLOCATION, 1)
            self.crl.setopt(pycurl.MAXREDIRS, 5)
            self.crl.setopt(pycurl.COOKIEFILE, self.cookie)
            self.crl.perform()
            f.close()

    def get_file(self, fname):
        f = open(fname, 'r')
        res = ''
        for ln in f.readlines():
            res += ln + '\n'
        f.close()
        return html5lib.parse(res, treebuilder = "lxml", namespaceHTMLElements = False), res

    def login(self):
        tree, src = self.get_url('http://vk.com/login.php?email='+self.email+'&pass='+self.password)
        # Тут нужно попарсить скрипт на предмет наличия ошибок при авторизации и выдавать код ошибки в резалте

    def get_audio(self, vkid = None):
        # Получение списка аудио
        # Надо записывать список url
        # При старте смотреть, есть ли список, если список есть, то подгружать его в память и перед загрузкой проверять наличие урла

        if vkid:
            upref = 'http://vk.com/audio?friend=' + str(vkid)
        else:
            upref = 'http://vk.com/audio'

        def get_a(offset = 0):
            if offset:
                if upref == 'http://vk.com/audio':
                    tree = self.get_url(upref + '?offset=' + str(offset))[0]
                else:
                    tree = self.get_url(upref + '&offset=' + str(offset))[0]
            else:
                tree = self.get_url(upref)[0]
            links = tree.findall('//td[@class="play_btn"]')
            names = tree.findall('//td[@class="info"]')
            return tree, links, names

        def fn_corr(fname):
            if fname:
                res = self.fn1.sub(' ', fname)
                res = self.fn2.sub('', res)
                res = self.fn3.sub('', res)
                res = self.fn4.sub('-', res)
                return res
            else:
                return None

        # tr - tree, ln - links, nm - names
        tr, ln, nm = get_a()

        summary = int(re.search('(?u)[0-9]+', tr.xpath('//div[@class="summary"]/b[@id="audio_summary"]/text()')[0]).group(0))
        cln = len(ln)
        cl = re.compile('(?u),[0-9]+$')

        lst = open('url.lst', 'w')
        #err = open('err.log', 'w')

        for c in range(summary/cln+1):
            if c:
                tr, ln, nm = get_a(c*cln)
            print '=-' * 15 + ' ' + str(c*50) + ' ' + '-=' * 15
            for i in range(len(ln)):
                url = cl.sub('', ln[i][1].values()[2])
                artist = nm[i].xpath('./div[@class="title_wrap"]/b/a/text()')[0]
                title = nm[i].xpath('.//span[@class="title"]/a/text()')
                if title:
                    title = title[0]
                else:
                    title = nm[i].xpath('.//span[@class="title"]/text()')[0]
                if title:
                    fn = fn_corr(artist + ' - ' + title)
                else:
                    fn = fn_corr(artist)
                if fn:
                    fn += '.mp3'
                    if not os.path.exists(fn):
                        print 'Download: ' + url + ' -> ' + fn
                        try:
                            self.get_dump(url, fn)
                            lst.write(url)
                        except:
                            print 'ERROR: Can\'t download file ' + fn
                            try:
                                self.errlog.write('ERROR: Can\'t download ' + url + ' to file ' + fn)
                            except:
                                print 'Can\'t write filename ' + fn + ' to logfile'
                    else:
                        print 'PASS: ' + fn

        lst.close()

tmp = vk_dwl('xxx@mail.ru', 'pass')
tmp.login()
tmp.get_audio('id')