X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fyouku.py;h=37be440faab9b68a5c656883a212f2a0d996de6e;hb=59ed87cbd9ea08c889514a05b646141004f432a1;hp=78caeb8b36e0be8cf4e97365d9e28251723059b7;hpb=eacd875f3be175a4ddd73763c73017ed04cce4ce;p=youtube-dl diff --git a/youtube_dl/extractor/youku.py b/youtube_dl/extractor/youku.py index 78caeb8b3..37be440fa 100644 --- a/youtube_dl/extractor/youku.py +++ b/youtube_dl/extractor/youku.py @@ -1,15 +1,16 @@ # coding: utf-8 from __future__ import unicode_literals -import base64 +import itertools +import random +import re +import string +import time from .common import InfoExtractor -from ..utils import ExtractorError - -from ..compat import ( - compat_urllib_parse, - compat_ord, - compat_urllib_request, +from ..utils import ( + ExtractorError, + get_element_by_attribute, ) @@ -24,12 +25,12 @@ class YoukuIE(InfoExtractor): ''' _TESTS = [{ + # MD5 is unstable 'url': 'http://v.youku.com/v_show/id_XMTc1ODE5Njcy.html', - 'md5': '5f3af4192eabacc4501508d54a8cabd7', 'info_dict': { - 'id': 'XMTc1ODE5Njcy_part1', + 'id': 'XMTc1ODE5Njcy', 'title': '★Smile﹗♡ Git Fresh -Booty Music舞蹈.', - 'ext': 'flv' + 'ext': 'mp4', } }, { 'url': 'http://player.youku.com/player.php/sid/XNDgyMDQ2NTQw/v.swf', @@ -38,199 +39,170 @@ class YoukuIE(InfoExtractor): 'url': 'http://v.youku.com/v_show/id_XODgxNjg1Mzk2_ev_1.html', 'info_dict': { 'id': 'XODgxNjg1Mzk2', + 'ext': 'mp4', 'title': '武媚娘传奇 85', }, - 'playlist_count': 11, }, { 'url': 'http://v.youku.com/v_show/id_XMTI1OTczNDM5Mg==.html', 'info_dict': { 'id': 'XMTI1OTczNDM5Mg', + 'ext': 'mp4', 'title': '花千骨 04', }, - 'playlist_count': 13, - 'skip': 'Available in China only', + }, { + 'url': 'http://v.youku.com/v_show/id_XNjA1NzA2Njgw.html', + 'note': 'Video protected with password', + 'info_dict': { + 'id': 'XNjA1NzA2Njgw', + 'ext': 'mp4', + 'title': '邢義田复旦讲座之想象中的胡人—从“左衽孔子”说起', + }, + 'params': { + 'videopassword': '100600', + }, + }, { + # /play/get.json contains streams with "channel_type":"tail" + 'url': 'http://v.youku.com/v_show/id_XOTUxMzg4NDMy.html', + 'info_dict': { + 'id': 'XOTUxMzg4NDMy', + 'ext': 'mp4', + 'title': '我的世界☆明月庄主☆车震猎杀☆杀人艺术Minecraft', + }, }] - def construct_video_urls(self, data1, data2): - # get sid, token - def yk_t(s1, s2): - ls = list(range(256)) - t = 0 - for i in range(256): - t = (t + ls[i] + compat_ord(s1[i % len(s1)])) % 256 - ls[i], ls[t] = ls[t], ls[i] - s = bytearray() - x, y = 0, 0 - for i in range(len(s2)): - y = (y + 1) % 256 - x = (x + ls[y]) % 256 - ls[x], ls[y] = ls[y], ls[x] - s.append(compat_ord(s2[i]) ^ ls[(ls[x] + ls[y]) % 256]) - return bytes(s) - - sid, token = yk_t( - b'becaf9be', base64.b64decode(data2['ep'].encode('ascii')) - ).decode('ascii').split('_') - - # get oip - oip = data2['ip'] - - # get fileid - string_ls = list( - 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ/\:._-1234567890') - shuffled_string_ls = [] - seed = data1['seed'] - N = len(string_ls) - for ii in range(N): - seed = (seed * 0xd3 + 0x754f) % 0x10000 - idx = seed * len(string_ls) // 0x10000 - shuffled_string_ls.append(string_ls[idx]) - del string_ls[idx] - - fileid_dict = {} - for format in data1['streamtypes']: - streamfileid = [ - int(i) for i in data1['streamfileids'][format].strip('*').split('*')] - fileid = ''.join( - [shuffled_string_ls[i] for i in streamfileid]) - fileid_dict[format] = fileid[:8] + '%s' + fileid[10:] - - def get_fileid(format, n): - fileid = fileid_dict[format] % hex(int(n))[2:].upper().zfill(2) - return fileid - - # get ep - def generate_ep(format, n): - fileid = get_fileid(format, n) - ep_t = yk_t( - b'bf7e5f01', - ('%s_%s_%s' % (sid, fileid, token)).encode('ascii') - ) - ep = base64.b64encode(ep_t).decode('ascii') - return ep - - # generate video_urls - video_urls_dict = {} - for format in data1['streamtypes']: - video_urls = [] - for dt in data1['segs'][format]: - n = str(int(dt['no'])) - param = { - 'K': dt['k'], - 'hd': self.get_hd(format), - 'myp': 0, - 'ts': dt['seconds'], - 'ypp': 0, - 'ctype': 12, - 'ev': 1, - 'token': token, - 'oip': oip, - 'ep': generate_ep(format, n) - } - video_url = \ - 'http://k.youku.com/player/getFlvPath/' + \ - 'sid/' + sid + \ - '_' + str(int(n) + 1).zfill(2) + \ - '/st/' + self.parse_ext_l(format) + \ - '/fileid/' + get_fileid(format, n) + '?' + \ - compat_urllib_parse.urlencode(param) - video_urls.append(video_url) - video_urls_dict[format] = video_urls - - return video_urls_dict - - def get_hd(self, fm): - hd_id_dict = { - 'flv': '0', - 'mp4': '1', - 'hd2': '2', - 'hd3': '3', - '3gp': '0', - '3gphd': '1' - } - return hd_id_dict[fm] - - def parse_ext_l(self, fm): - ext_dict = { - 'flv': 'flv', - 'mp4': 'mp4', - 'hd2': 'flv', - 'hd3': 'flv', - '3gp': 'flv', - '3gphd': 'mp4' - } - return ext_dict[fm] + @staticmethod + def get_ysuid(): + return '%d%s' % (int(time.time()), ''.join([ + random.choice(string.ascii_letters) for i in range(3)])) def get_format_name(self, fm): _dict = { '3gp': 'h6', '3gphd': 'h5', 'flv': 'h4', + 'flvhd': 'h4', 'mp4': 'h3', + 'mp4hd': 'h3', + 'mp4hd2': 'h4', + 'mp4hd3': 'h4', 'hd2': 'h2', - 'hd3': 'h1' + 'hd3': 'h1', } - return _dict[fm] + return _dict.get(fm) def _real_extract(self, url): video_id = self._match_id(url) - def retrieve_data(req_url, note): - req = compat_urllib_request.Request(req_url) - - cn_verification_proxy = self._downloader.params.get('cn_verification_proxy') - if cn_verification_proxy: - req.add_header('Ytdl-request-proxy', cn_verification_proxy) + self._set_cookie('youku.com', '__ysuid', self.get_ysuid()) + self._set_cookie('youku.com', 'xreferrer', 'http://www.youku.com') - raw_data = self._download_json(req, video_id, note=note) - return raw_data['data'][0] + _, urlh = self._download_webpage_handle( + 'https://log.mmstat.com/eg.js', video_id, 'Retrieving cna info') + # The etag header is '"foobar"'; let's remove the double quotes + cna = urlh.headers['etag'][1:-1] # request basic data - data1 = retrieve_data( - 'http://v.youku.com/player/getPlayList/VideoIDS/%s' % video_id, - 'Downloading JSON metadata 1') - data2 = retrieve_data( - 'http://v.youku.com/player/getPlayList/VideoIDS/%s/Pf/4/ctype/12/ev/1' % video_id, - 'Downloading JSON metadata 2') - - error_code = data1.get('error_code') - if error_code: - error = data1.get('error') - if error is not None and '因版权原因无法观看此视频' in error: + basic_data_params = { + 'vid': video_id, + 'ccode': '0401', + 'client_ip': '192.168.1.1', + 'utid': cna, + 'client_ts': time.time() / 1000, + } + + video_password = self._downloader.params.get('videopassword') + if video_password: + basic_data_params['password'] = video_password + + headers = { + 'Referer': url, + } + headers.update(self.geo_verification_headers()) + data = self._download_json( + 'https://ups.youku.com/ups/get.json', video_id, + 'Downloading JSON metadata', + query=basic_data_params, headers=headers)['data'] + + error = data.get('error') + if error: + error_note = error.get('note') + if error_note is not None and '因版权原因无法观看此视频' in error_note: raise ExtractorError( 'Youku said: Sorry, this video is available in China only', expected=True) + elif error_note and '该视频被设为私密' in error_note: + raise ExtractorError( + 'Youku said: Sorry, this video is private', expected=True) else: - msg = 'Youku server reported error %i' % error_code - if error is not None: - msg += ': ' + error + msg = 'Youku server reported error %i' % error.get('code') + if error_note is not None: + msg += ': ' + error_note raise ExtractorError(msg) - title = data1['title'] + # get video title + title = data['video']['title'] - # generate video_urls_dict - video_urls_dict = self.construct_video_urls(data1, data2) - - # construct info - entries = [{ - 'id': '%s_part%d' % (video_id, i + 1), - 'title': title, - 'formats': [], - # some formats are not available for all parts, we have to detect - # which one has all - } for i in range(max(len(v) for v in data1['segs'].values()))] - for fm in data1['streamtypes']: - video_urls = video_urls_dict[fm] - for video_url, seg, entry in zip(video_urls, data1['segs'][fm], entries): - entry['formats'].append({ - 'url': video_url, - 'format_id': self.get_format_name(fm), - 'ext': self.parse_ext_l(fm), - 'filesize': int(seg['size']), - }) + formats = [{ + 'url': stream['m3u8_url'], + 'format_id': self.get_format_name(stream.get('stream_type')), + 'ext': 'mp4', + 'protocol': 'm3u8_native', + 'filesize': int(stream.get('size')), + 'width': stream.get('width'), + 'height': stream.get('height'), + } for stream in data['stream'] if stream.get('channel_type') != 'tail'] + self._sort_formats(formats) return { - '_type': 'multi_video', 'id': video_id, 'title': title, - 'entries': entries, + 'formats': formats, } + + +class YoukuShowIE(InfoExtractor): + _VALID_URL = r'https?://(?:www\.)?youku\.com/show_page/id_(?P[0-9a-z]+)\.html' + IE_NAME = 'youku:show' + + _TEST = { + 'url': 'http://www.youku.com/show_page/id_zc7c670be07ff11e48b3f.html', + 'info_dict': { + 'id': 'zc7c670be07ff11e48b3f', + 'title': '花千骨 未删减版', + 'description': 'md5:578d4f2145ae3f9128d9d4d863312910', + }, + 'playlist_count': 50, + } + + _PAGE_SIZE = 40 + + def _find_videos_in_page(self, webpage): + videos = re.findall( + r'
  • ]+href="(?Phttps?://v\.youku\.com/[^"]+)"[^>]+title="(?P[^"]+)"', webpage) + return [ + self.url_result(video_url, YoukuIE.ie_key(), title) + for video_url, title in videos] + + def _real_extract(self, url): + show_id = self._match_id(url) + webpage = self._download_webpage(url, show_id) + + entries = self._find_videos_in_page(webpage) + + playlist_title = self._html_search_regex( + r'<span[^>]+class="name">([^<]+)</span>', webpage, 'playlist title', fatal=False) + detail_div = get_element_by_attribute('class', 'detail', webpage) or '' + playlist_description = self._html_search_regex( + r'<span[^>]+style="display:none"[^>]*>([^<]+)</span>', + detail_div, 'playlist description', fatal=False) + + for idx in itertools.count(1): + episodes_page = self._download_webpage( + 'http://www.youku.com/show_episode/id_%s.html' % show_id, + show_id, query={'divid': 'reload_%d' % (idx * self._PAGE_SIZE + 1)}, + note='Downloading episodes page %d' % idx) + new_entries = self._find_videos_in_page(episodes_page) + entries.extend(new_entries) + if len(new_entries) < self._PAGE_SIZE: + break + + return self.playlist_result(entries, show_id, playlist_title, playlist_description)