X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/9c8b63077a48f758bf0c9a7351669557071bbd74..0ae75ad42cc6f25af12b54201dcbadfc116f923f:/youtube_dl/extractor/youku.py?ds=inline diff --git a/youtube_dl/extractor/youku.py b/youtube_dl/extractor/youku.py index 9d88c17..e37f237 100644 --- a/youtube_dl/extractor/youku.py +++ b/youtube_dl/extractor/youku.py @@ -1,123 +1,334 @@ # coding: utf-8 +from __future__ import unicode_literals -import json -import math +import base64 +import itertools import random import re +import string import time from .common import InfoExtractor +from ..compat import ( + compat_urllib_parse_urlencode, + compat_ord, +) from ..utils import ( ExtractorError, + get_element_by_attribute, ) class YoukuIE(InfoExtractor): - _VALID_URL = r'(?:(?:http://)?(?:v|player)\.youku\.com/(?:v_show/id_|player\.php/sid/)|youku:)(?P[A-Za-z0-9]+)(?:\.html|/v\.swf|)' - _TEST = { - u"url": u"http://v.youku.com/v_show/id_XNDgyMDQ2NTQw.html", - u"file": u"XNDgyMDQ2NTQw_part00.flv", - u"md5": u"ffe3f2e435663dc2d1eea34faeff5b5b", - u"params": { u"test": False }, - u"info_dict": { - u"title": u"youtube-dl test video \"'/\\ä↭𝕐" + IE_NAME = 'youku' + IE_DESC = '优酷' + _VALID_URL = r'''(?x) + (?: + http://(?:v|player)\.youku\.com/(?:v_show/id_|player\.php/sid/)| + youku:) + (?P[A-Za-z0-9]+)(?:\.html|/v\.swf|) + ''' + + _TESTS = [{ + # MD5 is unstable + 'url': 'http://v.youku.com/v_show/id_XMTc1ODE5Njcy.html', + 'info_dict': { + 'id': 'XMTc1ODE5Njcy_part1', + 'title': '★Smile﹗♡ Git Fresh -Booty Music舞蹈.', + 'ext': 'flv' } - } + }, { + 'url': 'http://player.youku.com/player.php/sid/XNDgyMDQ2NTQw/v.swf', + 'only_matching': True, + }, { + 'url': 'http://v.youku.com/v_show/id_XODgxNjg1Mzk2_ev_1.html', + 'info_dict': { + 'id': 'XODgxNjg1Mzk2', + 'title': '武媚娘传奇 85', + }, + 'playlist_count': 11, + 'skip': 'Available in China only', + }, { + 'url': 'http://v.youku.com/v_show/id_XMTI1OTczNDM5Mg==.html', + 'info_dict': { + 'id': 'XMTI1OTczNDM5Mg', + 'title': '花千骨 04', + }, + 'playlist_count': 13, + }, { + 'url': 'http://v.youku.com/v_show/id_XNjA1NzA2Njgw.html', + 'note': 'Video protected with password', + 'info_dict': { + 'id': 'XNjA1NzA2Njgw', + 'title': '邢義田复旦讲座之想象中的胡人—从“左衽孔子”说起', + }, + 'playlist_count': 19, + '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', + 'title': '我的世界☆明月庄主☆车震猎杀☆杀人艺术Minecraft', + }, + 'playlist_count': 6, + }] + + def construct_video_urls(self, data): + # 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(data['security']['encrypt_string'].encode('ascii')) + ).decode('ascii').split('_') + + # get oip + oip = data['security']['ip'] + fileid_dict = {} + for stream in data['stream']: + if stream.get('channel_type') == 'tail': + continue + format = stream.get('stream_type') + fileid = stream['stream_fileid'] + fileid_dict[format] = fileid - def _gen_sid(self): - nowTime = int(time.time() * 1000) - random1 = random.randint(1000,1998) - random2 = random.randint(1000,9999) - - return "%d%d%d" %(nowTime,random1,random2) - - def _get_file_ID_mix_string(self, seed): - mixed = [] - source = list("abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ/\:._-1234567890") - seed = float(seed) - for i in range(len(source)): - seed = (seed * 211 + 30031 ) % 65536 - index = math.floor(seed / 65536 * len(source) ) - mixed.append(source[int(index)]) - source.remove(source[int(index)]) - #return ''.join(mixed) - return mixed - - def _get_file_id(self, fileId, seed): - mixed = self._get_file_ID_mix_string(seed) - ids = fileId.split('*') - realId = [] - for ch in ids: - if ch: - realId.append(mixed[int(ch)]) - return ''.join(realId) + def get_fileid(format, n): + number = hex(int(str(n), 10))[2:].upper() + if len(number) == 1: + number = '0' + number + streamfileids = fileid_dict[format] + fileid = streamfileids[0:8] + number + streamfileids[10:] + 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 stream in data['stream']: + if stream.get('channel_type') == 'tail': + continue + format = stream.get('stream_type') + video_urls = [] + for dt in stream['segs']: + n = str(stream['segs'].index(dt)) + param = { + 'K': dt['key'], + 'hd': self.get_hd(format), + 'myp': 0, + '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 + \ + '_00' + \ + '/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 + + @staticmethod + def get_ysuid(): + return '%d%s' % (int(time.time()), ''.join([ + random.choice(string.ascii_letters) for i in range(3)])) + + def get_hd(self, fm): + hd_id_dict = { + '3gp': '0', + '3gphd': '1', + 'flv': '0', + 'flvhd': '0', + 'mp4': '1', + 'mp4hd': '1', + 'mp4hd2': '1', + 'mp4hd3': '1', + 'hd2': '2', + 'hd3': '3', + } + return hd_id_dict[fm] + + def parse_ext_l(self, fm): + ext_dict = { + '3gp': 'flv', + '3gphd': 'mp4', + 'flv': 'flv', + 'flvhd': 'flv', + 'mp4': 'mp4', + 'mp4hd': 'mp4', + 'mp4hd2': 'flv', + 'mp4hd3': 'flv', + 'hd2': 'flv', + 'hd3': 'flv', + } + return ext_dict[fm] + + 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', + } + return _dict[fm] def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - if mobj is None: - raise ExtractorError(u'Invalid URL: %s' % url) - video_id = mobj.group('ID') - - info_url = 'http://v.youku.com/player/getPlayList/VideoIDS/' + video_id - - jsondata = self._download_webpage(info_url, video_id) - - self.report_extraction(video_id) - try: - config = json.loads(jsondata) - error_code = config['data'][0].get('error_code') - if error_code: - # -8 means blocked outside China. - error = config['data'][0].get('error') # Chinese and English, separated by newline. - raise ExtractorError(error or u'Server reported error %i' % error_code, - expected=True) - - video_title = config['data'][0]['title'] - seed = config['data'][0]['seed'] - - format = self._downloader.params.get('format', None) - supported_format = list(config['data'][0]['streamfileids'].keys()) - - if format is None or format == 'best': - if 'hd2' in supported_format: - format = 'hd2' - else: - format = 'flv' - ext = u'flv' - elif format == 'worst': - format = 'mp4' - ext = u'mp4' - else: - format = 'flv' - ext = u'flv' - - - fileid = config['data'][0]['streamfileids'][format] - keys = [s['k'] for s in config['data'][0]['segs'][format]] - # segs is usually a dictionary, but an empty *list* if an error occured. - except (UnicodeDecodeError, ValueError, KeyError): - raise ExtractorError(u'Unable to extract info section') - - files_info=[] - sid = self._gen_sid() - fileid = self._get_file_id(fileid, seed) - - #column 8,9 of fileid represent the segment number - #fileid[7:9] should be changed - for index, key in enumerate(keys): - - temp_fileid = '%s%02X%s' % (fileid[0:8], index, fileid[10:]) - download_url = 'http://f.youku.com/player/getFlvPath/sid/%s_%02X/st/flv/fileid/%s?k=%s' % (sid, index, temp_fileid, key) - - info = { - 'id': '%s_part%02d' % (video_id, index), - 'url': download_url, - 'uploader': None, - 'upload_date': None, - 'title': video_title, - 'ext': ext, + video_id = self._match_id(url) + + self._set_cookie('youku.com', '__ysuid', self.get_ysuid()) + + def retrieve_data(req_url, note): + headers = { + 'Referer': req_url, } - files_info.append(info) + headers.update(self.geo_verification_headers()) + self._set_cookie('youku.com', 'xreferrer', 'http://www.youku.com') + + raw_data = self._download_json(req_url, video_id, note=note, headers=headers) + + return raw_data['data'] + + video_password = self._downloader.params.get('videopassword') + + # request basic data + basic_data_url = 'http://play.youku.com/play/get.json?vid=%s&ct=12' % video_id + if video_password: + basic_data_url += '&pwd=%s' % video_password + + data = retrieve_data(basic_data_url, 'Downloading JSON metadata') + + 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.get('code') + if error_note is not None: + msg += ': ' + error_note + raise ExtractorError(msg) + + # get video title + title = data['video']['title'] + + # generate video_urls_dict + video_urls_dict = self.construct_video_urls(data) + + # 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.get('segs')) for v in data['stream']))] + for stream in data['stream']: + if stream.get('channel_type') == 'tail': + continue + fm = stream.get('stream_type') + video_urls = video_urls_dict[fm] + for video_url, seg, entry in zip(video_urls, stream['segs'], entries): + entry['formats'].append({ + 'url': video_url, + 'format_id': self.get_format_name(fm), + 'ext': self.parse_ext_l(fm), + 'filesize': int(seg['size']), + 'width': stream.get('width'), + 'height': stream.get('height'), + }) + + return { + '_type': 'multi_video', + 'id': video_id, + 'title': title, + 'entries': entries, + } + + +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 files_info + return self.playlist_result(entries, show_id, playlist_title, playlist_description)