]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/sohu.py
30760ca06be4b3fc112f3fe0200c74b665d64855
[youtubedl] / youtube_dl / extractor / sohu.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5
6 from .common import InfoExtractor
7 from ..compat import (
8 compat_str,
9 compat_urllib_parse_urlencode,
10 )
11 from ..utils import ExtractorError
12
13
14 class SohuIE(InfoExtractor):
15 _VALID_URL = r'https?://(?P<mytv>my\.)?tv\.sohu\.com/.+?/(?(mytv)|n)(?P<id>\d+)\.shtml.*?'
16
17 # Sohu videos give different MD5 sums on Travis CI and my machine
18 _TESTS = [{
19 'note': 'This video is available only in Mainland China',
20 'url': 'http://tv.sohu.com/20130724/n382479172.shtml#super',
21 'info_dict': {
22 'id': '382479172',
23 'ext': 'mp4',
24 'title': 'MV:Far East Movement《The Illest》',
25 },
26 'skip': 'On available in China',
27 }, {
28 'url': 'http://tv.sohu.com/20150305/n409385080.shtml',
29 'info_dict': {
30 'id': '409385080',
31 'ext': 'mp4',
32 'title': '《2015湖南卫视羊年元宵晚会》唐嫣《花好月圆》',
33 }
34 }, {
35 'url': 'http://my.tv.sohu.com/us/232799889/78693464.shtml',
36 'info_dict': {
37 'id': '78693464',
38 'ext': 'mp4',
39 'title': '【爱范品】第31期:MWC见不到的奇葩手机',
40 }
41 }, {
42 'note': 'Multipart video',
43 'url': 'http://my.tv.sohu.com/pl/8384802/78910339.shtml',
44 'info_dict': {
45 'id': '78910339',
46 'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
47 },
48 'playlist': [{
49 'info_dict': {
50 'id': '78910339_part1',
51 'ext': 'mp4',
52 'duration': 294,
53 'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
54 }
55 }, {
56 'info_dict': {
57 'id': '78910339_part2',
58 'ext': 'mp4',
59 'duration': 300,
60 'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
61 }
62 }, {
63 'info_dict': {
64 'id': '78910339_part3',
65 'ext': 'mp4',
66 'duration': 150,
67 'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
68 }
69 }]
70 }, {
71 'note': 'Video with title containing dash',
72 'url': 'http://my.tv.sohu.com/us/249884221/78932792.shtml',
73 'info_dict': {
74 'id': '78932792',
75 'ext': 'mp4',
76 'title': 'youtube-dl testing video',
77 },
78 'params': {
79 'skip_download': True
80 }
81 }]
82
83 def _real_extract(self, url):
84
85 def _fetch_data(vid_id, mytv=False):
86 if mytv:
87 base_data_url = 'http://my.tv.sohu.com/play/videonew.do?vid='
88 else:
89 base_data_url = 'http://hot.vrs.sohu.com/vrs_flash.action?vid='
90
91 return self._download_json(
92 base_data_url + vid_id, video_id,
93 'Downloading JSON data for %s' % vid_id,
94 headers=self.geo_verification_headers())
95
96 mobj = re.match(self._VALID_URL, url)
97 video_id = mobj.group('id')
98 mytv = mobj.group('mytv') is not None
99
100 webpage = self._download_webpage(url, video_id)
101
102 title = re.sub(r' - 搜狐视频$', '', self._og_search_title(webpage))
103
104 vid = self._html_search_regex(
105 r'var vid ?= ?["\'](\d+)["\']',
106 webpage, 'video path')
107 vid_data = _fetch_data(vid, mytv)
108 if vid_data['play'] != 1:
109 if vid_data.get('status') == 12:
110 raise ExtractorError(
111 'Sohu said: There\'s something wrong in the video.',
112 expected=True)
113 else:
114 raise ExtractorError(
115 'Sohu said: The video is only licensed to users in Mainland China.',
116 expected=True)
117
118 formats_json = {}
119 for format_id in ('nor', 'high', 'super', 'ori', 'h2644k', 'h2654k'):
120 vid_id = vid_data['data'].get('%sVid' % format_id)
121 if not vid_id:
122 continue
123 vid_id = compat_str(vid_id)
124 formats_json[format_id] = vid_data if vid == vid_id else _fetch_data(vid_id, mytv)
125
126 part_count = vid_data['data']['totalBlocks']
127
128 playlist = []
129 for i in range(part_count):
130 formats = []
131 for format_id, format_data in formats_json.items():
132 allot = format_data['allot']
133
134 data = format_data['data']
135 clips_url = data['clipsURL']
136 su = data['su']
137
138 video_url = 'newflv.sohu.ccgslb.net'
139 cdnId = None
140 retries = 0
141
142 while 'newflv.sohu.ccgslb.net' in video_url:
143 params = {
144 'prot': 9,
145 'file': clips_url[i],
146 'new': su[i],
147 'prod': 'flash',
148 'rb': 1,
149 }
150
151 if cdnId is not None:
152 params['idc'] = cdnId
153
154 download_note = 'Downloading %s video URL part %d of %d' % (
155 format_id, i + 1, part_count)
156
157 if retries > 0:
158 download_note += ' (retry #%d)' % retries
159 part_info = self._parse_json(self._download_webpage(
160 'http://%s/?%s' % (allot, compat_urllib_parse_urlencode(params)),
161 video_id, download_note), video_id)
162
163 video_url = part_info['url']
164 cdnId = part_info.get('nid')
165
166 retries += 1
167 if retries > 5:
168 raise ExtractorError('Failed to get video URL')
169
170 formats.append({
171 'url': video_url,
172 'format_id': format_id,
173 'filesize': data['clipsBytes'][i],
174 'width': data['width'],
175 'height': data['height'],
176 'fps': data['fps'],
177 })
178 self._sort_formats(formats)
179
180 playlist.append({
181 'id': '%s_part%d' % (video_id, i + 1),
182 'title': title,
183 'duration': vid_data['data']['clipsDuration'][i],
184 'formats': formats,
185 })
186
187 if len(playlist) == 1:
188 info = playlist[0]
189 info['id'] = video_id
190 else:
191 info = {
192 '_type': 'multi_video',
193 'entries': playlist,
194 'id': video_id,
195 'title': title,
196 }
197
198 return info