]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/vlive.py
Fix extraction from youtube.
[youtubedl] / youtube_dl / extractor / vlive.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5 import time
6 import itertools
7
8 from .common import InfoExtractor
9 from ..compat import (
10 compat_urllib_parse_urlencode,
11 compat_str,
12 )
13 from ..utils import (
14 dict_get,
15 ExtractorError,
16 float_or_none,
17 int_or_none,
18 remove_start,
19 try_get,
20 urlencode_postdata,
21 )
22
23
24 class VLiveIE(InfoExtractor):
25 IE_NAME = 'vlive'
26 _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/video/(?P<id>[0-9]+)'
27 _TESTS = [{
28 'url': 'http://www.vlive.tv/video/1326',
29 'md5': 'cc7314812855ce56de70a06a27314983',
30 'info_dict': {
31 'id': '1326',
32 'ext': 'mp4',
33 'title': "[V LIVE] Girl's Day's Broadcast",
34 'creator': "Girl's Day",
35 'view_count': int,
36 },
37 }, {
38 'url': 'http://www.vlive.tv/video/16937',
39 'info_dict': {
40 'id': '16937',
41 'ext': 'mp4',
42 'title': '[V LIVE] 첸백시 걍방',
43 'creator': 'EXO',
44 'view_count': int,
45 'subtitles': 'mincount:12',
46 },
47 'params': {
48 'skip_download': True,
49 },
50 }]
51
52 @classmethod
53 def suitable(cls, url):
54 return False if VLivePlaylistIE.suitable(url) else super(VLiveIE, cls).suitable(url)
55
56 def _real_extract(self, url):
57 video_id = self._match_id(url)
58
59 webpage = self._download_webpage(
60 'https://www.vlive.tv/video/%s' % video_id, video_id)
61
62 VIDEO_PARAMS_RE = r'\bvlive\.video\.init\(([^)]+)'
63 VIDEO_PARAMS_FIELD = 'video params'
64
65 params = self._parse_json(self._search_regex(
66 VIDEO_PARAMS_RE, webpage, VIDEO_PARAMS_FIELD, default=''), video_id,
67 transform_source=lambda s: '[' + s + ']', fatal=False)
68
69 if not params or len(params) < 7:
70 params = self._search_regex(
71 VIDEO_PARAMS_RE, webpage, VIDEO_PARAMS_FIELD)
72 params = [p.strip(r'"') for p in re.split(r'\s*,\s*', params)]
73
74 status, long_video_id, key = params[2], params[5], params[6]
75 status = remove_start(status, 'PRODUCT_')
76
77 if status in ('LIVE_ON_AIR', 'BIG_EVENT_ON_AIR'):
78 return self._live(video_id, webpage)
79 elif status in ('VOD_ON_AIR', 'BIG_EVENT_INTRO'):
80 if long_video_id and key:
81 return self._replay(video_id, webpage, long_video_id, key)
82 else:
83 status = 'COMING_SOON'
84
85 if status == 'LIVE_END':
86 raise ExtractorError('Uploading for replay. Please wait...',
87 expected=True)
88 elif status == 'COMING_SOON':
89 raise ExtractorError('Coming soon!', expected=True)
90 elif status == 'CANCELED':
91 raise ExtractorError('We are sorry, '
92 'but the live broadcast has been canceled.',
93 expected=True)
94 else:
95 raise ExtractorError('Unknown status %s' % status)
96
97 def _get_common_fields(self, webpage):
98 title = self._og_search_title(webpage)
99 creator = self._html_search_regex(
100 r'<div[^>]+class="info_area"[^>]*>\s*<a\s+[^>]*>([^<]+)',
101 webpage, 'creator', fatal=False)
102 thumbnail = self._og_search_thumbnail(webpage)
103 return {
104 'title': title,
105 'creator': creator,
106 'thumbnail': thumbnail,
107 }
108
109 def _live(self, video_id, webpage):
110 init_page = self._download_webpage(
111 'https://www.vlive.tv/video/init/view',
112 video_id, note='Downloading live webpage',
113 data=urlencode_postdata({'videoSeq': video_id}),
114 headers={
115 'Referer': 'https://www.vlive.tv/video/%s' % video_id,
116 'Content-Type': 'application/x-www-form-urlencoded'
117 })
118
119 live_params = self._search_regex(
120 r'"liveStreamInfo"\s*:\s*(".*"),',
121 init_page, 'live stream info')
122 live_params = self._parse_json(live_params, video_id)
123 live_params = self._parse_json(live_params, video_id)
124
125 formats = []
126 for vid in live_params.get('resolutions', []):
127 formats.extend(self._extract_m3u8_formats(
128 vid['cdnUrl'], video_id, 'mp4',
129 m3u8_id=vid.get('name'),
130 fatal=False, live=True))
131 self._sort_formats(formats)
132
133 info = self._get_common_fields(webpage)
134 info.update({
135 'title': self._live_title(info['title']),
136 'id': video_id,
137 'formats': formats,
138 'is_live': True,
139 })
140 return info
141
142 def _replay(self, video_id, webpage, long_video_id, key):
143 playinfo = self._download_json(
144 'http://global.apis.naver.com/rmcnmv/rmcnmv/vod_play_videoInfo.json?%s'
145 % compat_urllib_parse_urlencode({
146 'videoId': long_video_id,
147 'key': key,
148 'ptc': 'http',
149 'doct': 'json', # document type (xml or json)
150 'cpt': 'vtt', # captions type (vtt or ttml)
151 }), video_id)
152
153 formats = [{
154 'url': vid['source'],
155 'format_id': vid.get('encodingOption', {}).get('name'),
156 'abr': float_or_none(vid.get('bitrate', {}).get('audio')),
157 'vbr': float_or_none(vid.get('bitrate', {}).get('video')),
158 'width': int_or_none(vid.get('encodingOption', {}).get('width')),
159 'height': int_or_none(vid.get('encodingOption', {}).get('height')),
160 'filesize': int_or_none(vid.get('size')),
161 } for vid in playinfo.get('videos', {}).get('list', []) if vid.get('source')]
162 self._sort_formats(formats)
163
164 view_count = int_or_none(playinfo.get('meta', {}).get('count'))
165
166 subtitles = {}
167 for caption in playinfo.get('captions', {}).get('list', []):
168 lang = dict_get(caption, ('locale', 'language', 'country', 'label'))
169 if lang and caption.get('source'):
170 subtitles[lang] = [{
171 'ext': 'vtt',
172 'url': caption['source']}]
173
174 info = self._get_common_fields(webpage)
175 info.update({
176 'id': video_id,
177 'formats': formats,
178 'view_count': view_count,
179 'subtitles': subtitles,
180 })
181 return info
182
183
184 class VLiveChannelIE(InfoExtractor):
185 IE_NAME = 'vlive:channel'
186 _VALID_URL = r'https?://channels\.vlive\.tv/(?P<id>[0-9A-Z]+)'
187 _TEST = {
188 'url': 'http://channels.vlive.tv/FCD4B',
189 'info_dict': {
190 'id': 'FCD4B',
191 'title': 'MAMAMOO',
192 },
193 'playlist_mincount': 110
194 }
195 _APP_ID = '8c6cc7b45d2568fb668be6e05b6e5a3b'
196
197 def _real_extract(self, url):
198 channel_code = self._match_id(url)
199
200 webpage = self._download_webpage(
201 'http://channels.vlive.tv/%s/video' % channel_code, channel_code)
202
203 app_id = None
204
205 app_js_url = self._search_regex(
206 r'<script[^>]+src=(["\'])(?P<url>http.+?/app\.js.*?)\1',
207 webpage, 'app js', default=None, group='url')
208
209 if app_js_url:
210 app_js = self._download_webpage(
211 app_js_url, channel_code, 'Downloading app JS', fatal=False)
212 if app_js:
213 app_id = self._search_regex(
214 r'Global\.VFAN_APP_ID\s*=\s*[\'"]([^\'"]+)[\'"]',
215 app_js, 'app id', default=None)
216
217 app_id = app_id or self._APP_ID
218
219 channel_info = self._download_json(
220 'http://api.vfan.vlive.tv/vproxy/channelplus/decodeChannelCode',
221 channel_code, note='Downloading decode channel code',
222 query={
223 'app_id': app_id,
224 'channelCode': channel_code,
225 '_': int(time.time())
226 })
227
228 channel_seq = channel_info['result']['channelSeq']
229 channel_name = None
230 entries = []
231
232 for page_num in itertools.count(1):
233 video_list = self._download_json(
234 'http://api.vfan.vlive.tv/vproxy/channelplus/getChannelVideoList',
235 channel_code, note='Downloading channel list page #%d' % page_num,
236 query={
237 'app_id': app_id,
238 'channelSeq': channel_seq,
239 # Large values of maxNumOfRows (~300 or above) may cause
240 # empty responses (see [1]), e.g. this happens for [2] that
241 # has more than 300 videos.
242 # 1. https://github.com/rg3/youtube-dl/issues/13830
243 # 2. http://channels.vlive.tv/EDBF.
244 'maxNumOfRows': 100,
245 '_': int(time.time()),
246 'pageNo': page_num
247 }
248 )
249
250 if not channel_name:
251 channel_name = try_get(
252 video_list,
253 lambda x: x['result']['channelInfo']['channelName'],
254 compat_str)
255
256 videos = try_get(
257 video_list, lambda x: x['result']['videoList'], list)
258 if not videos:
259 break
260
261 for video in videos:
262 video_id = video.get('videoSeq')
263 if not video_id:
264 continue
265 video_id = compat_str(video_id)
266 entries.append(
267 self.url_result(
268 'http://www.vlive.tv/video/%s' % video_id,
269 ie=VLiveIE.ie_key(), video_id=video_id))
270
271 return self.playlist_result(
272 entries, channel_code, channel_name)
273
274
275 class VLivePlaylistIE(InfoExtractor):
276 IE_NAME = 'vlive:playlist'
277 _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/video/(?P<video_id>[0-9]+)/playlist/(?P<id>[0-9]+)'
278 _TEST = {
279 'url': 'http://www.vlive.tv/video/22867/playlist/22912',
280 'info_dict': {
281 'id': '22912',
282 'title': 'Valentine Day Message from TWICE'
283 },
284 'playlist_mincount': 9
285 }
286
287 def _real_extract(self, url):
288 mobj = re.match(self._VALID_URL, url)
289 video_id, playlist_id = mobj.group('video_id', 'id')
290
291 VIDEO_URL_TEMPLATE = 'http://www.vlive.tv/video/%s'
292 if self._downloader.params.get('noplaylist'):
293 self.to_screen(
294 'Downloading just video %s because of --no-playlist' % video_id)
295 return self.url_result(
296 VIDEO_URL_TEMPLATE % video_id,
297 ie=VLiveIE.ie_key(), video_id=video_id)
298
299 self.to_screen(
300 'Downloading playlist %s - add --no-playlist to just download video'
301 % playlist_id)
302
303 webpage = self._download_webpage(
304 'http://www.vlive.tv/video/%s/playlist/%s'
305 % (video_id, playlist_id), playlist_id)
306
307 item_ids = self._parse_json(
308 self._search_regex(
309 r'playlistVideoSeqs\s*=\s*(\[[^]]+\])', webpage,
310 'playlist video seqs'),
311 playlist_id)
312
313 entries = [
314 self.url_result(
315 VIDEO_URL_TEMPLATE % item_id, ie=VLiveIE.ie_key(),
316 video_id=compat_str(item_id))
317 for item_id in item_ids]
318
319 playlist_name = self._html_search_regex(
320 r'<div[^>]+class="[^"]*multicam_playlist[^>]*>\s*<h3[^>]+>([^<]+)',
321 webpage, 'playlist title', fatal=False)
322
323 return self.playlist_result(entries, playlist_id, playlist_name)