youtube_dlc/extractor/vlive.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import itertools
   5 import json
   6
   7 from .naver import NaverBaseIE
   8 from ..compat import (
   9     compat_HTTPError,
  10     compat_str,
  11 )
  12 from ..utils import (
  13     ExtractorError,
  14     int_or_none,
  15     merge_dicts,
  16     str_or_none,
  17     strip_or_none,
  18     try_get,
  19     urlencode_postdata,
  20 )
  21
  22
  23 class VLiveBaseIE(NaverBaseIE):
  24     _APP_ID = '8c6cc7b45d2568fb668be6e05b6e5a3b'
  25
  26
  27 class VLiveIE(VLiveBaseIE):
  28     IE_NAME = 'vlive'
  29     _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/(?:video|embed)/(?P<id>[0-9]+)'
  30     _NETRC_MACHINE = 'vlive'
  31     _TESTS = [{
  32         'url': 'http://www.vlive.tv/video/1326',
  33         'md5': 'cc7314812855ce56de70a06a27314983',
  34         'info_dict': {
  35             'id': '1326',
  36             'ext': 'mp4',
  37             'title': "Girl's Day's Broadcast",
  38             'creator': "Girl's Day",
  39             'view_count': int,
  40             'uploader_id': 'muploader_a',
  41         },
  42     }, {
  43         'url': 'http://www.vlive.tv/video/16937',
  44         'info_dict': {
  45             'id': '16937',
  46             'ext': 'mp4',
  47             'title': '첸백시 걍방',
  48             'creator': 'EXO',
  49             'view_count': int,
  50             'subtitles': 'mincount:12',
  51             'uploader_id': 'muploader_j',
  52         },
  53         'params': {
  54             'skip_download': True,
  55         },
  56     }, {
  57         'url': 'https://www.vlive.tv/video/129100',
  58         'md5': 'ca2569453b79d66e5b919e5d308bff6b',
  59         'info_dict': {
  60             'id': '129100',
  61             'ext': 'mp4',
  62             'title': '[V LIVE] [BTS+] Run BTS! 2019 - EP.71 :: Behind the scene',
  63             'creator': 'BTS+',
  64             'view_count': int,
  65             'subtitles': 'mincount:10',
  66         },
  67         'skip': 'This video is only available for CH+ subscribers',
  68     }, {
  69         'url': 'https://www.vlive.tv/embed/1326',
  70         'only_matching': True,
  71     }, {
  72         # works only with gcc=KR
  73         'url': 'https://www.vlive.tv/video/225019',
  74         'only_matching': True,
  75     }]
  76
  77     def _real_initialize(self):
  78         self._login()
  79
  80     def _login(self):
  81         email, password = self._get_login_info()
  82         if None in (email, password):
  83             return
  84
  85         def is_logged_in():
  86             login_info = self._download_json(
  87                 'https://www.vlive.tv/auth/loginInfo', None,
  88                 note='Downloading login info',
  89                 headers={'Referer': 'https://www.vlive.tv/home'})
  90             return try_get(
  91                 login_info, lambda x: x['message']['login'], bool) or False
  92
  93         LOGIN_URL = 'https://www.vlive.tv/auth/email/login'
  94         self._request_webpage(
  95             LOGIN_URL, None, note='Downloading login cookies')
  96
  97         self._download_webpage(
  98             LOGIN_URL, None, note='Logging in',
  99             data=urlencode_postdata({'email': email, 'pwd': password}),
 100             headers={
 101                 'Referer': LOGIN_URL,
 102                 'Content-Type': 'application/x-www-form-urlencoded'
 103             })
 104
 105         if not is_logged_in():
 106             raise ExtractorError('Unable to log in', expected=True)
 107
 108     def _call_api(self, path_template, video_id, fields=None):
 109         query = {'appId': self._APP_ID, 'gcc': 'KR'}
 110         if fields:
 111             query['fields'] = fields
 112         try:
 113             return self._download_json(
 114                 'https://www.vlive.tv/globalv-web/vam-web/' + path_template % video_id, video_id,
 115                 'Downloading %s JSON metadata' % path_template.split('/')[-1].split('-')[0],
 116                 headers={'Referer': 'https://www.vlive.tv/'}, query=query)
 117         except ExtractorError as e:
 118             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 119                 self.raise_login_required(json.loads(e.cause.read().decode())['message'])
 120             raise
 121
 122     def _real_extract(self, url):
 123         video_id = self._match_id(url)
 124
 125         post = self._call_api(
 126             'post/v1.0/officialVideoPost-%s', video_id,
 127             'author{nickname},channel{channelCode,channelName},officialVideo{commentCount,exposeStatus,likeCount,playCount,playTime,status,title,type,vodId}')
 128
 129         video = post['officialVideo']
 130
 131         def get_common_fields():
 132             channel = post.get('channel') or {}
 133             return {
 134                 'title': video.get('title'),
 135                 'creator': post.get('author', {}).get('nickname'),
 136                 'channel': channel.get('channelName'),
 137                 'channel_id': channel.get('channelCode'),
 138                 'duration': int_or_none(video.get('playTime')),
 139                 'view_count': int_or_none(video.get('playCount')),
 140                 'like_count': int_or_none(video.get('likeCount')),
 141                 'comment_count': int_or_none(video.get('commentCount')),
 142             }
 143
 144         video_type = video.get('type')
 145         if video_type == 'VOD':
 146             inkey = self._call_api('video/v1.0/vod/%s/inkey', video_id)['inkey']
 147             vod_id = video['vodId']
 148             return merge_dicts(
 149                 get_common_fields(),
 150                 self._extract_video_info(video_id, vod_id, inkey))
 151         elif video_type == 'LIVE':
 152             status = video.get('status')
 153             if status == 'ON_AIR':
 154                 stream_url = self._call_api(
 155                     'old/v3/live/%s/playInfo',
 156                     video_id)['result']['adaptiveStreamUrl']
 157                 formats = self._extract_m3u8_formats(stream_url, video_id, 'mp4')
 158                 info = get_common_fields()
 159                 info.update({
 160                     'title': self._live_title(video['title']),
 161                     'id': video_id,
 162                     'formats': formats,
 163                     'is_live': True,
 164                 })
 165                 return info
 166             elif status == 'ENDED':
 167                 raise ExtractorError(
 168                     'Uploading for replay. Please wait...', expected=True)
 169             elif status == 'RESERVED':
 170                 raise ExtractorError('Coming soon!', expected=True)
 171             elif video.get('exposeStatus') == 'CANCEL':
 172                 raise ExtractorError(
 173                     'We are sorry, but the live broadcast has been canceled.',
 174                     expected=True)
 175             else:
 176                 raise ExtractorError('Unknown status ' + status)
 177
 178
 179 class VLivePostIE(VLiveIE):
 180     IE_NAME = 'vlive:post'
 181     _VALID_URL = r'https?://(?:(?:www|m)\.)?vlive\.tv/post/(?P<id>\d-\d+)'
 182     _TESTS = [{
 183         # uploadType = SOS
 184         'url': 'https://www.vlive.tv/post/1-20088044',
 185         'info_dict': {
 186             'id': '1-20088044',
 187             'title': 'Hola estrellitas la tierra les dice hola (si era así no?) Ha...',
 188             'description': 'md5:fab8a1e50e6e51608907f46c7fa4b407',
 189         },
 190         'playlist_count': 3,
 191     }, {
 192         # uploadType = V
 193         'url': 'https://www.vlive.tv/post/1-20087926',
 194         'info_dict': {
 195             'id': '1-20087926',
 196             'title': 'James Corden: And so, the baby becamos the Papa💜😭💪😭',
 197         },
 198         'playlist_count': 1,
 199     }]
 200     _FVIDEO_TMPL = 'fvideo/v1.0/fvideo-%%s/%s'
 201     _SOS_TMPL = _FVIDEO_TMPL % 'sosPlayInfo'
 202     _INKEY_TMPL = _FVIDEO_TMPL % 'inKey'
 203
 204     def _real_extract(self, url):
 205         post_id = self._match_id(url)
 206
 207         post = self._call_api(
 208             'post/v1.0/post-%s', post_id,
 209             'attachments{video},officialVideo{videoSeq},plainBody,title')
 210
 211         video_seq = str_or_none(try_get(
 212             post, lambda x: x['officialVideo']['videoSeq']))
 213         if video_seq:
 214             return self.url_result(
 215                 'http://www.vlive.tv/video/' + video_seq,
 216                 VLiveIE.ie_key(), video_seq)
 217
 218         title = post['title']
 219         entries = []
 220         for idx, video in enumerate(post['attachments']['video'].values()):
 221             video_id = video.get('videoId')
 222             if not video_id:
 223                 continue
 224             upload_type = video.get('uploadType')
 225             upload_info = video.get('uploadInfo') or {}
 226             entry = None
 227             if upload_type == 'SOS':
 228                 download = self._call_api(
 229                     self._SOS_TMPL, video_id)['videoUrl']['download']
 230                 formats = []
 231                 for f_id, f_url in download.items():
 232                     formats.append({
 233                         'format_id': f_id,
 234                         'url': f_url,
 235                         'height': int_or_none(f_id[:-1]),
 236                     })
 237                 self._sort_formats(formats)
 238                 entry = {
 239                     'formats': formats,
 240                     'id': video_id,
 241                     'thumbnail': upload_info.get('imageUrl'),
 242                 }
 243             elif upload_type == 'V':
 244                 vod_id = upload_info.get('videoId')
 245                 if not vod_id:
 246                     continue
 247                 inkey = self._call_api(self._INKEY_TMPL, video_id)['inKey']
 248                 entry = self._extract_video_info(video_id, vod_id, inkey)
 249             if entry:
 250                 entry['title'] = '%s_part%s' % (title, idx)
 251                 entries.append(entry)
 252         return self.playlist_result(
 253             entries, post_id, title, strip_or_none(post.get('plainBody')))
 254
 255
 256 class VLiveChannelIE(VLiveBaseIE):
 257     IE_NAME = 'vlive:channel'
 258     _VALID_URL = r'https?://(?:channels\.vlive\.tv|(?:(?:www|m)\.)?vlive\.tv/channel)/(?P<id>[0-9A-Z]+)'
 259     _TESTS = [{
 260         'url': 'http://channels.vlive.tv/FCD4B',
 261         'info_dict': {
 262             'id': 'FCD4B',
 263             'title': 'MAMAMOO',
 264         },
 265         'playlist_mincount': 110
 266     }, {
 267         'url': 'https://www.vlive.tv/channel/FCD4B',
 268         'only_matching': True,
 269     }]
 270
 271     def _call_api(self, path, channel_key_suffix, channel_value, note, query):
 272         q = {
 273             'app_id': self._APP_ID,
 274             'channel' + channel_key_suffix: channel_value,
 275         }
 276         q.update(query)
 277         return self._download_json(
 278             'http://api.vfan.vlive.tv/vproxy/channelplus/' + path,
 279             channel_value, note='Downloading ' + note, query=q)['result']
 280
 281     def _real_extract(self, url):
 282         channel_code = self._match_id(url)
 283
 284         channel_seq = self._call_api(
 285             'decodeChannelCode', 'Code', channel_code,
 286             'decode channel code', {})['channelSeq']
 287
 288         channel_name = None
 289         entries = []
 290
 291         for page_num in itertools.count(1):
 292             video_list = self._call_api(
 293                 'getChannelVideoList', 'Seq', channel_seq,
 294                 'channel list page #%d' % page_num, {
 295                     # Large values of maxNumOfRows (~300 or above) may cause
 296                     # empty responses (see [1]), e.g. this happens for [2] that
 297                     # has more than 300 videos.
 298                     # 1. https://github.com/ytdl-org/youtube-dl/issues/13830
 299                     # 2. http://channels.vlive.tv/EDBF.
 300                     'maxNumOfRows': 100,
 301                     'pageNo': page_num
 302                 }
 303             )
 304
 305             if not channel_name:
 306                 channel_name = try_get(
 307                     video_list,
 308                     lambda x: x['channelInfo']['channelName'],
 309                     compat_str)
 310
 311             videos = try_get(
 312                 video_list, lambda x: x['videoList'], list)
 313             if not videos:
 314                 break
 315
 316             for video in videos:
 317                 video_id = video.get('videoSeq')
 318                 video_type = video.get('videoType')
 319
 320                 if not video_id or not video_type:
 321                     continue
 322                 video_id = compat_str(video_id)
 323
 324                 if video_type in ('PLAYLIST'):
 325                     playlist_videos = try_get(
 326                         video,
 327                         lambda x: x['videoPlaylist']['videoList'], list)
 328                     if not playlist_videos:
 329                         continue
 330
 331                     for playlist_video in playlist_videos:
 332                         playlist_video_id = playlist_video.get('videoSeq')
 333                         if not playlist_video_id:
 334                             continue
 335                         playlist_video_id = compat_str(playlist_video_id)
 336
 337                         entries.append(
 338                             self.url_result(
 339                                 'http://www.vlive.tv/video/%s' % playlist_video_id,
 340                                 ie=VLiveIE.ie_key(), video_id=playlist_video_id))
 341                 else:
 342                     entries.append(
 343                         self.url_result(
 344                             'http://www.vlive.tv/video/%s' % video_id,
 345                             ie=VLiveIE.ie_key(), video_id=video_id))
 346
 347         return self.playlist_result(
 348             entries, channel_code, channel_name)