_ Git - youtube-dl/blob - youtube_dl/extractor/twitch.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import itertools
   5 import re
   6 import random
   7 import json
   8
   9 from .common import InfoExtractor
  10 from ..compat import (
  11     compat_HTTPError,
  12     compat_kwargs,
  13     compat_parse_qs,
  14     compat_str,
  15     compat_urllib_parse_urlencode,
  16     compat_urllib_parse_urlparse,
  17 )
  18 from ..utils import (
  19     clean_html,
  20     ExtractorError,
  21     float_or_none,
  22     int_or_none,
  23     orderedSet,
  24     parse_duration,
  25     parse_iso8601,
  26     qualities,
  27     try_get,
  28     unified_timestamp,
  29     update_url_query,
  30     url_or_none,
  31     urljoin,
  32 )
  33
  34
  35 class TwitchBaseIE(InfoExtractor):
  36     _VALID_URL_BASE = r'https?://(?:(?:www|go|m)\.)?twitch\.tv'
  37
  38     _API_BASE = 'https://api.twitch.tv'
  39     _USHER_BASE = 'https://usher.ttvnw.net'
  40     _LOGIN_FORM_URL = 'https://www.twitch.tv/login'
  41     _LOGIN_POST_URL = 'https://passport.twitch.tv/login'
  42     _CLIENT_ID = 'jzkbprff40iqj646a697cyrvl0zt2m6'
  43     _NETRC_MACHINE = 'twitch'
  44
  45     def _handle_error(self, response):
  46         if not isinstance(response, dict):
  47             return
  48         error = response.get('error')
  49         if error:
  50             raise ExtractorError(
  51                 '%s returned error: %s - %s' % (self.IE_NAME, error, response.get('message')),
  52                 expected=True)
  53
  54     def _call_api(self, path, item_id, *args, **kwargs):
  55         kwargs.setdefault('headers', {})['Client-ID'] = self._CLIENT_ID
  56         response = self._download_json(
  57             '%s/%s' % (self._API_BASE, path), item_id,
  58             *args, **compat_kwargs(kwargs))
  59         self._handle_error(response)
  60         return response
  61
  62     def _real_initialize(self):
  63         self._login()
  64
  65     def _login(self):
  66         username, password = self._get_login_info()
  67         if username is None:
  68             return
  69
  70         def fail(message):
  71             raise ExtractorError(
  72                 'Unable to login. Twitch said: %s' % message, expected=True)
  73
  74         def login_step(page, urlh, note, data):
  75             form = self._hidden_inputs(page)
  76             form.update(data)
  77
  78             page_url = urlh.geturl()
  79             post_url = self._search_regex(
  80                 r'<form[^>]+action=(["\'])(?P<url>.+?)\1', page,
  81                 'post url', default=self._LOGIN_POST_URL, group='url')
  82             post_url = urljoin(page_url, post_url)
  83
  84             headers = {
  85                 'Referer': page_url,
  86                 'Origin': page_url,
  87                 'Content-Type': 'text/plain;charset=UTF-8'
  88             }
  89
  90             try:
  91                 response = self._download_json(
  92                     post_url, None, note,
  93                     data=json.dumps(form).encode(),
  94                     headers=headers)
  95             except ExtractorError as e:
  96                 if isinstance(e.cause, compat_HTTPError) and e.cause.code == 400:
  97                     response = self._parse_json(
  98                         e.cause.read().decode('utf-8'), None)
  99                     fail(response.get('error_description') or response.get('error_code'))
 100                 raise
 101
 102             if 'Authenticated successfully' in response.get('message', ''):
 103                 return None, None
 104
 105             redirect_url = urljoin(
 106                 post_url,
 107                 response.get('redirect') or response['redirect_path'])
 108             return self._download_webpage_handle(
 109                 redirect_url, None, 'Downloading login redirect page',
 110                 headers=headers)
 111
 112         login_page, handle = self._download_webpage_handle(
 113             self._LOGIN_FORM_URL, None, 'Downloading login page')
 114
 115         # Some TOR nodes and public proxies are blocked completely
 116         if 'blacklist_message' in login_page:
 117             fail(clean_html(login_page))
 118
 119         redirect_page, handle = login_step(
 120             login_page, handle, 'Logging in', {
 121                 'username': username,
 122                 'password': password,
 123                 'client_id': self._CLIENT_ID,
 124             })
 125
 126         # Successful login
 127         if not redirect_page:
 128             return
 129
 130         if re.search(r'(?i)<form[^>]+id="two-factor-submit"', redirect_page) is not None:
 131             # TODO: Add mechanism to request an SMS or phone call
 132             tfa_token = self._get_tfa_info('two-factor authentication token')
 133             login_step(redirect_page, handle, 'Submitting TFA token', {
 134                 'authy_token': tfa_token,
 135                 'remember_2fa': 'true',
 136             })
 137
 138     def _prefer_source(self, formats):
 139         try:
 140             source = next(f for f in formats if f['format_id'] == 'Source')
 141             source['preference'] = 10
 142         except StopIteration:
 143             pass  # No Source stream present
 144         self._sort_formats(formats)
 145
 146
 147 class TwitchItemBaseIE(TwitchBaseIE):
 148     def _download_info(self, item, item_id):
 149         return self._extract_info(self._call_api(
 150             'kraken/videos/%s%s' % (item, item_id), item_id,
 151             'Downloading %s info JSON' % self._ITEM_TYPE))
 152
 153     def _extract_media(self, item_id):
 154         info = self._download_info(self._ITEM_SHORTCUT, item_id)
 155         response = self._call_api(
 156             'api/videos/%s%s' % (self._ITEM_SHORTCUT, item_id), item_id,
 157             'Downloading %s playlist JSON' % self._ITEM_TYPE)
 158         entries = []
 159         chunks = response['chunks']
 160         qualities = list(chunks.keys())
 161         for num, fragment in enumerate(zip(*chunks.values()), start=1):
 162             formats = []
 163             for fmt_num, fragment_fmt in enumerate(fragment):
 164                 format_id = qualities[fmt_num]
 165                 fmt = {
 166                     'url': fragment_fmt['url'],
 167                     'format_id': format_id,
 168                     'quality': 1 if format_id == 'live' else 0,
 169                 }
 170                 m = re.search(r'^(?P<height>\d+)[Pp]', format_id)
 171                 if m:
 172                     fmt['height'] = int(m.group('height'))
 173                 formats.append(fmt)
 174             self._sort_formats(formats)
 175             entry = dict(info)
 176             entry['id'] = '%s_%d' % (entry['id'], num)
 177             entry['title'] = '%s part %d' % (entry['title'], num)
 178             entry['formats'] = formats
 179             entries.append(entry)
 180         return self.playlist_result(entries, info['id'], info['title'])
 181
 182     def _extract_info(self, info):
 183         status = info.get('status')
 184         if status == 'recording':
 185             is_live = True
 186         elif status == 'recorded':
 187             is_live = False
 188         else:
 189             is_live = None
 190         return {
 191             'id': info['_id'],
 192             'title': info.get('title') or 'Untitled Broadcast',
 193             'description': info.get('description'),
 194             'duration': int_or_none(info.get('length')),
 195             'thumbnail': info.get('preview'),
 196             'uploader': info.get('channel', {}).get('display_name'),
 197             'uploader_id': info.get('channel', {}).get('name'),
 198             'timestamp': parse_iso8601(info.get('recorded_at')),
 199             'view_count': int_or_none(info.get('views')),
 200             'is_live': is_live,
 201         }
 202
 203     def _real_extract(self, url):
 204         return self._extract_media(self._match_id(url))
 205
 206
 207 class TwitchVideoIE(TwitchItemBaseIE):
 208     IE_NAME = 'twitch:video'
 209     _VALID_URL = r'%s/[^/]+/b/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
 210     _ITEM_TYPE = 'video'
 211     _ITEM_SHORTCUT = 'a'
 212
 213     _TEST = {
 214         'url': 'http://www.twitch.tv/riotgames/b/577357806',
 215         'info_dict': {
 216             'id': 'a577357806',
 217             'title': 'Worlds Semifinals - Star Horn Royal Club vs. OMG',
 218         },
 219         'playlist_mincount': 12,
 220         'skip': 'HTTP Error 404: Not Found',
 221     }
 222
 223
 224 class TwitchChapterIE(TwitchItemBaseIE):
 225     IE_NAME = 'twitch:chapter'
 226     _VALID_URL = r'%s/[^/]+/c/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
 227     _ITEM_TYPE = 'chapter'
 228     _ITEM_SHORTCUT = 'c'
 229
 230     _TESTS = [{
 231         'url': 'http://www.twitch.tv/acracingleague/c/5285812',
 232         'info_dict': {
 233             'id': 'c5285812',
 234             'title': 'ACRL Off Season - Sports Cars @ Nordschleife',
 235         },
 236         'playlist_mincount': 3,
 237         'skip': 'HTTP Error 404: Not Found',
 238     }, {
 239         'url': 'http://www.twitch.tv/tsm_theoddone/c/2349361',
 240         'only_matching': True,
 241     }]
 242
 243
 244 class TwitchVodIE(TwitchItemBaseIE):
 245     IE_NAME = 'twitch:vod'
 246     _VALID_URL = r'''(?x)
 247                     https?://
 248                         (?:
 249                             (?:(?:www|go|m)\.)?twitch\.tv/(?:[^/]+/v(?:ideo)?|videos)/|
 250                             player\.twitch\.tv/\?.*?\bvideo=v
 251                         )
 252                         (?P<id>\d+)
 253                     '''
 254     _ITEM_TYPE = 'vod'
 255     _ITEM_SHORTCUT = 'v'
 256
 257     _TESTS = [{
 258         'url': 'http://www.twitch.tv/riotgames/v/6528877?t=5m10s',
 259         'info_dict': {
 260             'id': 'v6528877',
 261             'ext': 'mp4',
 262             'title': 'LCK Summer Split - Week 6 Day 1',
 263             'thumbnail': r're:^https?://.*\.jpg$',
 264             'duration': 17208,
 265             'timestamp': 1435131709,
 266             'upload_date': '20150624',
 267             'uploader': 'Riot Games',
 268             'uploader_id': 'riotgames',
 269             'view_count': int,
 270             'start_time': 310,
 271         },
 272         'params': {
 273             # m3u8 download
 274             'skip_download': True,
 275         },
 276     }, {
 277         # Untitled broadcast (title is None)
 278         'url': 'http://www.twitch.tv/belkao_o/v/11230755',
 279         'info_dict': {
 280             'id': 'v11230755',
 281             'ext': 'mp4',
 282             'title': 'Untitled Broadcast',
 283             'thumbnail': r're:^https?://.*\.jpg$',
 284             'duration': 1638,
 285             'timestamp': 1439746708,
 286             'upload_date': '20150816',
 287             'uploader': 'BelkAO_o',
 288             'uploader_id': 'belkao_o',
 289             'view_count': int,
 290         },
 291         'params': {
 292             # m3u8 download
 293             'skip_download': True,
 294         },
 295         'skip': 'HTTP Error 404: Not Found',
 296     }, {
 297         'url': 'http://player.twitch.tv/?t=5m10s&video=v6528877',
 298         'only_matching': True,
 299     }, {
 300         'url': 'https://www.twitch.tv/videos/6528877',
 301         'only_matching': True,
 302     }, {
 303         'url': 'https://m.twitch.tv/beagsandjam/v/247478721',
 304         'only_matching': True,
 305     }, {
 306         'url': 'https://www.twitch.tv/northernlion/video/291940395',
 307         'only_matching': True,
 308     }]
 309
 310     def _real_extract(self, url):
 311         item_id = self._match_id(url)
 312
 313         info = self._download_info(self._ITEM_SHORTCUT, item_id)
 314         access_token = self._call_api(
 315             'api/vods/%s/access_token' % item_id, item_id,
 316             'Downloading %s access token' % self._ITEM_TYPE)
 317
 318         formats = self._extract_m3u8_formats(
 319             '%s/vod/%s?%s' % (
 320                 self._USHER_BASE, item_id,
 321                 compat_urllib_parse_urlencode({
 322                     'allow_source': 'true',
 323                     'allow_audio_only': 'true',
 324                     'allow_spectre': 'true',
 325                     'player': 'twitchweb',
 326                     'nauth': access_token['token'],
 327                     'nauthsig': access_token['sig'],
 328                 })),
 329             item_id, 'mp4', entry_protocol='m3u8_native')
 330
 331         self._prefer_source(formats)
 332         info['formats'] = formats
 333
 334         parsed_url = compat_urllib_parse_urlparse(url)
 335         query = compat_parse_qs(parsed_url.query)
 336         if 't' in query:
 337             info['start_time'] = parse_duration(query['t'][0])
 338
 339         if info.get('timestamp') is not None:
 340             info['subtitles'] = {
 341                 'rechat': [{
 342                     'url': update_url_query(
 343                         'https://rechat.twitch.tv/rechat-messages', {
 344                             'video_id': 'v%s' % item_id,
 345                             'start': info['timestamp'],
 346                         }),
 347                     'ext': 'json',
 348                 }],
 349             }
 350
 351         return info
 352
 353
 354 class TwitchPlaylistBaseIE(TwitchBaseIE):
 355     _PLAYLIST_PATH = 'kraken/channels/%s/videos/?offset=%d&limit=%d'
 356     _PAGE_LIMIT = 100
 357
 358     def _extract_playlist(self, channel_id):
 359         info = self._call_api(
 360             'kraken/channels/%s' % channel_id,
 361             channel_id, 'Downloading channel info JSON')
 362         channel_name = info.get('display_name') or info.get('name')
 363         entries = []
 364         offset = 0
 365         limit = self._PAGE_LIMIT
 366         broken_paging_detected = False
 367         counter_override = None
 368         for counter in itertools.count(1):
 369             response = self._call_api(
 370                 self._PLAYLIST_PATH % (channel_id, offset, limit),
 371                 channel_id,
 372                 'Downloading %s JSON page %s'
 373                 % (self._PLAYLIST_TYPE, counter_override or counter))
 374             page_entries = self._extract_playlist_page(response)
 375             if not page_entries:
 376                 break
 377             total = int_or_none(response.get('_total'))
 378             # Since the beginning of March 2016 twitch's paging mechanism
 379             # is completely broken on the twitch side. It simply ignores
 380             # a limit and returns the whole offset number of videos.
 381             # Working around by just requesting all videos at once.
 382             # Upd: pagination bug was fixed by twitch on 15.03.2016.
 383             if not broken_paging_detected and total and len(page_entries) > limit:
 384                 self.report_warning(
 385                     'Twitch pagination is broken on twitch side, requesting all videos at once',
 386                     channel_id)
 387                 broken_paging_detected = True
 388                 offset = total
 389                 counter_override = '(all at once)'
 390                 continue
 391             entries.extend(page_entries)
 392             if broken_paging_detected or total and len(page_entries) >= total:
 393                 break
 394             offset += limit
 395         return self.playlist_result(
 396             [self._make_url_result(entry) for entry in orderedSet(entries)],
 397             channel_id, channel_name)
 398
 399     def _make_url_result(self, url):
 400         try:
 401             video_id = 'v%s' % TwitchVodIE._match_id(url)
 402             return self.url_result(url, TwitchVodIE.ie_key(), video_id=video_id)
 403         except AssertionError:
 404             return self.url_result(url)
 405
 406     def _extract_playlist_page(self, response):
 407         videos = response.get('videos')
 408         return [video['url'] for video in videos] if videos else []
 409
 410     def _real_extract(self, url):
 411         return self._extract_playlist(self._match_id(url))
 412
 413
 414 class TwitchProfileIE(TwitchPlaylistBaseIE):
 415     IE_NAME = 'twitch:profile'
 416     _VALID_URL = r'%s/(?P<id>[^/]+)/profile/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
 417     _PLAYLIST_TYPE = 'profile'
 418
 419     _TESTS = [{
 420         'url': 'http://www.twitch.tv/vanillatv/profile',
 421         'info_dict': {
 422             'id': 'vanillatv',
 423             'title': 'VanillaTV',
 424         },
 425         'playlist_mincount': 412,
 426     }, {
 427         'url': 'http://m.twitch.tv/vanillatv/profile',
 428         'only_matching': True,
 429     }]
 430
 431
 432 class TwitchVideosBaseIE(TwitchPlaylistBaseIE):
 433     _VALID_URL_VIDEOS_BASE = r'%s/(?P<id>[^/]+)/videos' % TwitchBaseIE._VALID_URL_BASE
 434     _PLAYLIST_PATH = TwitchPlaylistBaseIE._PLAYLIST_PATH + '&broadcast_type='
 435
 436
 437 class TwitchAllVideosIE(TwitchVideosBaseIE):
 438     IE_NAME = 'twitch:videos:all'
 439     _VALID_URL = r'%s/all' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
 440     _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'archive,upload,highlight'
 441     _PLAYLIST_TYPE = 'all videos'
 442
 443     _TESTS = [{
 444         'url': 'https://www.twitch.tv/spamfish/videos/all',
 445         'info_dict': {
 446             'id': 'spamfish',
 447             'title': 'Spamfish',
 448         },
 449         'playlist_mincount': 869,
 450     }, {
 451         'url': 'https://m.twitch.tv/spamfish/videos/all',
 452         'only_matching': True,
 453     }]
 454
 455
 456 class TwitchUploadsIE(TwitchVideosBaseIE):
 457     IE_NAME = 'twitch:videos:uploads'
 458     _VALID_URL = r'%s/uploads' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
 459     _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'upload'
 460     _PLAYLIST_TYPE = 'uploads'
 461
 462     _TESTS = [{
 463         'url': 'https://www.twitch.tv/spamfish/videos/uploads',
 464         'info_dict': {
 465             'id': 'spamfish',
 466             'title': 'Spamfish',
 467         },
 468         'playlist_mincount': 0,
 469     }, {
 470         'url': 'https://m.twitch.tv/spamfish/videos/uploads',
 471         'only_matching': True,
 472     }]
 473
 474
 475 class TwitchPastBroadcastsIE(TwitchVideosBaseIE):
 476     IE_NAME = 'twitch:videos:past-broadcasts'
 477     _VALID_URL = r'%s/past-broadcasts' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
 478     _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'archive'
 479     _PLAYLIST_TYPE = 'past broadcasts'
 480
 481     _TESTS = [{
 482         'url': 'https://www.twitch.tv/spamfish/videos/past-broadcasts',
 483         'info_dict': {
 484             'id': 'spamfish',
 485             'title': 'Spamfish',
 486         },
 487         'playlist_mincount': 0,
 488     }, {
 489         'url': 'https://m.twitch.tv/spamfish/videos/past-broadcasts',
 490         'only_matching': True,
 491     }]
 492
 493
 494 class TwitchHighlightsIE(TwitchVideosBaseIE):
 495     IE_NAME = 'twitch:videos:highlights'
 496     _VALID_URL = r'%s/highlights' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
 497     _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'highlight'
 498     _PLAYLIST_TYPE = 'highlights'
 499
 500     _TESTS = [{
 501         'url': 'https://www.twitch.tv/spamfish/videos/highlights',
 502         'info_dict': {
 503             'id': 'spamfish',
 504             'title': 'Spamfish',
 505         },
 506         'playlist_mincount': 805,
 507     }, {
 508         'url': 'https://m.twitch.tv/spamfish/videos/highlights',
 509         'only_matching': True,
 510     }]
 511
 512
 513 class TwitchStreamIE(TwitchBaseIE):
 514     IE_NAME = 'twitch:stream'
 515     _VALID_URL = r'''(?x)
 516                     https?://
 517                         (?:
 518                             (?:(?:www|go|m)\.)?twitch\.tv/|
 519                             player\.twitch\.tv/\?.*?\bchannel=
 520                         )
 521                         (?P<id>[^/#?]+)
 522                     '''
 523
 524     _TESTS = [{
 525         'url': 'http://www.twitch.tv/shroomztv',
 526         'info_dict': {
 527             'id': '12772022048',
 528             'display_id': 'shroomztv',
 529             'ext': 'mp4',
 530             'title': 're:^ShroomzTV [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
 531             'description': 'H1Z1 - lonewolfing with ShroomzTV | A3 Battle Royale later - @ShroomzTV',
 532             'is_live': True,
 533             'timestamp': 1421928037,
 534             'upload_date': '20150122',
 535             'uploader': 'ShroomzTV',
 536             'uploader_id': 'shroomztv',
 537             'view_count': int,
 538         },
 539         'params': {
 540             # m3u8 download
 541             'skip_download': True,
 542         },
 543     }, {
 544         'url': 'http://www.twitch.tv/miracle_doto#profile-0',
 545         'only_matching': True,
 546     }, {
 547         'url': 'https://player.twitch.tv/?channel=lotsofs',
 548         'only_matching': True,
 549     }, {
 550         'url': 'https://go.twitch.tv/food',
 551         'only_matching': True,
 552     }, {
 553         'url': 'https://m.twitch.tv/food',
 554         'only_matching': True,
 555     }]
 556
 557     @classmethod
 558     def suitable(cls, url):
 559         return (False
 560                 if any(ie.suitable(url) for ie in (
 561                     TwitchVideoIE,
 562                     TwitchChapterIE,
 563                     TwitchVodIE,
 564                     TwitchProfileIE,
 565                     TwitchAllVideosIE,
 566                     TwitchUploadsIE,
 567                     TwitchPastBroadcastsIE,
 568                     TwitchHighlightsIE))
 569                 else super(TwitchStreamIE, cls).suitable(url))
 570
 571     def _real_extract(self, url):
 572         channel_id = self._match_id(url)
 573
 574         stream = self._call_api(
 575             'kraken/streams/%s?stream_type=all' % channel_id, channel_id,
 576             'Downloading stream JSON').get('stream')
 577
 578         if not stream:
 579             raise ExtractorError('%s is offline' % channel_id, expected=True)
 580
 581         # Channel name may be typed if different case than the original channel name
 582         # (e.g. http://www.twitch.tv/TWITCHPLAYSPOKEMON) that will lead to constructing
 583         # an invalid m3u8 URL. Working around by use of original channel name from stream
 584         # JSON and fallback to lowercase if it's not available.
 585         channel_id = stream.get('channel', {}).get('name') or channel_id.lower()
 586
 587         access_token = self._call_api(
 588             'api/channels/%s/access_token' % channel_id, channel_id,
 589             'Downloading channel access token')
 590
 591         query = {
 592             'allow_source': 'true',
 593             'allow_audio_only': 'true',
 594             'allow_spectre': 'true',
 595             'p': random.randint(1000000, 10000000),
 596             'player': 'twitchweb',
 597             'segment_preference': '4',
 598             'sig': access_token['sig'].encode('utf-8'),
 599             'token': access_token['token'].encode('utf-8'),
 600         }
 601         formats = self._extract_m3u8_formats(
 602             '%s/api/channel/hls/%s.m3u8?%s'
 603             % (self._USHER_BASE, channel_id, compat_urllib_parse_urlencode(query)),
 604             channel_id, 'mp4')
 605         self._prefer_source(formats)
 606
 607         view_count = stream.get('viewers')
 608         timestamp = parse_iso8601(stream.get('created_at'))
 609
 610         channel = stream['channel']
 611         title = self._live_title(channel.get('display_name') or channel.get('name'))
 612         description = channel.get('status')
 613
 614         thumbnails = []
 615         for thumbnail_key, thumbnail_url in stream['preview'].items():
 616             m = re.search(r'(?P<width>\d+)x(?P<height>\d+)\.jpg$', thumbnail_key)
 617             if not m:
 618                 continue
 619             thumbnails.append({
 620                 'url': thumbnail_url,
 621                 'width': int(m.group('width')),
 622                 'height': int(m.group('height')),
 623             })
 624
 625         return {
 626             'id': compat_str(stream['_id']),
 627             'display_id': channel_id,
 628             'title': title,
 629             'description': description,
 630             'thumbnails': thumbnails,
 631             'uploader': channel.get('display_name'),
 632             'uploader_id': channel.get('name'),
 633             'timestamp': timestamp,
 634             'view_count': view_count,
 635             'formats': formats,
 636             'is_live': True,
 637         }
 638
 639
 640 class TwitchClipsIE(TwitchBaseIE):
 641     IE_NAME = 'twitch:clips'
 642     _VALID_URL = r'https?://clips\.twitch\.tv/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 643
 644     _TESTS = [{
 645         'url': 'https://clips.twitch.tv/FaintLightGullWholeWheat',
 646         'md5': '761769e1eafce0ffebfb4089cb3847cd',
 647         'info_dict': {
 648             'id': '42850523',
 649             'ext': 'mp4',
 650             'title': 'EA Play 2016 Live from the Novo Theatre',
 651             'thumbnail': r're:^https?://.*\.jpg',
 652             'timestamp': 1465767393,
 653             'upload_date': '20160612',
 654             'creator': 'EA',
 655             'uploader': 'stereotype_',
 656             'uploader_id': '43566419',
 657         },
 658     }, {
 659         # multiple formats
 660         'url': 'https://clips.twitch.tv/rflegendary/UninterestedBeeDAESuppy',
 661         'only_matching': True,
 662     }]
 663
 664     def _real_extract(self, url):
 665         video_id = self._match_id(url)
 666
 667         status = self._download_json(
 668             'https://clips.twitch.tv/api/v2/clips/%s/status' % video_id,
 669             video_id)
 670
 671         formats = []
 672
 673         for option in status['quality_options']:
 674             if not isinstance(option, dict):
 675                 continue
 676             source = url_or_none(option.get('source'))
 677             if not source:
 678                 continue
 679             formats.append({
 680                 'url': source,
 681                 'format_id': option.get('quality'),
 682                 'height': int_or_none(option.get('quality')),
 683                 'fps': int_or_none(option.get('frame_rate')),
 684             })
 685
 686         self._sort_formats(formats)
 687
 688         info = {
 689             'formats': formats,
 690         }
 691
 692         clip = self._call_api(
 693             'kraken/clips/%s' % video_id, video_id, fatal=False, headers={
 694                 'Accept': 'application/vnd.twitchtv.v5+json',
 695             })
 696
 697         if clip:
 698             quality_key = qualities(('tiny', 'small', 'medium'))
 699             thumbnails = []
 700             thumbnails_dict = clip.get('thumbnails')
 701             if isinstance(thumbnails_dict, dict):
 702                 for thumbnail_id, thumbnail_url in thumbnails_dict.items():
 703                     thumbnails.append({
 704                         'id': thumbnail_id,
 705                         'url': thumbnail_url,
 706                         'preference': quality_key(thumbnail_id),
 707                     })
 708
 709             info.update({
 710                 'id': clip.get('tracking_id') or video_id,
 711                 'title': clip.get('title') or video_id,
 712                 'duration': float_or_none(clip.get('duration')),
 713                 'views': int_or_none(clip.get('views')),
 714                 'timestamp': unified_timestamp(clip.get('created_at')),
 715                 'thumbnails': thumbnails,
 716                 'creator': try_get(clip, lambda x: x['broadcaster']['display_name'], compat_str),
 717                 'uploader': try_get(clip, lambda x: x['curator']['display_name'], compat_str),
 718                 'uploader_id': try_get(clip, lambda x: x['curator']['id'], compat_str),
 719             })
 720         else:
 721             info.update({
 722                 'title': video_id,
 723                 'id': video_id,
 724             })
 725
 726         return info