Raphaël G. Git Repositories - youtubedl/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_kwargs,
  12     compat_parse_qs,
  13     compat_str,
  14     compat_urllib_parse_urlencode,
  15     compat_urllib_parse_urlparse,
  16 )
  17 from ..utils import (
  18     clean_html,
  19     ExtractorError,
  20     float_or_none,
  21     int_or_none,
  22     orderedSet,
  23     parse_duration,
  24     parse_iso8601,
  25     qualities,
  26     try_get,
  27     unified_timestamp,
  28     update_url_query,
  29     url_or_none,
  30     urljoin,
  31 )
  32
  33
  34 class TwitchBaseIE(InfoExtractor):
  35     _VALID_URL_BASE = r'https?://(?:(?:www|go|m)\.)?twitch\.tv'
  36
  37     _API_BASE = 'https://api.twitch.tv'
  38     _USHER_BASE = 'https://usher.ttvnw.net'
  39     _LOGIN_FORM_URL = 'https://www.twitch.tv/login'
  40     _LOGIN_POST_URL = 'https://passport.twitch.tv/login'
  41     _CLIENT_ID = 'kimne78kx3ncx6brgo4mv6wki5h1ko'
  42     _NETRC_MACHINE = 'twitch'
  43
  44     def _handle_error(self, response):
  45         if not isinstance(response, dict):
  46             return
  47         error = response.get('error')
  48         if error:
  49             raise ExtractorError(
  50                 '%s returned error: %s - %s' % (self.IE_NAME, error, response.get('message')),
  51                 expected=True)
  52
  53     def _call_api(self, path, item_id, *args, **kwargs):
  54         kwargs.setdefault('headers', {})['Client-ID'] = self._CLIENT_ID
  55         response = self._download_json(
  56             '%s/%s' % (self._API_BASE, path), item_id,
  57             *args, **compat_kwargs(kwargs))
  58         self._handle_error(response)
  59         return response
  60
  61     def _real_initialize(self):
  62         self._login()
  63
  64     def _login(self):
  65         username, password = self._get_login_info()
  66         if username is None:
  67             return
  68
  69         def fail(message):
  70             raise ExtractorError(
  71                 'Unable to login. Twitch said: %s' % message, expected=True)
  72
  73         def login_step(page, urlh, note, data):
  74             form = self._hidden_inputs(page)
  75             form.update(data)
  76
  77             page_url = urlh.geturl()
  78             post_url = self._search_regex(
  79                 r'<form[^>]+action=(["\'])(?P<url>.+?)\1', page,
  80                 'post url', default=self._LOGIN_POST_URL, group='url')
  81             post_url = urljoin(page_url, post_url)
  82
  83             headers = {
  84                 'Referer': page_url,
  85                 'Origin': page_url,
  86                 'Content-Type': 'text/plain;charset=UTF-8',
  87             }
  88
  89             response = self._download_json(
  90                 post_url, None, note, data=json.dumps(form).encode(),
  91                 headers=headers, expected_status=400)
  92             error = response.get('error_description') or response.get('error_code')
  93             if error:
  94                 fail(error)
  95
  96             if 'Authenticated successfully' in response.get('message', ''):
  97                 return None, None
  98
  99             redirect_url = urljoin(
 100                 post_url,
 101                 response.get('redirect') or response['redirect_path'])
 102             return self._download_webpage_handle(
 103                 redirect_url, None, 'Downloading login redirect page',
 104                 headers=headers)
 105
 106         login_page, handle = self._download_webpage_handle(
 107             self._LOGIN_FORM_URL, None, 'Downloading login page')
 108
 109         # Some TOR nodes and public proxies are blocked completely
 110         if 'blacklist_message' in login_page:
 111             fail(clean_html(login_page))
 112
 113         redirect_page, handle = login_step(
 114             login_page, handle, 'Logging in', {
 115                 'username': username,
 116                 'password': password,
 117                 'client_id': self._CLIENT_ID,
 118             })
 119
 120         # Successful login
 121         if not redirect_page:
 122             return
 123
 124         if re.search(r'(?i)<form[^>]+id="two-factor-submit"', redirect_page) is not None:
 125             # TODO: Add mechanism to request an SMS or phone call
 126             tfa_token = self._get_tfa_info('two-factor authentication token')
 127             login_step(redirect_page, handle, 'Submitting TFA token', {
 128                 'authy_token': tfa_token,
 129                 'remember_2fa': 'true',
 130             })
 131
 132     def _prefer_source(self, formats):
 133         try:
 134             source = next(f for f in formats if f['format_id'] == 'Source')
 135             source['preference'] = 10
 136         except StopIteration:
 137             pass  # No Source stream present
 138         self._sort_formats(formats)
 139
 140
 141 class TwitchItemBaseIE(TwitchBaseIE):
 142     def _download_info(self, item, item_id):
 143         return self._extract_info(self._call_api(
 144             'kraken/videos/%s%s' % (item, item_id), item_id,
 145             'Downloading %s info JSON' % self._ITEM_TYPE))
 146
 147     def _extract_media(self, item_id):
 148         info = self._download_info(self._ITEM_SHORTCUT, item_id)
 149         response = self._call_api(
 150             'api/videos/%s%s' % (self._ITEM_SHORTCUT, item_id), item_id,
 151             'Downloading %s playlist JSON' % self._ITEM_TYPE)
 152         entries = []
 153         chunks = response['chunks']
 154         qualities = list(chunks.keys())
 155         for num, fragment in enumerate(zip(*chunks.values()), start=1):
 156             formats = []
 157             for fmt_num, fragment_fmt in enumerate(fragment):
 158                 format_id = qualities[fmt_num]
 159                 fmt = {
 160                     'url': fragment_fmt['url'],
 161                     'format_id': format_id,
 162                     'quality': 1 if format_id == 'live' else 0,
 163                 }
 164                 m = re.search(r'^(?P<height>\d+)[Pp]', format_id)
 165                 if m:
 166                     fmt['height'] = int(m.group('height'))
 167                 formats.append(fmt)
 168             self._sort_formats(formats)
 169             entry = dict(info)
 170             entry['id'] = '%s_%d' % (entry['id'], num)
 171             entry['title'] = '%s part %d' % (entry['title'], num)
 172             entry['formats'] = formats
 173             entries.append(entry)
 174         return self.playlist_result(entries, info['id'], info['title'])
 175
 176     def _extract_info(self, info):
 177         status = info.get('status')
 178         if status == 'recording':
 179             is_live = True
 180         elif status == 'recorded':
 181             is_live = False
 182         else:
 183             is_live = None
 184         return {
 185             'id': info['_id'],
 186             'title': info.get('title') or 'Untitled Broadcast',
 187             'description': info.get('description'),
 188             'duration': int_or_none(info.get('length')),
 189             'thumbnail': info.get('preview'),
 190             'uploader': info.get('channel', {}).get('display_name'),
 191             'uploader_id': info.get('channel', {}).get('name'),
 192             'timestamp': parse_iso8601(info.get('recorded_at')),
 193             'view_count': int_or_none(info.get('views')),
 194             'is_live': is_live,
 195         }
 196
 197     def _real_extract(self, url):
 198         return self._extract_media(self._match_id(url))
 199
 200
 201 class TwitchVideoIE(TwitchItemBaseIE):
 202     IE_NAME = 'twitch:video'
 203     _VALID_URL = r'%s/[^/]+/b/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
 204     _ITEM_TYPE = 'video'
 205     _ITEM_SHORTCUT = 'a'
 206
 207     _TEST = {
 208         'url': 'http://www.twitch.tv/riotgames/b/577357806',
 209         'info_dict': {
 210             'id': 'a577357806',
 211             'title': 'Worlds Semifinals - Star Horn Royal Club vs. OMG',
 212         },
 213         'playlist_mincount': 12,
 214         'skip': 'HTTP Error 404: Not Found',
 215     }
 216
 217
 218 class TwitchChapterIE(TwitchItemBaseIE):
 219     IE_NAME = 'twitch:chapter'
 220     _VALID_URL = r'%s/[^/]+/c/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
 221     _ITEM_TYPE = 'chapter'
 222     _ITEM_SHORTCUT = 'c'
 223
 224     _TESTS = [{
 225         'url': 'http://www.twitch.tv/acracingleague/c/5285812',
 226         'info_dict': {
 227             'id': 'c5285812',
 228             'title': 'ACRL Off Season - Sports Cars @ Nordschleife',
 229         },
 230         'playlist_mincount': 3,
 231         'skip': 'HTTP Error 404: Not Found',
 232     }, {
 233         'url': 'http://www.twitch.tv/tsm_theoddone/c/2349361',
 234         'only_matching': True,
 235     }]
 236
 237
 238 class TwitchVodIE(TwitchItemBaseIE):
 239     IE_NAME = 'twitch:vod'
 240     _VALID_URL = r'''(?x)
 241                     https?://
 242                         (?:
 243                             (?:(?:www|go|m)\.)?twitch\.tv/(?:[^/]+/v(?:ideo)?|videos)/|
 244                             player\.twitch\.tv/\?.*?\bvideo=v
 245                         )
 246                         (?P<id>\d+)
 247                     '''
 248     _ITEM_TYPE = 'vod'
 249     _ITEM_SHORTCUT = 'v'
 250
 251     _TESTS = [{
 252         'url': 'http://www.twitch.tv/riotgames/v/6528877?t=5m10s',
 253         'info_dict': {
 254             'id': 'v6528877',
 255             'ext': 'mp4',
 256             'title': 'LCK Summer Split - Week 6 Day 1',
 257             'thumbnail': r're:^https?://.*\.jpg$',
 258             'duration': 17208,
 259             'timestamp': 1435131709,
 260             'upload_date': '20150624',
 261             'uploader': 'Riot Games',
 262             'uploader_id': 'riotgames',
 263             'view_count': int,
 264             'start_time': 310,
 265         },
 266         'params': {
 267             # m3u8 download
 268             'skip_download': True,
 269         },
 270     }, {
 271         # Untitled broadcast (title is None)
 272         'url': 'http://www.twitch.tv/belkao_o/v/11230755',
 273         'info_dict': {
 274             'id': 'v11230755',
 275             'ext': 'mp4',
 276             'title': 'Untitled Broadcast',
 277             'thumbnail': r're:^https?://.*\.jpg$',
 278             'duration': 1638,
 279             'timestamp': 1439746708,
 280             'upload_date': '20150816',
 281             'uploader': 'BelkAO_o',
 282             'uploader_id': 'belkao_o',
 283             'view_count': int,
 284         },
 285         'params': {
 286             # m3u8 download
 287             'skip_download': True,
 288         },
 289         'skip': 'HTTP Error 404: Not Found',
 290     }, {
 291         'url': 'http://player.twitch.tv/?t=5m10s&video=v6528877',
 292         'only_matching': True,
 293     }, {
 294         'url': 'https://www.twitch.tv/videos/6528877',
 295         'only_matching': True,
 296     }, {
 297         'url': 'https://m.twitch.tv/beagsandjam/v/247478721',
 298         'only_matching': True,
 299     }, {
 300         'url': 'https://www.twitch.tv/northernlion/video/291940395',
 301         'only_matching': True,
 302     }]
 303
 304     def _real_extract(self, url):
 305         item_id = self._match_id(url)
 306
 307         info = self._download_info(self._ITEM_SHORTCUT, item_id)
 308         access_token = self._call_api(
 309             'api/vods/%s/access_token' % item_id, item_id,
 310             'Downloading %s access token' % self._ITEM_TYPE)
 311
 312         formats = self._extract_m3u8_formats(
 313             '%s/vod/%s?%s' % (
 314                 self._USHER_BASE, item_id,
 315                 compat_urllib_parse_urlencode({
 316                     'allow_source': 'true',
 317                     'allow_audio_only': 'true',
 318                     'allow_spectre': 'true',
 319                     'player': 'twitchweb',
 320                     'nauth': access_token['token'],
 321                     'nauthsig': access_token['sig'],
 322                 })),
 323             item_id, 'mp4', entry_protocol='m3u8_native')
 324
 325         self._prefer_source(formats)
 326         info['formats'] = formats
 327
 328         parsed_url = compat_urllib_parse_urlparse(url)
 329         query = compat_parse_qs(parsed_url.query)
 330         if 't' in query:
 331             info['start_time'] = parse_duration(query['t'][0])
 332
 333         if info.get('timestamp') is not None:
 334             info['subtitles'] = {
 335                 'rechat': [{
 336                     'url': update_url_query(
 337                         'https://rechat.twitch.tv/rechat-messages', {
 338                             'video_id': 'v%s' % item_id,
 339                             'start': info['timestamp'],
 340                         }),
 341                     'ext': 'json',
 342                 }],
 343             }
 344
 345         return info
 346
 347
 348 class TwitchPlaylistBaseIE(TwitchBaseIE):
 349     _PLAYLIST_PATH = 'kraken/channels/%s/videos/?offset=%d&limit=%d'
 350     _PAGE_LIMIT = 100
 351
 352     def _extract_playlist(self, channel_id):
 353         info = self._call_api(
 354             'kraken/channels/%s' % channel_id,
 355             channel_id, 'Downloading channel info JSON')
 356         channel_name = info.get('display_name') or info.get('name')
 357         entries = []
 358         offset = 0
 359         limit = self._PAGE_LIMIT
 360         broken_paging_detected = False
 361         counter_override = None
 362         for counter in itertools.count(1):
 363             response = self._call_api(
 364                 self._PLAYLIST_PATH % (channel_id, offset, limit),
 365                 channel_id,
 366                 'Downloading %s JSON page %s'
 367                 % (self._PLAYLIST_TYPE, counter_override or counter))
 368             page_entries = self._extract_playlist_page(response)
 369             if not page_entries:
 370                 break
 371             total = int_or_none(response.get('_total'))
 372             # Since the beginning of March 2016 twitch's paging mechanism
 373             # is completely broken on the twitch side. It simply ignores
 374             # a limit and returns the whole offset number of videos.
 375             # Working around by just requesting all videos at once.
 376             # Upd: pagination bug was fixed by twitch on 15.03.2016.
 377             if not broken_paging_detected and total and len(page_entries) > limit:
 378                 self.report_warning(
 379                     'Twitch pagination is broken on twitch side, requesting all videos at once',
 380                     channel_id)
 381                 broken_paging_detected = True
 382                 offset = total
 383                 counter_override = '(all at once)'
 384                 continue
 385             entries.extend(page_entries)
 386             if broken_paging_detected or total and len(page_entries) >= total:
 387                 break
 388             offset += limit
 389         return self.playlist_result(
 390             [self._make_url_result(entry) for entry in orderedSet(entries)],
 391             channel_id, channel_name)
 392
 393     def _make_url_result(self, url):
 394         try:
 395             video_id = 'v%s' % TwitchVodIE._match_id(url)
 396             return self.url_result(url, TwitchVodIE.ie_key(), video_id=video_id)
 397         except AssertionError:
 398             return self.url_result(url)
 399
 400     def _extract_playlist_page(self, response):
 401         videos = response.get('videos')
 402         return [video['url'] for video in videos] if videos else []
 403
 404     def _real_extract(self, url):
 405         return self._extract_playlist(self._match_id(url))
 406
 407
 408 class TwitchProfileIE(TwitchPlaylistBaseIE):
 409     IE_NAME = 'twitch:profile'
 410     _VALID_URL = r'%s/(?P<id>[^/]+)/profile/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
 411     _PLAYLIST_TYPE = 'profile'
 412
 413     _TESTS = [{
 414         'url': 'http://www.twitch.tv/vanillatv/profile',
 415         'info_dict': {
 416             'id': 'vanillatv',
 417             'title': 'VanillaTV',
 418         },
 419         'playlist_mincount': 412,
 420     }, {
 421         'url': 'http://m.twitch.tv/vanillatv/profile',
 422         'only_matching': True,
 423     }]
 424
 425
 426 class TwitchVideosBaseIE(TwitchPlaylistBaseIE):
 427     _VALID_URL_VIDEOS_BASE = r'%s/(?P<id>[^/]+)/videos' % TwitchBaseIE._VALID_URL_BASE
 428     _PLAYLIST_PATH = TwitchPlaylistBaseIE._PLAYLIST_PATH + '&broadcast_type='
 429
 430
 431 class TwitchAllVideosIE(TwitchVideosBaseIE):
 432     IE_NAME = 'twitch:videos:all'
 433     _VALID_URL = r'%s/all' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
 434     _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'archive,upload,highlight'
 435     _PLAYLIST_TYPE = 'all videos'
 436
 437     _TESTS = [{
 438         'url': 'https://www.twitch.tv/spamfish/videos/all',
 439         'info_dict': {
 440             'id': 'spamfish',
 441             'title': 'Spamfish',
 442         },
 443         'playlist_mincount': 869,
 444     }, {
 445         'url': 'https://m.twitch.tv/spamfish/videos/all',
 446         'only_matching': True,
 447     }]
 448
 449
 450 class TwitchUploadsIE(TwitchVideosBaseIE):
 451     IE_NAME = 'twitch:videos:uploads'
 452     _VALID_URL = r'%s/uploads' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
 453     _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'upload'
 454     _PLAYLIST_TYPE = 'uploads'
 455
 456     _TESTS = [{
 457         'url': 'https://www.twitch.tv/spamfish/videos/uploads',
 458         'info_dict': {
 459             'id': 'spamfish',
 460             'title': 'Spamfish',
 461         },
 462         'playlist_mincount': 0,
 463     }, {
 464         'url': 'https://m.twitch.tv/spamfish/videos/uploads',
 465         'only_matching': True,
 466     }]
 467
 468
 469 class TwitchPastBroadcastsIE(TwitchVideosBaseIE):
 470     IE_NAME = 'twitch:videos:past-broadcasts'
 471     _VALID_URL = r'%s/past-broadcasts' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
 472     _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'archive'
 473     _PLAYLIST_TYPE = 'past broadcasts'
 474
 475     _TESTS = [{
 476         'url': 'https://www.twitch.tv/spamfish/videos/past-broadcasts',
 477         'info_dict': {
 478             'id': 'spamfish',
 479             'title': 'Spamfish',
 480         },
 481         'playlist_mincount': 0,
 482     }, {
 483         'url': 'https://m.twitch.tv/spamfish/videos/past-broadcasts',
 484         'only_matching': True,
 485     }]
 486
 487
 488 class TwitchHighlightsIE(TwitchVideosBaseIE):
 489     IE_NAME = 'twitch:videos:highlights'
 490     _VALID_URL = r'%s/highlights' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
 491     _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'highlight'
 492     _PLAYLIST_TYPE = 'highlights'
 493
 494     _TESTS = [{
 495         'url': 'https://www.twitch.tv/spamfish/videos/highlights',
 496         'info_dict': {
 497             'id': 'spamfish',
 498             'title': 'Spamfish',
 499         },
 500         'playlist_mincount': 805,
 501     }, {
 502         'url': 'https://m.twitch.tv/spamfish/videos/highlights',
 503         'only_matching': True,
 504     }]
 505
 506
 507 class TwitchStreamIE(TwitchBaseIE):
 508     IE_NAME = 'twitch:stream'
 509     _VALID_URL = r'''(?x)
 510                     https?://
 511                         (?:
 512                             (?:(?:www|go|m)\.)?twitch\.tv/|
 513                             player\.twitch\.tv/\?.*?\bchannel=
 514                         )
 515                         (?P<id>[^/#?]+)
 516                     '''
 517
 518     _TESTS = [{
 519         'url': 'http://www.twitch.tv/shroomztv',
 520         'info_dict': {
 521             'id': '12772022048',
 522             'display_id': 'shroomztv',
 523             'ext': 'mp4',
 524             'title': 're:^ShroomzTV [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
 525             'description': 'H1Z1 - lonewolfing with ShroomzTV | A3 Battle Royale later - @ShroomzTV',
 526             'is_live': True,
 527             'timestamp': 1421928037,
 528             'upload_date': '20150122',
 529             'uploader': 'ShroomzTV',
 530             'uploader_id': 'shroomztv',
 531             'view_count': int,
 532         },
 533         'params': {
 534             # m3u8 download
 535             'skip_download': True,
 536         },
 537     }, {
 538         'url': 'http://www.twitch.tv/miracle_doto#profile-0',
 539         'only_matching': True,
 540     }, {
 541         'url': 'https://player.twitch.tv/?channel=lotsofs',
 542         'only_matching': True,
 543     }, {
 544         'url': 'https://go.twitch.tv/food',
 545         'only_matching': True,
 546     }, {
 547         'url': 'https://m.twitch.tv/food',
 548         'only_matching': True,
 549     }]
 550
 551     @classmethod
 552     def suitable(cls, url):
 553         return (False
 554                 if any(ie.suitable(url) for ie in (
 555                     TwitchVideoIE,
 556                     TwitchChapterIE,
 557                     TwitchVodIE,
 558                     TwitchProfileIE,
 559                     TwitchAllVideosIE,
 560                     TwitchUploadsIE,
 561                     TwitchPastBroadcastsIE,
 562                     TwitchHighlightsIE))
 563                 else super(TwitchStreamIE, cls).suitable(url))
 564
 565     def _real_extract(self, url):
 566         channel_id = self._match_id(url)
 567
 568         stream = self._call_api(
 569             'kraken/streams/%s?stream_type=all' % channel_id, channel_id,
 570             'Downloading stream JSON').get('stream')
 571
 572         if not stream:
 573             raise ExtractorError('%s is offline' % channel_id, expected=True)
 574
 575         # Channel name may be typed if different case than the original channel name
 576         # (e.g. http://www.twitch.tv/TWITCHPLAYSPOKEMON) that will lead to constructing
 577         # an invalid m3u8 URL. Working around by use of original channel name from stream
 578         # JSON and fallback to lowercase if it's not available.
 579         channel_id = stream.get('channel', {}).get('name') or channel_id.lower()
 580
 581         access_token = self._call_api(
 582             'api/channels/%s/access_token' % channel_id, channel_id,
 583             'Downloading channel access token')
 584
 585         query = {
 586             'allow_source': 'true',
 587             'allow_audio_only': 'true',
 588             'allow_spectre': 'true',
 589             'p': random.randint(1000000, 10000000),
 590             'player': 'twitchweb',
 591             'segment_preference': '4',
 592             'sig': access_token['sig'].encode('utf-8'),
 593             'token': access_token['token'].encode('utf-8'),
 594         }
 595         formats = self._extract_m3u8_formats(
 596             '%s/api/channel/hls/%s.m3u8?%s'
 597             % (self._USHER_BASE, channel_id, compat_urllib_parse_urlencode(query)),
 598             channel_id, 'mp4')
 599         self._prefer_source(formats)
 600
 601         view_count = stream.get('viewers')
 602         timestamp = parse_iso8601(stream.get('created_at'))
 603
 604         channel = stream['channel']
 605         title = self._live_title(channel.get('display_name') or channel.get('name'))
 606         description = channel.get('status')
 607
 608         thumbnails = []
 609         for thumbnail_key, thumbnail_url in stream['preview'].items():
 610             m = re.search(r'(?P<width>\d+)x(?P<height>\d+)\.jpg$', thumbnail_key)
 611             if not m:
 612                 continue
 613             thumbnails.append({
 614                 'url': thumbnail_url,
 615                 'width': int(m.group('width')),
 616                 'height': int(m.group('height')),
 617             })
 618
 619         return {
 620             'id': compat_str(stream['_id']),
 621             'display_id': channel_id,
 622             'title': title,
 623             'description': description,
 624             'thumbnails': thumbnails,
 625             'uploader': channel.get('display_name'),
 626             'uploader_id': channel.get('name'),
 627             'timestamp': timestamp,
 628             'view_count': view_count,
 629             'formats': formats,
 630             'is_live': True,
 631         }
 632
 633
 634 class TwitchClipsIE(TwitchBaseIE):
 635     IE_NAME = 'twitch:clips'
 636     _VALID_URL = r'https?://clips\.twitch\.tv/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 637
 638     _TESTS = [{
 639         'url': 'https://clips.twitch.tv/FaintLightGullWholeWheat',
 640         'md5': '761769e1eafce0ffebfb4089cb3847cd',
 641         'info_dict': {
 642             'id': '42850523',
 643             'ext': 'mp4',
 644             'title': 'EA Play 2016 Live from the Novo Theatre',
 645             'thumbnail': r're:^https?://.*\.jpg',
 646             'timestamp': 1465767393,
 647             'upload_date': '20160612',
 648             'creator': 'EA',
 649             'uploader': 'stereotype_',
 650             'uploader_id': '43566419',
 651         },
 652     }, {
 653         # multiple formats
 654         'url': 'https://clips.twitch.tv/rflegendary/UninterestedBeeDAESuppy',
 655         'only_matching': True,
 656     }]
 657
 658     def _real_extract(self, url):
 659         video_id = self._match_id(url)
 660
 661         status = self._download_json(
 662             'https://clips.twitch.tv/api/v2/clips/%s/status' % video_id,
 663             video_id)
 664
 665         formats = []
 666
 667         for option in status['quality_options']:
 668             if not isinstance(option, dict):
 669                 continue
 670             source = url_or_none(option.get('source'))
 671             if not source:
 672                 continue
 673             formats.append({
 674                 'url': source,
 675                 'format_id': option.get('quality'),
 676                 'height': int_or_none(option.get('quality')),
 677                 'fps': int_or_none(option.get('frame_rate')),
 678             })
 679
 680         self._sort_formats(formats)
 681
 682         info = {
 683             'formats': formats,
 684         }
 685
 686         clip = self._call_api(
 687             'kraken/clips/%s' % video_id, video_id, fatal=False, headers={
 688                 'Accept': 'application/vnd.twitchtv.v5+json',
 689             })
 690
 691         if clip:
 692             quality_key = qualities(('tiny', 'small', 'medium'))
 693             thumbnails = []
 694             thumbnails_dict = clip.get('thumbnails')
 695             if isinstance(thumbnails_dict, dict):
 696                 for thumbnail_id, thumbnail_url in thumbnails_dict.items():
 697                     thumbnails.append({
 698                         'id': thumbnail_id,
 699                         'url': thumbnail_url,
 700                         'preference': quality_key(thumbnail_id),
 701                     })
 702
 703             info.update({
 704                 'id': clip.get('tracking_id') or video_id,
 705                 'title': clip.get('title') or video_id,
 706                 'duration': float_or_none(clip.get('duration')),
 707                 'views': int_or_none(clip.get('views')),
 708                 'timestamp': unified_timestamp(clip.get('created_at')),
 709                 'thumbnails': thumbnails,
 710                 'creator': try_get(clip, lambda x: x['broadcaster']['display_name'], compat_str),
 711                 'uploader': try_get(clip, lambda x: x['curator']['display_name'], compat_str),
 712                 'uploader_id': try_get(clip, lambda x: x['curator']['id'], compat_str),
 713             })
 714         else:
 715             info.update({
 716                 'title': video_id,
 717                 'id': video_id,
 718             })
 719
 720         return info