2 from __future__ import unicode_literals
9 from .common import InfoExtractor
10 from ..compat import (
14 compat_urllib_parse_urlencode,
15 compat_urllib_parse_urlparse,
32 class TwitchBaseIE(InfoExtractor):
33 _VALID_URL_BASE = r'https?://(?:(?:www|go|m)\.)?twitch\.tv'
35 _API_BASE = 'https://api.twitch.tv'
36 _USHER_BASE = 'https://usher.ttvnw.net'
37 _LOGIN_FORM_URL = 'https://www.twitch.tv/login'
38 _LOGIN_POST_URL = 'https://passport.twitch.tv/login'
39 _CLIENT_ID = 'kimne78kx3ncx6brgo4mv6wki5h1ko'
40 _NETRC_MACHINE = 'twitch'
42 def _handle_error(self, response):
43 if not isinstance(response, dict):
45 error = response.get('error')
48 '%s returned error: %s - %s' % (self.IE_NAME, error, response.get('message')),
51 def _call_api(self, path, item_id, *args, **kwargs):
52 headers = kwargs.get('headers', {}).copy()
53 headers['Client-ID'] = self._CLIENT_ID
54 kwargs['headers'] = headers
55 response = self._download_json(
56 '%s/%s' % (self._API_BASE, path), item_id,
57 *args, **compat_kwargs(kwargs))
58 self._handle_error(response)
61 def _real_initialize(self):
65 username, password = self._get_login_info()
71 'Unable to login. Twitch said: %s' % message, expected=True)
73 def login_step(page, urlh, note, data):
74 form = self._hidden_inputs(page)
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)
86 'Content-Type': 'text/plain;charset=UTF-8',
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')
96 if 'Authenticated successfully' in response.get('message', ''):
99 redirect_url = urljoin(
101 response.get('redirect') or response['redirect_path'])
102 return self._download_webpage_handle(
103 redirect_url, None, 'Downloading login redirect page',
106 login_page, handle = self._download_webpage_handle(
107 self._LOGIN_FORM_URL, None, 'Downloading login page')
109 # Some TOR nodes and public proxies are blocked completely
110 if 'blacklist_message' in login_page:
111 fail(clean_html(login_page))
113 redirect_page, handle = login_step(
114 login_page, handle, 'Logging in', {
115 'username': username,
116 'password': password,
117 'client_id': self._CLIENT_ID,
121 if not redirect_page:
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',
132 def _prefer_source(self, formats):
134 source = next(f for f in formats if f['format_id'] == 'Source')
135 source['quality'] = 10
136 except StopIteration:
138 if '/chunked/' in f['url']:
141 'format_note': 'Source',
143 self._sort_formats(formats)
146 class TwitchItemBaseIE(TwitchBaseIE):
147 def _download_info(self, item, item_id):
148 return self._extract_info(self._call_api(
149 'kraken/videos/%s%s' % (item, item_id), item_id,
150 'Downloading %s info JSON' % self._ITEM_TYPE))
152 def _extract_media(self, item_id):
153 info = self._download_info(self._ITEM_SHORTCUT, item_id)
154 response = self._call_api(
155 'api/videos/%s%s' % (self._ITEM_SHORTCUT, item_id), item_id,
156 'Downloading %s playlist JSON' % self._ITEM_TYPE)
158 chunks = response['chunks']
159 qualities = list(chunks.keys())
160 for num, fragment in enumerate(zip(*chunks.values()), start=1):
162 for fmt_num, fragment_fmt in enumerate(fragment):
163 format_id = qualities[fmt_num]
165 'url': fragment_fmt['url'],
166 'format_id': format_id,
167 'quality': 1 if format_id == 'live' else 0,
169 m = re.search(r'^(?P<height>\d+)[Pp]', format_id)
171 fmt['height'] = int(m.group('height'))
173 self._sort_formats(formats)
175 entry['id'] = '%s_%d' % (entry['id'], num)
176 entry['title'] = '%s part %d' % (entry['title'], num)
177 entry['formats'] = formats
178 entries.append(entry)
179 return self.playlist_result(entries, info['id'], info['title'])
181 def _extract_info(self, info):
182 status = info.get('status')
183 if status == 'recording':
185 elif status == 'recorded':
191 'title': info.get('title') or 'Untitled Broadcast',
192 'description': info.get('description'),
193 'duration': int_or_none(info.get('length')),
194 'thumbnail': info.get('preview'),
195 'uploader': info.get('channel', {}).get('display_name'),
196 'uploader_id': info.get('channel', {}).get('name'),
197 'timestamp': parse_iso8601(info.get('recorded_at')),
198 'view_count': int_or_none(info.get('views')),
202 def _real_extract(self, url):
203 return self._extract_media(self._match_id(url))
206 class TwitchVideoIE(TwitchItemBaseIE):
207 IE_NAME = 'twitch:video'
208 _VALID_URL = r'%s/[^/]+/b/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
213 'url': 'http://www.twitch.tv/riotgames/b/577357806',
216 'title': 'Worlds Semifinals - Star Horn Royal Club vs. OMG',
218 'playlist_mincount': 12,
219 'skip': 'HTTP Error 404: Not Found',
223 class TwitchChapterIE(TwitchItemBaseIE):
224 IE_NAME = 'twitch:chapter'
225 _VALID_URL = r'%s/[^/]+/c/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
226 _ITEM_TYPE = 'chapter'
230 'url': 'http://www.twitch.tv/acracingleague/c/5285812',
233 'title': 'ACRL Off Season - Sports Cars @ Nordschleife',
235 'playlist_mincount': 3,
236 'skip': 'HTTP Error 404: Not Found',
238 'url': 'http://www.twitch.tv/tsm_theoddone/c/2349361',
239 'only_matching': True,
243 class TwitchVodIE(TwitchItemBaseIE):
244 IE_NAME = 'twitch:vod'
245 _VALID_URL = r'''(?x)
248 (?:(?:www|go|m)\.)?twitch\.tv/(?:[^/]+/v(?:ideo)?|videos)/|
249 player\.twitch\.tv/\?.*?\bvideo=v?
257 'url': 'http://www.twitch.tv/riotgames/v/6528877?t=5m10s',
261 'title': 'LCK Summer Split - Week 6 Day 1',
262 'thumbnail': r're:^https?://.*\.jpg$',
264 'timestamp': 1435131709,
265 'upload_date': '20150624',
266 'uploader': 'Riot Games',
267 'uploader_id': 'riotgames',
273 'skip_download': True,
276 # Untitled broadcast (title is None)
277 'url': 'http://www.twitch.tv/belkao_o/v/11230755',
281 'title': 'Untitled Broadcast',
282 'thumbnail': r're:^https?://.*\.jpg$',
284 'timestamp': 1439746708,
285 'upload_date': '20150816',
286 'uploader': 'BelkAO_o',
287 'uploader_id': 'belkao_o',
292 'skip_download': True,
294 'skip': 'HTTP Error 404: Not Found',
296 'url': 'http://player.twitch.tv/?t=5m10s&video=v6528877',
297 'only_matching': True,
299 'url': 'https://www.twitch.tv/videos/6528877',
300 'only_matching': True,
302 'url': 'https://m.twitch.tv/beagsandjam/v/247478721',
303 'only_matching': True,
305 'url': 'https://www.twitch.tv/northernlion/video/291940395',
306 'only_matching': True,
308 'url': 'https://player.twitch.tv/?video=480452374',
309 'only_matching': True,
312 def _real_extract(self, url):
313 item_id = self._match_id(url)
315 info = self._download_info(self._ITEM_SHORTCUT, item_id)
316 access_token = self._call_api(
317 'api/vods/%s/access_token' % item_id, item_id,
318 'Downloading %s access token' % self._ITEM_TYPE)
320 formats = self._extract_m3u8_formats(
321 '%s/vod/%s.m3u8?%s' % (
322 self._USHER_BASE, item_id,
323 compat_urllib_parse_urlencode({
324 'allow_source': 'true',
325 'allow_audio_only': 'true',
326 'allow_spectre': 'true',
327 'player': 'twitchweb',
328 'playlist_include_framerate': 'true',
329 'nauth': access_token['token'],
330 'nauthsig': access_token['sig'],
332 item_id, 'mp4', entry_protocol='m3u8_native')
334 self._prefer_source(formats)
335 info['formats'] = formats
337 parsed_url = compat_urllib_parse_urlparse(url)
338 query = compat_parse_qs(parsed_url.query)
340 info['start_time'] = parse_duration(query['t'][0])
342 if info.get('timestamp') is not None:
343 info['subtitles'] = {
345 'url': update_url_query(
346 'https://api.twitch.tv/v5/videos/%s/comments' % item_id, {
347 'client_id': self._CLIENT_ID,
356 class TwitchPlaylistBaseIE(TwitchBaseIE):
357 _PLAYLIST_PATH = 'kraken/channels/%s/videos/?offset=%d&limit=%d'
360 def _extract_playlist(self, channel_id):
361 info = self._call_api(
362 'kraken/channels/%s' % channel_id,
363 channel_id, 'Downloading channel info JSON')
364 channel_name = info.get('display_name') or info.get('name')
367 limit = self._PAGE_LIMIT
368 broken_paging_detected = False
369 counter_override = None
370 for counter in itertools.count(1):
371 response = self._call_api(
372 self._PLAYLIST_PATH % (channel_id, offset, limit),
374 'Downloading %s JSON page %s'
375 % (self._PLAYLIST_TYPE, counter_override or counter))
376 page_entries = self._extract_playlist_page(response)
379 total = int_or_none(response.get('_total'))
380 # Since the beginning of March 2016 twitch's paging mechanism
381 # is completely broken on the twitch side. It simply ignores
382 # a limit and returns the whole offset number of videos.
383 # Working around by just requesting all videos at once.
384 # Upd: pagination bug was fixed by twitch on 15.03.2016.
385 if not broken_paging_detected and total and len(page_entries) > limit:
387 'Twitch pagination is broken on twitch side, requesting all videos at once',
389 broken_paging_detected = True
391 counter_override = '(all at once)'
393 entries.extend(page_entries)
394 if broken_paging_detected or total and len(page_entries) >= total:
397 return self.playlist_result(
398 [self._make_url_result(entry) for entry in orderedSet(entries)],
399 channel_id, channel_name)
401 def _make_url_result(self, url):
403 video_id = 'v%s' % TwitchVodIE._match_id(url)
404 return self.url_result(url, TwitchVodIE.ie_key(), video_id=video_id)
405 except AssertionError:
406 return self.url_result(url)
408 def _extract_playlist_page(self, response):
409 videos = response.get('videos')
410 return [video['url'] for video in videos] if videos else []
412 def _real_extract(self, url):
413 return self._extract_playlist(self._match_id(url))
416 class TwitchProfileIE(TwitchPlaylistBaseIE):
417 IE_NAME = 'twitch:profile'
418 _VALID_URL = r'%s/(?P<id>[^/]+)/profile/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
419 _PLAYLIST_TYPE = 'profile'
422 'url': 'http://www.twitch.tv/vanillatv/profile',
425 'title': 'VanillaTV',
427 'playlist_mincount': 412,
429 'url': 'http://m.twitch.tv/vanillatv/profile',
430 'only_matching': True,
434 class TwitchVideosBaseIE(TwitchPlaylistBaseIE):
435 _VALID_URL_VIDEOS_BASE = r'%s/(?P<id>[^/]+)/videos' % TwitchBaseIE._VALID_URL_BASE
436 _PLAYLIST_PATH = TwitchPlaylistBaseIE._PLAYLIST_PATH + '&broadcast_type='
439 class TwitchAllVideosIE(TwitchVideosBaseIE):
440 IE_NAME = 'twitch:videos:all'
441 _VALID_URL = r'%s/all' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
442 _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'archive,upload,highlight'
443 _PLAYLIST_TYPE = 'all videos'
446 'url': 'https://www.twitch.tv/spamfish/videos/all',
451 'playlist_mincount': 869,
453 'url': 'https://m.twitch.tv/spamfish/videos/all',
454 'only_matching': True,
458 class TwitchUploadsIE(TwitchVideosBaseIE):
459 IE_NAME = 'twitch:videos:uploads'
460 _VALID_URL = r'%s/uploads' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
461 _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'upload'
462 _PLAYLIST_TYPE = 'uploads'
465 'url': 'https://www.twitch.tv/spamfish/videos/uploads',
470 'playlist_mincount': 0,
472 'url': 'https://m.twitch.tv/spamfish/videos/uploads',
473 'only_matching': True,
477 class TwitchPastBroadcastsIE(TwitchVideosBaseIE):
478 IE_NAME = 'twitch:videos:past-broadcasts'
479 _VALID_URL = r'%s/past-broadcasts' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
480 _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'archive'
481 _PLAYLIST_TYPE = 'past broadcasts'
484 'url': 'https://www.twitch.tv/spamfish/videos/past-broadcasts',
489 'playlist_mincount': 0,
491 'url': 'https://m.twitch.tv/spamfish/videos/past-broadcasts',
492 'only_matching': True,
496 class TwitchHighlightsIE(TwitchVideosBaseIE):
497 IE_NAME = 'twitch:videos:highlights'
498 _VALID_URL = r'%s/highlights' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
499 _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'highlight'
500 _PLAYLIST_TYPE = 'highlights'
503 'url': 'https://www.twitch.tv/spamfish/videos/highlights',
508 'playlist_mincount': 805,
510 'url': 'https://m.twitch.tv/spamfish/videos/highlights',
511 'only_matching': True,
515 class TwitchStreamIE(TwitchBaseIE):
516 IE_NAME = 'twitch:stream'
517 _VALID_URL = r'''(?x)
520 (?:(?:www|go|m)\.)?twitch\.tv/|
521 player\.twitch\.tv/\?.*?\bchannel=
527 'url': 'http://www.twitch.tv/shroomztv',
530 'display_id': 'shroomztv',
532 'title': 're:^ShroomzTV [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
533 'description': 'H1Z1 - lonewolfing with ShroomzTV | A3 Battle Royale later - @ShroomzTV',
535 'timestamp': 1421928037,
536 'upload_date': '20150122',
537 'uploader': 'ShroomzTV',
538 'uploader_id': 'shroomztv',
543 'skip_download': True,
546 'url': 'http://www.twitch.tv/miracle_doto#profile-0',
547 'only_matching': True,
549 'url': 'https://player.twitch.tv/?channel=lotsofs',
550 'only_matching': True,
552 'url': 'https://go.twitch.tv/food',
553 'only_matching': True,
555 'url': 'https://m.twitch.tv/food',
556 'only_matching': True,
560 def suitable(cls, url):
562 if any(ie.suitable(url) for ie in (
569 TwitchPastBroadcastsIE,
572 else super(TwitchStreamIE, cls).suitable(url))
574 def _real_extract(self, url):
575 channel_id = self._match_id(url)
577 stream = self._call_api(
578 'kraken/streams/%s?stream_type=all' % channel_id.lower(),
579 channel_id, 'Downloading stream JSON').get('stream')
582 raise ExtractorError('%s is offline' % channel_id, expected=True)
584 # Channel name may be typed if different case than the original channel name
585 # (e.g. http://www.twitch.tv/TWITCHPLAYSPOKEMON) that will lead to constructing
586 # an invalid m3u8 URL. Working around by use of original channel name from stream
587 # JSON and fallback to lowercase if it's not available.
588 channel_id = stream.get('channel', {}).get('name') or channel_id.lower()
590 access_token = self._call_api(
591 'api/channels/%s/access_token' % channel_id, channel_id,
592 'Downloading channel access token')
595 'allow_source': 'true',
596 'allow_audio_only': 'true',
597 'allow_spectre': 'true',
598 'p': random.randint(1000000, 10000000),
599 'player': 'twitchweb',
600 'playlist_include_framerate': 'true',
601 'segment_preference': '4',
602 'sig': access_token['sig'].encode('utf-8'),
603 'token': access_token['token'].encode('utf-8'),
605 formats = self._extract_m3u8_formats(
606 '%s/api/channel/hls/%s.m3u8?%s'
607 % (self._USHER_BASE, channel_id, compat_urllib_parse_urlencode(query)),
609 self._prefer_source(formats)
611 view_count = stream.get('viewers')
612 timestamp = parse_iso8601(stream.get('created_at'))
614 channel = stream['channel']
615 title = self._live_title(channel.get('display_name') or channel.get('name'))
616 description = channel.get('status')
619 for thumbnail_key, thumbnail_url in stream['preview'].items():
620 m = re.search(r'(?P<width>\d+)x(?P<height>\d+)\.jpg$', thumbnail_key)
624 'url': thumbnail_url,
625 'width': int(m.group('width')),
626 'height': int(m.group('height')),
630 'id': compat_str(stream['_id']),
631 'display_id': channel_id,
633 'description': description,
634 'thumbnails': thumbnails,
635 'uploader': channel.get('display_name'),
636 'uploader_id': channel.get('name'),
637 'timestamp': timestamp,
638 'view_count': view_count,
644 class TwitchClipsIE(TwitchBaseIE):
645 IE_NAME = 'twitch:clips'
646 _VALID_URL = r'''(?x)
649 clips\.twitch\.tv/(?:embed\?.*?\bclip=|(?:[^/]+/)*)|
650 (?:(?:www|go|m)\.)?twitch\.tv/[^/]+/clip/
656 'url': 'https://clips.twitch.tv/FaintLightGullWholeWheat',
657 'md5': '761769e1eafce0ffebfb4089cb3847cd',
661 'title': 'EA Play 2016 Live from the Novo Theatre',
662 'thumbnail': r're:^https?://.*\.jpg',
663 'timestamp': 1465767393,
664 'upload_date': '20160612',
666 'uploader': 'stereotype_',
667 'uploader_id': '43566419',
671 'url': 'https://clips.twitch.tv/rflegendary/UninterestedBeeDAESuppy',
672 'only_matching': True,
674 'url': 'https://www.twitch.tv/sergeynixon/clip/StormyThankfulSproutFutureMan',
675 'only_matching': True,
677 'url': 'https://clips.twitch.tv/embed?clip=InquisitiveBreakableYogurtJebaited',
678 'only_matching': True,
680 'url': 'https://m.twitch.tv/rossbroadcast/clip/ConfidentBraveHumanChefFrank',
681 'only_matching': True,
683 'url': 'https://go.twitch.tv/rossbroadcast/clip/ConfidentBraveHumanChefFrank',
684 'only_matching': True,
687 def _real_extract(self, url):
688 video_id = self._match_id(url)
690 clip = self._download_json(
691 'https://gql.twitch.tv/gql', video_id, data=json.dumps({
704 tiny: thumbnailURL(width: 86, height: 45)
705 small: thumbnailURL(width: 260, height: 147)
706 medium: thumbnailURL(width: 480, height: 272)
716 }).encode(), headers={
717 'Client-ID': self._CLIENT_ID,
721 raise ExtractorError(
722 'This clip is no longer available', expected=True)
725 for option in clip.get('videoQualities', []):
726 if not isinstance(option, dict):
728 source = url_or_none(option.get('sourceURL'))
733 'format_id': option.get('quality'),
734 'height': int_or_none(option.get('quality')),
735 'fps': int_or_none(option.get('frameRate')),
737 self._sort_formats(formats)
740 for thumbnail_id in ('tiny', 'small', 'medium'):
741 thumbnail_url = clip.get(thumbnail_id)
742 if not thumbnail_url:
746 'url': thumbnail_url,
748 mobj = re.search(r'-(\d+)x(\d+)\.', thumbnail_url)
751 'height': int(mobj.group(2)),
752 'width': int(mobj.group(1)),
754 thumbnails.append(thumb)
757 'id': clip.get('id') or video_id,
758 'title': clip.get('title') or video_id,
760 'duration': int_or_none(clip.get('durationSeconds')),
761 'views': int_or_none(clip.get('viewCount')),
762 'timestamp': unified_timestamp(clip.get('createdAt')),
763 'thumbnails': thumbnails,
764 'creator': try_get(clip, lambda x: x['broadcaster']['displayName'], compat_str),
765 'uploader': try_get(clip, lambda x: x['curator']['displayName'], compat_str),
766 'uploader_id': try_get(clip, lambda x: x['curator']['id'], compat_str),