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
,
34 class TwitchBaseIE(InfoExtractor
):
35 _VALID_URL_BASE
= r
'https?://(?:(?:www|go|m)\.)?twitch\.tv'
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'
44 def _handle_error(self
, response
):
45 if not isinstance(response
, dict):
47 error
= response
.get('error')
50 '%s returned error: %s - %s' % (self
.IE_NAME
, error
, response
.get('message')),
53 def _call_api(self
, path
, item_id
, *args
, **kwargs
):
54 headers
= kwargs
.get('headers', {}).copy()
55 headers
['Client-ID'] = self
._CLIENT
_ID
56 kwargs
['headers'] = headers
57 response
= self
._download
_json
(
58 '%s/%s' % (self
._API
_BASE
, path
), item_id
,
59 *args
, **compat_kwargs(kwargs
))
60 self
._handle
_error
(response
)
63 def _real_initialize(self
):
67 username
, password
= self
._get
_login
_info
()
73 'Unable to login. Twitch said: %s' % message
, expected
=True)
75 def login_step(page
, urlh
, note
, data
):
76 form
= self
._hidden
_inputs
(page
)
79 page_url
= urlh
.geturl()
80 post_url
= self
._search
_regex
(
81 r
'<form[^>]+action=(["\'])(?P
<url
>.+?
)\
1', page,
82 'post url
', default=self._LOGIN_POST_URL, group='url
')
83 post_url = urljoin(page_url, post_url)
88 'Content
-Type
': 'text
/plain
;charset
=UTF
-8',
91 response = self._download_json(
92 post_url, None, note, data=json.dumps(form).encode(),
93 headers=headers, expected_status=400)
94 error = response.get('error_description
') or response.get('error_code
')
98 if 'Authenticated successfully
' in response.get('message
', ''):
101 redirect_url = urljoin(
103 response.get('redirect
') or response['redirect_path
'])
104 return self._download_webpage_handle(
105 redirect_url, None, 'Downloading login redirect page
',
108 login_page, handle = self._download_webpage_handle(
109 self._LOGIN_FORM_URL, None, 'Downloading login page
')
111 # Some TOR nodes and public proxies are blocked completely
112 if 'blacklist_message
' in login_page:
113 fail(clean_html(login_page))
115 redirect_page, handle = login_step(
116 login_page, handle, 'Logging
in', {
117 'username
': username,
118 'password
': password,
119 'client_id
': self._CLIENT_ID,
123 if not redirect_page:
126 if re.search(r'(?i
)<form
[^
>]+id="two-factor-submit"', redirect_page) is not None:
127 # TODO: Add mechanism to request an SMS or phone call
128 tfa_token = self._get_tfa_info('two
-factor authentication token
')
129 login_step(redirect_page, handle, 'Submitting TFA token
', {
130 'authy_token
': tfa_token,
131 'remember_2fa
': 'true
',
134 def _prefer_source(self, formats):
136 source = next(f for f in formats if f['format_id
'] == 'Source
')
137 source['preference
'] = 10
138 except StopIteration:
139 pass # No Source stream present
140 self._sort_formats(formats)
143 class TwitchItemBaseIE(TwitchBaseIE):
144 def _download_info(self, item, item_id):
145 return self._extract_info(self._call_api(
146 'kraken
/videos
/%s%s' % (item, item_id), item_id,
147 'Downloading
%s info JSON
' % self._ITEM_TYPE))
149 def _extract_media(self, item_id):
150 info = self._download_info(self._ITEM_SHORTCUT, item_id)
151 response = self._call_api(
152 'api
/videos
/%s%s' % (self._ITEM_SHORTCUT, item_id), item_id,
153 'Downloading
%s playlist JSON
' % self._ITEM_TYPE)
155 chunks = response['chunks
']
156 qualities = list(chunks.keys())
157 for num, fragment in enumerate(zip(*chunks.values()), start=1):
159 for fmt_num, fragment_fmt in enumerate(fragment):
160 format_id = qualities[fmt_num]
162 'url
': fragment_fmt['url
'],
163 'format_id
': format_id,
164 'quality
': 1 if format_id == 'live
' else 0,
166 m = re.search(r'^
(?P
<height
>\d
+)[Pp
]', format_id)
168 fmt['height
'] = int(m.group('height
'))
170 self._sort_formats(formats)
172 entry['id'] = '%s_%d' % (entry['id'], num)
173 entry['title
'] = '%s part
%d' % (entry['title
'], num)
174 entry['formats
'] = formats
175 entries.append(entry)
176 return self.playlist_result(entries, info['id'], info['title
'])
178 def _extract_info(self, info):
179 status = info.get('status
')
180 if status == 'recording
':
182 elif status == 'recorded
':
188 'title
': info.get('title
') or 'Untitled Broadcast
',
189 'description
': info.get('description
'),
190 'duration
': int_or_none(info.get('length
')),
191 'thumbnail
': info.get('preview
'),
192 'uploader
': info.get('channel
', {}).get('display_name
'),
193 'uploader_id
': info.get('channel
', {}).get('name
'),
194 'timestamp
': parse_iso8601(info.get('recorded_at
')),
195 'view_count
': int_or_none(info.get('views
')),
199 def _real_extract(self, url):
200 return self._extract_media(self._match_id(url))
203 class TwitchVideoIE(TwitchItemBaseIE):
204 IE_NAME = 'twitch
:video
'
205 _VALID_URL = r'%s/[^
/]+/b
/(?P
<id>\d
+)' % TwitchBaseIE._VALID_URL_BASE
210 'url
': 'http
://www
.twitch
.tv
/riotgames
/b
/577357806',
213 'title
': 'Worlds Semifinals
- Star Horn Royal Club vs
. OMG
',
215 'playlist_mincount
': 12,
216 'skip
': 'HTTP Error
404: Not Found
',
220 class TwitchChapterIE(TwitchItemBaseIE):
221 IE_NAME = 'twitch
:chapter
'
222 _VALID_URL = r'%s/[^
/]+/c
/(?P
<id>\d
+)' % TwitchBaseIE._VALID_URL_BASE
223 _ITEM_TYPE = 'chapter
'
227 'url
': 'http
://www
.twitch
.tv
/acracingleague
/c
/5285812',
230 'title
': 'ACRL Off Season
- Sports Cars
@ Nordschleife
',
232 'playlist_mincount
': 3,
233 'skip
': 'HTTP Error
404: Not Found
',
235 'url
': 'http
://www
.twitch
.tv
/tsm_theoddone
/c
/2349361',
236 'only_matching
': True,
240 class TwitchVodIE(TwitchItemBaseIE):
241 IE_NAME = 'twitch
:vod
'
242 _VALID_URL = r'''(?x)
245 (?:(?:www|go|m)\.)?twitch\.tv/(?:[^/]+/v(?:ideo)?|videos)/|
246 player\.twitch\.tv/\?.*?\bvideo=v
254 'url
': 'http
://www
.twitch
.tv
/riotgames
/v
/6528877?t
=5m10s
',
258 'title
': 'LCK Summer Split
- Week
6 Day
1',
259 'thumbnail
': r're
:^https?
://.*\
.jpg$
',
261 'timestamp
': 1435131709,
262 'upload_date
': '20150624',
263 'uploader
': 'Riot Games
',
264 'uploader_id
': 'riotgames
',
270 'skip_download
': True,
273 # Untitled broadcast (title is None)
274 'url
': 'http
://www
.twitch
.tv
/belkao_o
/v
/11230755',
278 'title
': 'Untitled Broadcast
',
279 'thumbnail
': r're
:^https?
://.*\
.jpg$
',
281 'timestamp
': 1439746708,
282 'upload_date
': '20150816',
283 'uploader
': 'BelkAO_o
',
284 'uploader_id
': 'belkao_o
',
289 'skip_download
': True,
291 'skip
': 'HTTP Error
404: Not Found
',
293 'url
': 'http
://player
.twitch
.tv
/?t
=5m10s
&video
=v6528877
',
294 'only_matching
': True,
296 'url
': 'https
://www
.twitch
.tv
/videos
/6528877',
297 'only_matching
': True,
299 'url
': 'https
://m
.twitch
.tv
/beagsandjam
/v
/247478721',
300 'only_matching
': True,
302 'url
': 'https
://www
.twitch
.tv
/northernlion
/video
/291940395',
303 'only_matching
': True,
306 def _real_extract(self, url):
307 item_id = self._match_id(url)
309 info = self._download_info(self._ITEM_SHORTCUT, item_id)
310 access_token = self._call_api(
311 'api
/vods
/%s/access_token
' % item_id, item_id,
312 'Downloading
%s access token
' % self._ITEM_TYPE)
314 formats = self._extract_m3u8_formats(
316 self._USHER_BASE, item_id,
317 compat_urllib_parse_urlencode({
318 'allow_source
': 'true
',
319 'allow_audio_only
': 'true
',
320 'allow_spectre
': 'true
',
321 'player
': 'twitchweb
',
322 'nauth
': access_token['token
'],
323 'nauthsig
': access_token['sig
'],
325 item_id, 'mp4
', entry_protocol='m3u8_native
')
327 self._prefer_source(formats)
328 info['formats
'] = formats
330 parsed_url = compat_urllib_parse_urlparse(url)
331 query = compat_parse_qs(parsed_url.query)
333 info['start_time
'] = parse_duration(query['t
'][0])
335 if info.get('timestamp
') is not None:
336 info['subtitles
'] = {
338 'url
': update_url_query(
339 'https
://rechat
.twitch
.tv
/rechat
-messages
', {
340 'video_id
': 'v
%s' % item_id,
341 'start
': info['timestamp
'],
350 class TwitchPlaylistBaseIE(TwitchBaseIE):
351 _PLAYLIST_PATH = 'kraken
/channels
/%s/videos
/?offset
=%d&limit
=%d'
354 def _extract_playlist(self, channel_id):
355 info = self._call_api(
356 'kraken
/channels
/%s' % channel_id,
357 channel_id, 'Downloading channel info JSON
')
358 channel_name = info.get('display_name
') or info.get('name
')
361 limit = self._PAGE_LIMIT
362 broken_paging_detected = False
363 counter_override = None
364 for counter in itertools.count(1):
365 response = self._call_api(
366 self._PLAYLIST_PATH % (channel_id, offset, limit),
368 'Downloading
%s JSON page
%s'
369 % (self._PLAYLIST_TYPE, counter_override or counter))
370 page_entries = self._extract_playlist_page(response)
373 total = int_or_none(response.get('_total
'))
374 # Since the beginning of March 2016 twitch's paging mechanism
375 # is completely broken on the twitch side. It simply ignores
376 # a limit and returns the whole offset number of videos.
377 # Working around by just requesting all videos at once.
378 # Upd: pagination bug was fixed by twitch on 15.03.2016.
379 if not broken_paging_detected
and total
and len(page_entries
) > limit
:
381 'Twitch pagination is broken on twitch side, requesting all videos at once',
383 broken_paging_detected
= True
385 counter_override
= '(all at once)'
387 entries
.extend(page_entries
)
388 if broken_paging_detected
or total
and len(page_entries
) >= total
:
391 return self
.playlist_result(
392 [self
._make
_url
_result
(entry
) for entry
in orderedSet(entries
)],
393 channel_id
, channel_name
)
395 def _make_url_result(self
, url
):
397 video_id
= 'v%s' % TwitchVodIE
._match
_id
(url
)
398 return self
.url_result(url
, TwitchVodIE
.ie_key(), video_id
=video_id
)
399 except AssertionError:
400 return self
.url_result(url
)
402 def _extract_playlist_page(self
, response
):
403 videos
= response
.get('videos')
404 return [video
['url'] for video
in videos
] if videos
else []
406 def _real_extract(self
, url
):
407 return self
._extract
_playlist
(self
._match
_id
(url
))
410 class TwitchProfileIE(TwitchPlaylistBaseIE
):
411 IE_NAME
= 'twitch:profile'
412 _VALID_URL
= r
'%s/(?P<id>[^/]+)/profile/?(?:\#.*)?$' % TwitchBaseIE
._VALID
_URL
_BASE
413 _PLAYLIST_TYPE
= 'profile'
416 'url': 'http://www.twitch.tv/vanillatv/profile',
419 'title': 'VanillaTV',
421 'playlist_mincount': 412,
423 'url': 'http://m.twitch.tv/vanillatv/profile',
424 'only_matching': True,
428 class TwitchVideosBaseIE(TwitchPlaylistBaseIE
):
429 _VALID_URL_VIDEOS_BASE
= r
'%s/(?P<id>[^/]+)/videos' % TwitchBaseIE
._VALID
_URL
_BASE
430 _PLAYLIST_PATH
= TwitchPlaylistBaseIE
._PLAYLIST
_PATH
+ '&broadcast_type='
433 class TwitchAllVideosIE(TwitchVideosBaseIE
):
434 IE_NAME
= 'twitch:videos:all'
435 _VALID_URL
= r
'%s/all' % TwitchVideosBaseIE
._VALID
_URL
_VIDEOS
_BASE
436 _PLAYLIST_PATH
= TwitchVideosBaseIE
._PLAYLIST
_PATH
+ 'archive,upload,highlight'
437 _PLAYLIST_TYPE
= 'all videos'
440 'url': 'https://www.twitch.tv/spamfish/videos/all',
445 'playlist_mincount': 869,
447 'url': 'https://m.twitch.tv/spamfish/videos/all',
448 'only_matching': True,
452 class TwitchUploadsIE(TwitchVideosBaseIE
):
453 IE_NAME
= 'twitch:videos:uploads'
454 _VALID_URL
= r
'%s/uploads' % TwitchVideosBaseIE
._VALID
_URL
_VIDEOS
_BASE
455 _PLAYLIST_PATH
= TwitchVideosBaseIE
._PLAYLIST
_PATH
+ 'upload'
456 _PLAYLIST_TYPE
= 'uploads'
459 'url': 'https://www.twitch.tv/spamfish/videos/uploads',
464 'playlist_mincount': 0,
466 'url': 'https://m.twitch.tv/spamfish/videos/uploads',
467 'only_matching': True,
471 class TwitchPastBroadcastsIE(TwitchVideosBaseIE
):
472 IE_NAME
= 'twitch:videos:past-broadcasts'
473 _VALID_URL
= r
'%s/past-broadcasts' % TwitchVideosBaseIE
._VALID
_URL
_VIDEOS
_BASE
474 _PLAYLIST_PATH
= TwitchVideosBaseIE
._PLAYLIST
_PATH
+ 'archive'
475 _PLAYLIST_TYPE
= 'past broadcasts'
478 'url': 'https://www.twitch.tv/spamfish/videos/past-broadcasts',
483 'playlist_mincount': 0,
485 'url': 'https://m.twitch.tv/spamfish/videos/past-broadcasts',
486 'only_matching': True,
490 class TwitchHighlightsIE(TwitchVideosBaseIE
):
491 IE_NAME
= 'twitch:videos:highlights'
492 _VALID_URL
= r
'%s/highlights' % TwitchVideosBaseIE
._VALID
_URL
_VIDEOS
_BASE
493 _PLAYLIST_PATH
= TwitchVideosBaseIE
._PLAYLIST
_PATH
+ 'highlight'
494 _PLAYLIST_TYPE
= 'highlights'
497 'url': 'https://www.twitch.tv/spamfish/videos/highlights',
502 'playlist_mincount': 805,
504 'url': 'https://m.twitch.tv/spamfish/videos/highlights',
505 'only_matching': True,
509 class TwitchStreamIE(TwitchBaseIE
):
510 IE_NAME
= 'twitch:stream'
511 _VALID_URL
= r
'''(?x)
514 (?:(?:www|go|m)\.)?twitch\.tv/|
515 player\.twitch\.tv/\?.*?\bchannel=
521 'url': 'http://www.twitch.tv/shroomztv',
524 'display_id': 'shroomztv',
526 'title': 're:^ShroomzTV [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
527 'description': 'H1Z1 - lonewolfing with ShroomzTV | A3 Battle Royale later - @ShroomzTV',
529 'timestamp': 1421928037,
530 'upload_date': '20150122',
531 'uploader': 'ShroomzTV',
532 'uploader_id': 'shroomztv',
537 'skip_download': True,
540 'url': 'http://www.twitch.tv/miracle_doto#profile-0',
541 'only_matching': True,
543 'url': 'https://player.twitch.tv/?channel=lotsofs',
544 'only_matching': True,
546 'url': 'https://go.twitch.tv/food',
547 'only_matching': True,
549 'url': 'https://m.twitch.tv/food',
550 'only_matching': True,
554 def suitable(cls
, url
):
556 if any(ie
.suitable(url
) for ie
in (
563 TwitchPastBroadcastsIE
,
566 else super(TwitchStreamIE
, cls
).suitable(url
))
568 def _real_extract(self
, url
):
569 channel_id
= self
._match
_id
(url
)
571 stream
= self
._call
_api
(
572 'kraken/streams/%s?stream_type=all' % channel_id
, channel_id
,
573 'Downloading stream JSON').get('stream')
576 raise ExtractorError('%s is offline' % channel_id
, expected
=True)
578 # Channel name may be typed if different case than the original channel name
579 # (e.g. http://www.twitch.tv/TWITCHPLAYSPOKEMON) that will lead to constructing
580 # an invalid m3u8 URL. Working around by use of original channel name from stream
581 # JSON and fallback to lowercase if it's not available.
582 channel_id
= stream
.get('channel', {}).get('name') or channel_id
.lower()
584 access_token
= self
._call
_api
(
585 'api/channels/%s/access_token' % channel_id
, channel_id
,
586 'Downloading channel access token')
589 'allow_source': 'true',
590 'allow_audio_only': 'true',
591 'allow_spectre': 'true',
592 'p': random
.randint(1000000, 10000000),
593 'player': 'twitchweb',
594 'segment_preference': '4',
595 'sig': access_token
['sig'].encode('utf-8'),
596 'token': access_token
['token'].encode('utf-8'),
598 formats
= self
._extract
_m
3u8_formats
(
599 '%s/api/channel/hls/%s.m3u8?%s'
600 % (self
._USHER
_BASE
, channel_id
, compat_urllib_parse_urlencode(query
)),
602 self
._prefer
_source
(formats
)
604 view_count
= stream
.get('viewers')
605 timestamp
= parse_iso8601(stream
.get('created_at'))
607 channel
= stream
['channel']
608 title
= self
._live
_title
(channel
.get('display_name') or channel
.get('name'))
609 description
= channel
.get('status')
612 for thumbnail_key
, thumbnail_url
in stream
['preview'].items():
613 m
= re
.search(r
'(?P<width>\d+)x(?P<height>\d+)\.jpg$', thumbnail_key
)
617 'url': thumbnail_url
,
618 'width': int(m
.group('width')),
619 'height': int(m
.group('height')),
623 'id': compat_str(stream
['_id']),
624 'display_id': channel_id
,
626 'description': description
,
627 'thumbnails': thumbnails
,
628 'uploader': channel
.get('display_name'),
629 'uploader_id': channel
.get('name'),
630 'timestamp': timestamp
,
631 'view_count': view_count
,
637 class TwitchClipsIE(TwitchBaseIE
):
638 IE_NAME
= 'twitch:clips'
639 _VALID_URL
= r
'https?://(?:clips\.twitch\.tv/(?:[^/]+/)*|(?:www\.)?twitch\.tv/[^/]+/clip/)(?P<id>[^/?#&]+)'
642 'url': 'https://clips.twitch.tv/FaintLightGullWholeWheat',
643 'md5': '761769e1eafce0ffebfb4089cb3847cd',
647 'title': 'EA Play 2016 Live from the Novo Theatre',
648 'thumbnail': r
're:^https?://.*\.jpg',
649 'timestamp': 1465767393,
650 'upload_date': '20160612',
652 'uploader': 'stereotype_',
653 'uploader_id': '43566419',
657 'url': 'https://clips.twitch.tv/rflegendary/UninterestedBeeDAESuppy',
658 'only_matching': True,
660 'url': 'https://www.twitch.tv/sergeynixon/clip/StormyThankfulSproutFutureMan',
661 'only_matching': True,
664 def _real_extract(self
, url
):
665 video_id
= self
._match
_id
(url
)
667 status
= self
._download
_json
(
668 'https://clips.twitch.tv/api/v2/clips/%s/status' % video_id
,
673 for option
in status
['quality_options']:
674 if not isinstance(option
, dict):
676 source
= url_or_none(option
.get('source'))
681 'format_id': option
.get('quality'),
682 'height': int_or_none(option
.get('quality')),
683 'fps': int_or_none(option
.get('frame_rate')),
686 self
._sort
_formats
(formats
)
692 clip
= self
._call
_api
(
693 'kraken/clips/%s' % video_id
, video_id
, fatal
=False, headers
={
694 'Accept': 'application/vnd.twitchtv.v5+json',
698 quality_key
= qualities(('tiny', 'small', 'medium'))
700 thumbnails_dict
= clip
.get('thumbnails')
701 if isinstance(thumbnails_dict
, dict):
702 for thumbnail_id
, thumbnail_url
in thumbnails_dict
.items():
705 'url': thumbnail_url
,
706 'preference': quality_key(thumbnail_id
),
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
),