import itertools
import re
import random
+import json
from .common import InfoExtractor
from ..compat import (
+ compat_kwargs,
compat_parse_qs,
compat_str,
- compat_urllib_parse,
+ compat_urllib_parse_urlencode,
compat_urllib_parse_urlparse,
- compat_urlparse,
)
from ..utils import (
- encode_dict,
+ clean_html,
ExtractorError,
int_or_none,
+ orderedSet,
parse_duration,
parse_iso8601,
- sanitized_Request,
+ qualities,
+ str_or_none,
+ try_get,
+ unified_timestamp,
+ update_url_query,
+ url_or_none,
+ urljoin,
)
class TwitchBaseIE(InfoExtractor):
- _VALID_URL_BASE = r'https?://(?:www\.)?twitch\.tv'
+ _VALID_URL_BASE = r'https?://(?:(?:www|go|m)\.)?twitch\.tv'
_API_BASE = 'https://api.twitch.tv'
- _USHER_BASE = 'http://usher.twitch.tv'
- _LOGIN_URL = 'http://www.twitch.tv/login'
+ _USHER_BASE = 'https://usher.ttvnw.net'
+ _LOGIN_FORM_URL = 'https://www.twitch.tv/login'
+ _LOGIN_POST_URL = 'https://passport.twitch.tv/login'
+ _CLIENT_ID = 'kimne78kx3ncx6brgo4mv6wki5h1ko'
_NETRC_MACHINE = 'twitch'
def _handle_error(self, response):
'%s returned error: %s - %s' % (self.IE_NAME, error, response.get('message')),
expected=True)
- def _download_json(self, url, video_id, note='Downloading JSON metadata'):
- headers = {
- 'Referer': 'http://api.twitch.tv/crossdomain/receiver.html?v=2',
- 'X-Requested-With': 'XMLHttpRequest',
- }
- for cookie in self._downloader.cookiejar:
- if cookie.name == 'api_token':
- headers['Twitch-Api-Token'] = cookie.value
- request = sanitized_Request(url, headers=headers)
- response = super(TwitchBaseIE, self)._download_json(request, video_id, note)
+ def _call_api(self, path, item_id, *args, **kwargs):
+ headers = kwargs.get('headers', {}).copy()
+ headers.update({
+ 'Accept': 'application/vnd.twitchtv.v5+json; charset=UTF-8',
+ 'Client-ID': self._CLIENT_ID,
+ })
+ kwargs.update({
+ 'headers': headers,
+ 'expected_status': (400, 410),
+ })
+ response = self._download_json(
+ '%s/%s' % (self._API_BASE, path), item_id,
+ *args, **compat_kwargs(kwargs))
self._handle_error(response)
return response
self._login()
def _login(self):
- (username, password) = self._get_login_info()
+ username, password = self._get_login_info()
if username is None:
return
- login_page, handle = self._download_webpage_handle(
- self._LOGIN_URL, None, 'Downloading login page')
+ def fail(message):
+ raise ExtractorError(
+ 'Unable to login. Twitch said: %s' % message, expected=True)
- login_form = self._hidden_inputs(login_page)
+ def login_step(page, urlh, note, data):
+ form = self._hidden_inputs(page)
+ form.update(data)
- login_form.update({
- 'username': username,
- 'password': password,
- })
+ page_url = urlh.geturl()
+ post_url = self._search_regex(
+ r'<form[^>]+action=(["\'])(?P<url>.+?)\1', page,
+ 'post url', default=self._LOGIN_POST_URL, group='url')
+ post_url = urljoin(page_url, post_url)
- redirect_url = handle.geturl()
+ headers = {
+ 'Referer': page_url,
+ 'Origin': page_url,
+ 'Content-Type': 'text/plain;charset=UTF-8',
+ }
- post_url = self._search_regex(
- r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
- 'post url', default=redirect_url, group='url')
+ response = self._download_json(
+ post_url, None, note, data=json.dumps(form).encode(),
+ headers=headers, expected_status=400)
+ error = response.get('error_description') or response.get('error_code')
+ if error:
+ fail(error)
+
+ if 'Authenticated successfully' in response.get('message', ''):
+ return None, None
+
+ redirect_url = urljoin(
+ post_url,
+ response.get('redirect') or response['redirect_path'])
+ return self._download_webpage_handle(
+ redirect_url, None, 'Downloading login redirect page',
+ headers=headers)
- if not post_url.startswith('http'):
- post_url = compat_urlparse.urljoin(redirect_url, post_url)
+ login_page, handle = self._download_webpage_handle(
+ self._LOGIN_FORM_URL, None, 'Downloading login page')
- request = sanitized_Request(
- post_url, compat_urllib_parse.urlencode(encode_dict(login_form)).encode('utf-8'))
- request.add_header('Referer', redirect_url)
- response = self._download_webpage(
- request, None, 'Logging in as %s' % username)
+ # Some TOR nodes and public proxies are blocked completely
+ if 'blacklist_message' in login_page:
+ fail(clean_html(login_page))
- error_message = self._search_regex(
- r'<div[^>]+class="subwindow_notice"[^>]*>([^<]+)</div>',
- response, 'error message', default=None)
- if error_message:
- raise ExtractorError(
- 'Unable to login. Twitch said: %s' % error_message, expected=True)
+ redirect_page, handle = login_step(
+ login_page, handle, 'Logging in', {
+ 'username': username,
+ 'password': password,
+ 'client_id': self._CLIENT_ID,
+ })
+
+ # Successful login
+ if not redirect_page:
+ return
- if '>Reset your password<' in response:
- self.report_warning('Twitch asks you to reset your password, go to https://secure.twitch.tv/reset/submit')
+ if re.search(r'(?i)<form[^>]+id="two-factor-submit"', redirect_page) is not None:
+ # TODO: Add mechanism to request an SMS or phone call
+ tfa_token = self._get_tfa_info('two-factor authentication token')
+ login_step(redirect_page, handle, 'Submitting TFA token', {
+ 'authy_token': tfa_token,
+ 'remember_2fa': 'true',
+ })
def _prefer_source(self, formats):
try:
source = next(f for f in formats if f['format_id'] == 'Source')
- source['preference'] = 10
+ source['quality'] = 10
except StopIteration:
- pass # No Source stream present
+ for f in formats:
+ if '/chunked/' in f['url']:
+ f.update({
+ 'quality': 10,
+ 'format_note': 'Source',
+ })
self._sort_formats(formats)
class TwitchItemBaseIE(TwitchBaseIE):
def _download_info(self, item, item_id):
- return self._extract_info(self._download_json(
- '%s/kraken/videos/%s%s' % (self._API_BASE, item, item_id), item_id,
+ return self._extract_info(self._call_api(
+ 'kraken/videos/%s%s' % (item, item_id), item_id,
'Downloading %s info JSON' % self._ITEM_TYPE))
def _extract_media(self, item_id):
info = self._download_info(self._ITEM_SHORTCUT, item_id)
- response = self._download_json(
- '%s/api/videos/%s%s' % (self._API_BASE, self._ITEM_SHORTCUT, item_id), item_id,
+ response = self._call_api(
+ 'api/videos/%s%s' % (self._ITEM_SHORTCUT, item_id), item_id,
'Downloading %s playlist JSON' % self._ITEM_TYPE)
entries = []
chunks = response['chunks']
return self.playlist_result(entries, info['id'], info['title'])
def _extract_info(self, info):
+ status = info.get('status')
+ if status == 'recording':
+ is_live = True
+ elif status == 'recorded':
+ is_live = False
+ else:
+ is_live = None
+ _QUALITIES = ('small', 'medium', 'large')
+ quality_key = qualities(_QUALITIES)
+ thumbnails = []
+ preview = info.get('preview')
+ if isinstance(preview, dict):
+ for thumbnail_id, thumbnail_url in preview.items():
+ thumbnail_url = url_or_none(thumbnail_url)
+ if not thumbnail_url:
+ continue
+ if thumbnail_id not in _QUALITIES:
+ continue
+ thumbnails.append({
+ 'url': thumbnail_url,
+ 'preference': quality_key(thumbnail_id),
+ })
return {
'id': info['_id'],
'title': info.get('title') or 'Untitled Broadcast',
'description': info.get('description'),
'duration': int_or_none(info.get('length')),
- 'thumbnail': info.get('preview'),
+ 'thumbnails': thumbnails,
'uploader': info.get('channel', {}).get('display_name'),
'uploader_id': info.get('channel', {}).get('name'),
'timestamp': parse_iso8601(info.get('recorded_at')),
'view_count': int_or_none(info.get('views')),
+ 'is_live': is_live,
}
def _real_extract(self, url):
'title': 'Worlds Semifinals - Star Horn Royal Club vs. OMG',
},
'playlist_mincount': 12,
+ 'skip': 'HTTP Error 404: Not Found',
}
'title': 'ACRL Off Season - Sports Cars @ Nordschleife',
},
'playlist_mincount': 3,
+ 'skip': 'HTTP Error 404: Not Found',
}, {
'url': 'http://www.twitch.tv/tsm_theoddone/c/2349361',
'only_matching': True,
class TwitchVodIE(TwitchItemBaseIE):
IE_NAME = 'twitch:vod'
- _VALID_URL = r'%s/[^/]+/v/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
+ _VALID_URL = r'''(?x)
+ https?://
+ (?:
+ (?:(?:www|go|m)\.)?twitch\.tv/(?:[^/]+/v(?:ideo)?|videos)/|
+ player\.twitch\.tv/\?.*?\bvideo=v?
+ )
+ (?P<id>\d+)
+ '''
_ITEM_TYPE = 'vod'
_ITEM_SHORTCUT = 'v'
'id': 'v6528877',
'ext': 'mp4',
'title': 'LCK Summer Split - Week 6 Day 1',
- 'thumbnail': 're:^https?://.*\.jpg$',
+ 'thumbnail': r're:^https?://.*\.jpg$',
'duration': 17208,
'timestamp': 1435131709,
'upload_date': '20150624',
'id': 'v11230755',
'ext': 'mp4',
'title': 'Untitled Broadcast',
- 'thumbnail': 're:^https?://.*\.jpg$',
+ 'thumbnail': r're:^https?://.*\.jpg$',
'duration': 1638,
'timestamp': 1439746708,
'upload_date': '20150816',
# m3u8 download
'skip_download': True,
},
+ 'skip': 'HTTP Error 404: Not Found',
+ }, {
+ 'url': 'http://player.twitch.tv/?t=5m10s&video=v6528877',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://www.twitch.tv/videos/6528877',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://m.twitch.tv/beagsandjam/v/247478721',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://www.twitch.tv/northernlion/video/291940395',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://player.twitch.tv/?video=480452374',
+ 'only_matching': True,
}]
def _real_extract(self, url):
item_id = self._match_id(url)
info = self._download_info(self._ITEM_SHORTCUT, item_id)
- access_token = self._download_json(
- '%s/api/vods/%s/access_token' % (self._API_BASE, item_id), item_id,
+ access_token = self._call_api(
+ 'api/vods/%s/access_token' % item_id, item_id,
'Downloading %s access token' % self._ITEM_TYPE)
formats = self._extract_m3u8_formats(
- '%s/vod/%s?%s' % (
+ '%s/vod/%s.m3u8?%s' % (
self._USHER_BASE, item_id,
- compat_urllib_parse.urlencode({
+ compat_urllib_parse_urlencode({
'allow_source': 'true',
+ 'allow_audio_only': 'true',
'allow_spectre': 'true',
'player': 'twitchweb',
+ 'playlist_include_framerate': 'true',
'nauth': access_token['token'],
'nauthsig': access_token['sig'],
})),
- item_id, 'mp4')
+ item_id, 'mp4', entry_protocol='m3u8_native')
self._prefer_source(formats)
info['formats'] = formats
if 't' in query:
info['start_time'] = parse_duration(query['t'][0])
+ if info.get('timestamp') is not None:
+ info['subtitles'] = {
+ 'rechat': [{
+ 'url': update_url_query(
+ 'https://api.twitch.tv/v5/videos/%s/comments' % item_id, {
+ 'client_id': self._CLIENT_ID,
+ }),
+ 'ext': 'json',
+ }],
+ }
+
return info
class TwitchPlaylistBaseIE(TwitchBaseIE):
- _PLAYLIST_URL = '%s/kraken/channels/%%s/videos/?offset=%%d&limit=%%d' % TwitchBaseIE._API_BASE
+ _PLAYLIST_PATH = 'kraken/channels/%s/videos/?offset=%d&limit=%d'
_PAGE_LIMIT = 100
def _extract_playlist(self, channel_id):
- info = self._download_json(
- '%s/kraken/channels/%s' % (self._API_BASE, channel_id),
+ info = self._call_api(
+ 'kraken/channels/%s' % channel_id,
channel_id, 'Downloading channel info JSON')
channel_name = info.get('display_name') or info.get('name')
entries = []
offset = 0
limit = self._PAGE_LIMIT
+ broken_paging_detected = False
+ counter_override = None
for counter in itertools.count(1):
- response = self._download_json(
- self._PLAYLIST_URL % (channel_id, offset, limit),
- channel_id, 'Downloading %s videos JSON page %d' % (self._PLAYLIST_TYPE, counter))
+ response = self._call_api(
+ self._PLAYLIST_PATH % (channel_id, offset, limit),
+ channel_id,
+ 'Downloading %s JSON page %s'
+ % (self._PLAYLIST_TYPE, counter_override or counter))
page_entries = self._extract_playlist_page(response)
if not page_entries:
break
+ total = int_or_none(response.get('_total'))
+ # Since the beginning of March 2016 twitch's paging mechanism
+ # is completely broken on the twitch side. It simply ignores
+ # a limit and returns the whole offset number of videos.
+ # Working around by just requesting all videos at once.
+ # Upd: pagination bug was fixed by twitch on 15.03.2016.
+ if not broken_paging_detected and total and len(page_entries) > limit:
+ self.report_warning(
+ 'Twitch pagination is broken on twitch side, requesting all videos at once',
+ channel_id)
+ broken_paging_detected = True
+ offset = total
+ counter_override = '(all at once)'
+ continue
entries.extend(page_entries)
+ if broken_paging_detected or total and len(page_entries) >= total:
+ break
offset += limit
return self.playlist_result(
- [self.url_result(entry) for entry in set(entries)],
+ [self._make_url_result(entry) for entry in orderedSet(entries)],
channel_id, channel_name)
+ def _make_url_result(self, url):
+ try:
+ video_id = 'v%s' % TwitchVodIE._match_id(url)
+ return self.url_result(url, TwitchVodIE.ie_key(), video_id=video_id)
+ except AssertionError:
+ return self.url_result(url)
+
def _extract_playlist_page(self, response):
videos = response.get('videos')
return [video['url'] for video in videos] if videos else []
_VALID_URL = r'%s/(?P<id>[^/]+)/profile/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
_PLAYLIST_TYPE = 'profile'
- _TEST = {
+ _TESTS = [{
'url': 'http://www.twitch.tv/vanillatv/profile',
'info_dict': {
'id': 'vanillatv',
'title': 'VanillaTV',
},
'playlist_mincount': 412,
- }
+ }, {
+ 'url': 'http://m.twitch.tv/vanillatv/profile',
+ 'only_matching': True,
+ }]
-class TwitchPastBroadcastsIE(TwitchPlaylistBaseIE):
- IE_NAME = 'twitch:past_broadcasts'
- _VALID_URL = r'%s/(?P<id>[^/]+)/profile/past_broadcasts/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
- _PLAYLIST_URL = TwitchPlaylistBaseIE._PLAYLIST_URL + '&broadcasts=true'
- _PLAYLIST_TYPE = 'past broadcasts'
+class TwitchVideosBaseIE(TwitchPlaylistBaseIE):
+ _VALID_URL_VIDEOS_BASE = r'%s/(?P<id>[^/]+)/videos' % TwitchBaseIE._VALID_URL_BASE
+ _PLAYLIST_PATH = TwitchPlaylistBaseIE._PLAYLIST_PATH + '&broadcast_type='
- _TEST = {
- 'url': 'http://www.twitch.tv/spamfish/profile/past_broadcasts',
+
+class TwitchAllVideosIE(TwitchVideosBaseIE):
+ IE_NAME = 'twitch:videos:all'
+ _VALID_URL = r'%s/all' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
+ _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'archive,upload,highlight'
+ _PLAYLIST_TYPE = 'all videos'
+
+ _TESTS = [{
+ 'url': 'https://www.twitch.tv/spamfish/videos/all',
'info_dict': {
'id': 'spamfish',
'title': 'Spamfish',
},
- 'playlist_mincount': 54,
- }
+ 'playlist_mincount': 869,
+ }, {
+ 'url': 'https://m.twitch.tv/spamfish/videos/all',
+ 'only_matching': True,
+ }]
-class TwitchBookmarksIE(TwitchPlaylistBaseIE):
- IE_NAME = 'twitch:bookmarks'
- _VALID_URL = r'%s/(?P<id>[^/]+)/profile/bookmarks/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
- _PLAYLIST_URL = '%s/api/bookmark/?user=%%s&offset=%%d&limit=%%d' % TwitchBaseIE._API_BASE
- _PLAYLIST_TYPE = 'bookmarks'
+class TwitchUploadsIE(TwitchVideosBaseIE):
+ IE_NAME = 'twitch:videos:uploads'
+ _VALID_URL = r'%s/uploads' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
+ _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'upload'
+ _PLAYLIST_TYPE = 'uploads'
- _TEST = {
- 'url': 'http://www.twitch.tv/ognos/profile/bookmarks',
+ _TESTS = [{
+ 'url': 'https://www.twitch.tv/spamfish/videos/uploads',
'info_dict': {
- 'id': 'ognos',
- 'title': 'Ognos',
+ 'id': 'spamfish',
+ 'title': 'Spamfish',
},
- 'playlist_mincount': 3,
- }
+ 'playlist_mincount': 0,
+ }, {
+ 'url': 'https://m.twitch.tv/spamfish/videos/uploads',
+ 'only_matching': True,
+ }]
- def _extract_playlist_page(self, response):
- entries = []
- for bookmark in response.get('bookmarks', []):
- video = bookmark.get('video')
- if not video:
- continue
- entries.append(video['url'])
- return entries
+
+class TwitchPastBroadcastsIE(TwitchVideosBaseIE):
+ IE_NAME = 'twitch:videos:past-broadcasts'
+ _VALID_URL = r'%s/past-broadcasts' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
+ _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'archive'
+ _PLAYLIST_TYPE = 'past broadcasts'
+
+ _TESTS = [{
+ 'url': 'https://www.twitch.tv/spamfish/videos/past-broadcasts',
+ 'info_dict': {
+ 'id': 'spamfish',
+ 'title': 'Spamfish',
+ },
+ 'playlist_mincount': 0,
+ }, {
+ 'url': 'https://m.twitch.tv/spamfish/videos/past-broadcasts',
+ 'only_matching': True,
+ }]
+
+
+class TwitchHighlightsIE(TwitchVideosBaseIE):
+ IE_NAME = 'twitch:videos:highlights'
+ _VALID_URL = r'%s/highlights' % TwitchVideosBaseIE._VALID_URL_VIDEOS_BASE
+ _PLAYLIST_PATH = TwitchVideosBaseIE._PLAYLIST_PATH + 'highlight'
+ _PLAYLIST_TYPE = 'highlights'
+
+ _TESTS = [{
+ 'url': 'https://www.twitch.tv/spamfish/videos/highlights',
+ 'info_dict': {
+ 'id': 'spamfish',
+ 'title': 'Spamfish',
+ },
+ 'playlist_mincount': 805,
+ }, {
+ 'url': 'https://m.twitch.tv/spamfish/videos/highlights',
+ 'only_matching': True,
+ }]
class TwitchStreamIE(TwitchBaseIE):
IE_NAME = 'twitch:stream'
- _VALID_URL = r'%s/(?P<id>[^/#?]+)/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
+ _VALID_URL = r'''(?x)
+ https?://
+ (?:
+ (?:(?:www|go|m)\.)?twitch\.tv/|
+ player\.twitch\.tv/\?.*?\bchannel=
+ )
+ (?P<id>[^/#?]+)
+ '''
_TESTS = [{
'url': 'http://www.twitch.tv/shroomztv',
}, {
'url': 'http://www.twitch.tv/miracle_doto#profile-0',
'only_matching': True,
+ }, {
+ 'url': 'https://player.twitch.tv/?channel=lotsofs',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://go.twitch.tv/food',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://m.twitch.tv/food',
+ 'only_matching': True,
}]
+ @classmethod
+ def suitable(cls, url):
+ return (False
+ if any(ie.suitable(url) for ie in (
+ TwitchVideoIE,
+ TwitchChapterIE,
+ TwitchVodIE,
+ TwitchProfileIE,
+ TwitchAllVideosIE,
+ TwitchUploadsIE,
+ TwitchPastBroadcastsIE,
+ TwitchHighlightsIE,
+ TwitchClipsIE))
+ else super(TwitchStreamIE, cls).suitable(url))
+
def _real_extract(self, url):
- channel_id = self._match_id(url)
+ channel_name = self._match_id(url)
+
+ access_token = self._call_api(
+ 'api/channels/%s/access_token' % channel_name, channel_name,
+ 'Downloading access token JSON')
- stream = self._download_json(
- '%s/kraken/streams/%s' % (self._API_BASE, channel_id), channel_id,
- 'Downloading stream JSON').get('stream')
+ token = access_token['token']
+ channel_id = compat_str(self._parse_json(
+ token, channel_name)['channel_id'])
+
+ stream = self._call_api(
+ 'kraken/streams/%s?stream_type=all' % channel_id,
+ channel_id, 'Downloading stream JSON').get('stream')
- # Fallback on profile extraction if stream is offline
if not stream:
- return self.url_result(
- 'http://www.twitch.tv/%s/profile' % channel_id,
- 'TwitchProfile', channel_id)
+ raise ExtractorError('%s is offline' % channel_id, expected=True)
# Channel name may be typed if different case than the original channel name
# (e.g. http://www.twitch.tv/TWITCHPLAYSPOKEMON) that will lead to constructing
# an invalid m3u8 URL. Working around by use of original channel name from stream
# JSON and fallback to lowercase if it's not available.
- channel_id = stream.get('channel', {}).get('name') or channel_id.lower()
-
- access_token = self._download_json(
- '%s/api/channels/%s/access_token' % (self._API_BASE, channel_id), channel_id,
- 'Downloading channel access token')
+ channel_name = try_get(
+ stream, lambda x: x['channel']['name'],
+ compat_str) or channel_name.lower()
query = {
'allow_source': 'true',
+ 'allow_audio_only': 'true',
+ 'allow_spectre': 'true',
'p': random.randint(1000000, 10000000),
'player': 'twitchweb',
+ 'playlist_include_framerate': 'true',
'segment_preference': '4',
'sig': access_token['sig'].encode('utf-8'),
- 'token': access_token['token'].encode('utf-8'),
+ 'token': token.encode('utf-8'),
}
formats = self._extract_m3u8_formats(
'%s/api/channel/hls/%s.m3u8?%s'
- % (self._USHER_BASE, channel_id, compat_urllib_parse.urlencode(query)),
+ % (self._USHER_BASE, channel_name, compat_urllib_parse_urlencode(query)),
channel_id, 'mp4')
self._prefer_source(formats)
})
return {
- 'id': compat_str(stream['_id']),
- 'display_id': channel_id,
+ 'id': str_or_none(stream.get('_id')) or channel_id,
+ 'display_id': channel_name,
'title': title,
'description': description,
'thumbnails': thumbnails,
'formats': formats,
'is_live': True,
}
+
+
+class TwitchClipsIE(TwitchBaseIE):
+ IE_NAME = 'twitch:clips'
+ _VALID_URL = r'''(?x)
+ https?://
+ (?:
+ clips\.twitch\.tv/(?:embed\?.*?\bclip=|(?:[^/]+/)*)|
+ (?:(?:www|go|m)\.)?twitch\.tv/[^/]+/clip/
+ )
+ (?P<id>[^/?#&]+)
+ '''
+
+ _TESTS = [{
+ 'url': 'https://clips.twitch.tv/FaintLightGullWholeWheat',
+ 'md5': '761769e1eafce0ffebfb4089cb3847cd',
+ 'info_dict': {
+ 'id': '42850523',
+ 'ext': 'mp4',
+ 'title': 'EA Play 2016 Live from the Novo Theatre',
+ 'thumbnail': r're:^https?://.*\.jpg',
+ 'timestamp': 1465767393,
+ 'upload_date': '20160612',
+ 'creator': 'EA',
+ 'uploader': 'stereotype_',
+ 'uploader_id': '43566419',
+ },
+ }, {
+ # multiple formats
+ 'url': 'https://clips.twitch.tv/rflegendary/UninterestedBeeDAESuppy',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://www.twitch.tv/sergeynixon/clip/StormyThankfulSproutFutureMan',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://clips.twitch.tv/embed?clip=InquisitiveBreakableYogurtJebaited',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://m.twitch.tv/rossbroadcast/clip/ConfidentBraveHumanChefFrank',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://go.twitch.tv/rossbroadcast/clip/ConfidentBraveHumanChefFrank',
+ 'only_matching': True,
+ }]
+
+ def _real_extract(self, url):
+ video_id = self._match_id(url)
+
+ clip = self._download_json(
+ 'https://gql.twitch.tv/gql', video_id, data=json.dumps({
+ 'query': '''{
+ clip(slug: "%s") {
+ broadcaster {
+ displayName
+ }
+ createdAt
+ curator {
+ displayName
+ id
+ }
+ durationSeconds
+ id
+ tiny: thumbnailURL(width: 86, height: 45)
+ small: thumbnailURL(width: 260, height: 147)
+ medium: thumbnailURL(width: 480, height: 272)
+ title
+ videoQualities {
+ frameRate
+ quality
+ sourceURL
+ }
+ viewCount
+ }
+}''' % video_id,
+ }).encode(), headers={
+ 'Client-ID': self._CLIENT_ID,
+ })['data']['clip']
+
+ if not clip:
+ raise ExtractorError(
+ 'This clip is no longer available', expected=True)
+
+ formats = []
+ for option in clip.get('videoQualities', []):
+ if not isinstance(option, dict):
+ continue
+ source = url_or_none(option.get('sourceURL'))
+ if not source:
+ continue
+ formats.append({
+ 'url': source,
+ 'format_id': option.get('quality'),
+ 'height': int_or_none(option.get('quality')),
+ 'fps': int_or_none(option.get('frameRate')),
+ })
+ self._sort_formats(formats)
+
+ thumbnails = []
+ for thumbnail_id in ('tiny', 'small', 'medium'):
+ thumbnail_url = clip.get(thumbnail_id)
+ if not thumbnail_url:
+ continue
+ thumb = {
+ 'id': thumbnail_id,
+ 'url': thumbnail_url,
+ }
+ mobj = re.search(r'-(\d+)x(\d+)\.', thumbnail_url)
+ if mobj:
+ thumb.update({
+ 'height': int(mobj.group(2)),
+ 'width': int(mobj.group(1)),
+ })
+ thumbnails.append(thumb)
+
+ return {
+ 'id': clip.get('id') or video_id,
+ 'title': clip.get('title') or video_id,
+ 'formats': formats,
+ 'duration': int_or_none(clip.get('durationSeconds')),
+ 'views': int_or_none(clip.get('viewCount')),
+ 'timestamp': unified_timestamp(clip.get('createdAt')),
+ 'thumbnails': thumbnails,
+ 'creator': try_get(clip, lambda x: x['broadcaster']['displayName'], compat_str),
+ 'uploader': try_get(clip, lambda x: x['curator']['displayName'], compat_str),
+ 'uploader_id': try_get(clip, lambda x: x['curator']['id'], compat_str),
+ }
Rapsys Git / youtubedl / blobdiff