+
+ return info
+
+
+class TwitchPlaylistBaseIE(TwitchBaseIE):
+ _PLAYLIST_PATH = 'kraken/channels/%s/videos/?offset=%d&limit=%d'
+ _PAGE_LIMIT = 100
+
+ def _extract_playlist(self, 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._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._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 []
+
+ def _real_extract(self, url):
+ return self._extract_playlist(self._match_id(url))
+
+
+class TwitchProfileIE(TwitchPlaylistBaseIE):
+ IE_NAME = 'twitch:profile'
+ _VALID_URL = r'%s/(?P<id>[^/]+)/profile/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
+ _PLAYLIST_TYPE = 'profile'
+
+ _TEST = {
+ 'url': 'http://www.twitch.tv/vanillatv/profile',
+ 'info_dict': {
+ 'id': 'vanillatv',
+ 'title': 'VanillaTV',
+ },
+ 'playlist_mincount': 412,
+ }
+
+
+class TwitchVideosBaseIE(TwitchPlaylistBaseIE):
+ _VALID_URL_VIDEOS_BASE = r'%s/(?P<id>[^/]+)/videos' % TwitchBaseIE._VALID_URL_BASE
+ _PLAYLIST_PATH = TwitchPlaylistBaseIE._PLAYLIST_PATH + '&broadcast_type='
+
+
+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'
+
+ _TEST = {
+ 'url': 'https://www.twitch.tv/spamfish/videos/all',
+ 'info_dict': {
+ 'id': 'spamfish',
+ 'title': 'Spamfish',
+ },
+ 'playlist_mincount': 869,
+ }
+
+
+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': 'https://www.twitch.tv/spamfish/videos/uploads',
+ 'info_dict': {
+ 'id': 'spamfish',
+ 'title': 'Spamfish',
+ },
+ 'playlist_mincount': 0,
+ }
+
+
+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'
+
+ _TEST = {
+ 'url': 'https://www.twitch.tv/spamfish/videos/past-broadcasts',
+ 'info_dict': {
+ 'id': 'spamfish',
+ 'title': 'Spamfish',
+ },
+ 'playlist_mincount': 0,
+ }
+
+
+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'
+
+ _TEST = {
+ 'url': 'https://www.twitch.tv/spamfish/videos/highlights',
+ 'info_dict': {
+ 'id': 'spamfish',
+ 'title': 'Spamfish',
+ },
+ 'playlist_mincount': 805,
+ }
+
+
+class TwitchStreamIE(TwitchBaseIE):
+ IE_NAME = 'twitch:stream'
+ _VALID_URL = r'''(?x)
+ https?://
+ (?:
+ (?:(?:www|go)\.)?twitch\.tv/|
+ player\.twitch\.tv/\?.*?\bchannel=
+ )
+ (?P<id>[^/#?]+)
+ '''
+
+ _TESTS = [{
+ 'url': 'http://www.twitch.tv/shroomztv',
+ 'info_dict': {
+ 'id': '12772022048',
+ 'display_id': 'shroomztv',
+ 'ext': 'mp4',
+ 'title': 're:^ShroomzTV [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
+ 'description': 'H1Z1 - lonewolfing with ShroomzTV | A3 Battle Royale later - @ShroomzTV',
+ 'is_live': True,
+ 'timestamp': 1421928037,
+ 'upload_date': '20150122',
+ 'uploader': 'ShroomzTV',
+ 'uploader_id': 'shroomztv',
+ 'view_count': int,
+ },
+ 'params': {
+ # m3u8 download
+ 'skip_download': True,
+ },
+ }, {
+ '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,
+ }]
+
+ @classmethod
+ def suitable(cls, url):
+ return (False
+ if any(ie.suitable(url) for ie in (
+ TwitchVideoIE,
+ TwitchChapterIE,
+ TwitchVodIE,
+ TwitchProfileIE,
+ TwitchAllVideosIE,
+ TwitchUploadsIE,
+ TwitchPastBroadcastsIE,
+ TwitchHighlightsIE))
+ else super(TwitchStreamIE, cls).suitable(url))
+
+ def _real_extract(self, url):
+ channel_id = self._match_id(url)
+
+ stream = self._call_api(
+ 'kraken/streams/%s?stream_type=all' % channel_id, channel_id,
+ 'Downloading stream JSON').get('stream')
+
+ if not stream:
+ 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._call_api(
+ 'api/channels/%s/access_token' % channel_id, channel_id,
+ 'Downloading channel access token')
+
+ query = {
+ 'allow_source': 'true',
+ 'allow_audio_only': 'true',
+ 'allow_spectre': 'true',
+ 'p': random.randint(1000000, 10000000),
+ 'player': 'twitchweb',
+ 'segment_preference': '4',
+ 'sig': access_token['sig'].encode('utf-8'),
+ 'token': access_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)),
+ channel_id, 'mp4')
+ self._prefer_source(formats)
+
+ view_count = stream.get('viewers')
+ timestamp = parse_iso8601(stream.get('created_at'))
+
+ channel = stream['channel']
+ title = self._live_title(channel.get('display_name') or channel.get('name'))
+ description = channel.get('status')
+
+ thumbnails = []
+ for thumbnail_key, thumbnail_url in stream['preview'].items():
+ m = re.search(r'(?P<width>\d+)x(?P<height>\d+)\.jpg$', thumbnail_key)
+ if not m:
+ continue
+ thumbnails.append({
+ 'url': thumbnail_url,
+ 'width': int(m.group('width')),
+ 'height': int(m.group('height')),
+ })
+
+ return {
+ 'id': compat_str(stream['_id']),
+ 'display_id': channel_id,
+ 'title': title,
+ 'description': description,
+ 'thumbnails': thumbnails,
+ 'uploader': channel.get('display_name'),
+ 'uploader_id': channel.get('name'),
+ 'timestamp': timestamp,
+ 'view_count': view_count,
+ 'formats': formats,
+ 'is_live': True,
+ }
+
+
+class TwitchClipsIE(InfoExtractor):
+ IE_NAME = 'twitch:clips'
+ _VALID_URL = r'https?://clips\.twitch\.tv/(?:[^/]+/)*(?P<id>[^/?#&]+)'
+
+ _TESTS = [{
+ 'url': 'https://clips.twitch.tv/ea/AggressiveCobraPoooound',
+ 'md5': '761769e1eafce0ffebfb4089cb3847cd',
+ 'info_dict': {
+ 'id': 'AggressiveCobraPoooound',
+ 'ext': 'mp4',
+ 'title': 'EA Play 2016 Live from the Novo Theatre',
+ 'thumbnail': r're:^https?://.*\.jpg',
+ 'creator': 'EA',
+ 'uploader': 'stereotype_',
+ 'uploader_id': 'stereotype_',
+ },
+ }, {
+ # multiple formats
+ 'url': 'https://clips.twitch.tv/rflegendary/UninterestedBeeDAESuppy',
+ 'only_matching': True,
+ }]
+
+ def _real_extract(self, url):
+ video_id = self._match_id(url)
+
+ webpage = self._download_webpage(url, video_id)
+
+ clip = self._parse_json(
+ self._search_regex(
+ r'(?s)clipInfo\s*=\s*({.+?});', webpage, 'clip info'),
+ video_id, transform_source=js_to_json)
+
+ title = clip.get('title') or clip.get('channel_title') or self._og_search_title(webpage)
+
+ formats = [{
+ 'url': option['source'],
+ 'format_id': option.get('quality'),
+ 'height': int_or_none(option.get('quality')),
+ } for option in clip.get('quality_options', []) if option.get('source')]
+
+ if not formats:
+ formats = [{
+ 'url': clip['clip_video_url'],
+ }]
+
+ self._sort_formats(formats)
+
+ return {
+ 'id': video_id,
+ 'title': title,
+ 'thumbnail': self._og_search_thumbnail(webpage),
+ 'creator': clip.get('broadcaster_display_name') or clip.get('broadcaster_login'),
+ 'uploader': clip.get('curator_login'),
+ 'uploader_id': clip.get('curator_display_name'),
+ 'formats': formats,
+ }
Rapsys Git / youtubedl / blobdiff