2 from __future__ import unicode_literals
8 from .common import InfoExtractor
12 compat_urllib_request,
20 class TwitchBaseIE(InfoExtractor):
21 _VALID_URL_BASE = r'https?://(?:www\.)?twitch\.tv'
23 _API_BASE = 'https://api.twitch.tv'
24 _USHER_BASE = 'http://usher.twitch.tv'
25 _LOGIN_URL = 'https://secure.twitch.tv/login'
26 _LOGIN_POST_URL = 'https://passport.twitch.tv/authorize'
27 _NETRC_MACHINE = 'twitch'
29 def _handle_error(self, response):
30 if not isinstance(response, dict):
32 error = response.get('error')
35 '%s returned error: %s - %s' % (self.IE_NAME, error, response.get('message')),
38 def _download_json(self, url, video_id, note='Downloading JSON metadata'):
40 'Referer': 'http://api.twitch.tv/crossdomain/receiver.html?v=2',
41 'X-Requested-With': 'XMLHttpRequest',
43 for cookie in self._downloader.cookiejar:
44 if cookie.name == 'api_token':
45 headers['Twitch-Api-Token'] = cookie.value
46 request = compat_urllib_request.Request(url, headers=headers)
47 response = super(TwitchBaseIE, self)._download_json(request, video_id, note)
48 self._handle_error(response)
51 def _real_initialize(self):
55 (username, password) = self._get_login_info()
59 login_page = self._download_webpage(
60 self._LOGIN_URL, None, 'Downloading login page')
62 login_form = dict(re.findall(
63 r'<input\s+type="hidden"\s+name="([^"]+)"\s+(?:id="[^"]+"\s+)?value="([^"]*)"',
71 request = compat_urllib_request.Request(
72 self._LOGIN_POST_URL, compat_urllib_parse.urlencode(login_form).encode('utf-8'))
73 request.add_header('Referer', self._LOGIN_URL)
74 response = self._download_webpage(
75 request, None, 'Logging in as %s' % username)
77 error_message = self._search_regex(
78 r'<div[^>]+class="subwindow_notice"[^>]*>([^<]+)</div>',
79 response, 'error message', default=None)
82 'Unable to login. Twitch said: %s' % error_message, expected=True)
84 def _prefer_source(self, formats):
86 source = next(f for f in formats if f['format_id'] == 'Source')
87 source['preference'] = 10
89 pass # No Source stream present
90 self._sort_formats(formats)
93 class TwitchItemBaseIE(TwitchBaseIE):
94 def _download_info(self, item, item_id):
95 return self._extract_info(self._download_json(
96 '%s/kraken/videos/%s%s' % (self._API_BASE, item, item_id), item_id,
97 'Downloading %s info JSON' % self._ITEM_TYPE))
99 def _extract_media(self, item_id):
100 info = self._download_info(self._ITEM_SHORTCUT, item_id)
101 response = self._download_json(
102 '%s/api/videos/%s%s' % (self._API_BASE, self._ITEM_SHORTCUT, item_id), item_id,
103 'Downloading %s playlist JSON' % self._ITEM_TYPE)
105 chunks = response['chunks']
106 qualities = list(chunks.keys())
107 for num, fragment in enumerate(zip(*chunks.values()), start=1):
109 for fmt_num, fragment_fmt in enumerate(fragment):
110 format_id = qualities[fmt_num]
112 'url': fragment_fmt['url'],
113 'format_id': format_id,
114 'quality': 1 if format_id == 'live' else 0,
116 m = re.search(r'^(?P<height>\d+)[Pp]', format_id)
118 fmt['height'] = int(m.group('height'))
120 self._sort_formats(formats)
122 entry['id'] = '%s_%d' % (entry['id'], num)
123 entry['title'] = '%s part %d' % (entry['title'], num)
124 entry['formats'] = formats
125 entries.append(entry)
126 return self.playlist_result(entries, info['id'], info['title'])
128 def _extract_info(self, info):
131 'title': info['title'],
132 'description': info['description'],
133 'duration': info['length'],
134 'thumbnail': info['preview'],
135 'uploader': info['channel']['display_name'],
136 'uploader_id': info['channel']['name'],
137 'timestamp': parse_iso8601(info['recorded_at']),
138 'view_count': info['views'],
141 def _real_extract(self, url):
142 return self._extract_media(self._match_id(url))
145 class TwitchVideoIE(TwitchItemBaseIE):
146 IE_NAME = 'twitch:video'
147 _VALID_URL = r'%s/[^/]+/b/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
152 'url': 'http://www.twitch.tv/riotgames/b/577357806',
155 'title': 'Worlds Semifinals - Star Horn Royal Club vs. OMG',
157 'playlist_mincount': 12,
161 class TwitchChapterIE(TwitchItemBaseIE):
162 IE_NAME = 'twitch:chapter'
163 _VALID_URL = r'%s/[^/]+/c/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
164 _ITEM_TYPE = 'chapter'
168 'url': 'http://www.twitch.tv/acracingleague/c/5285812',
171 'title': 'ACRL Off Season - Sports Cars @ Nordschleife',
173 'playlist_mincount': 3,
175 'url': 'http://www.twitch.tv/tsm_theoddone/c/2349361',
176 'only_matching': True,
180 class TwitchVodIE(TwitchItemBaseIE):
181 IE_NAME = 'twitch:vod'
182 _VALID_URL = r'%s/[^/]+/v/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
187 'url': 'http://www.twitch.tv/riotgames/v/6528877',
191 'title': 'LCK Summer Split - Week 6 Day 1',
192 'thumbnail': 're:^https?://.*\.jpg$',
194 'timestamp': 1435131709,
195 'upload_date': '20150624',
196 'uploader': 'Riot Games',
197 'uploader_id': 'riotgames',
202 'skip_download': True,
206 def _real_extract(self, url):
207 item_id = self._match_id(url)
208 info = self._download_info(self._ITEM_SHORTCUT, item_id)
209 access_token = self._download_json(
210 '%s/api/vods/%s/access_token' % (self._API_BASE, item_id), item_id,
211 'Downloading %s access token' % self._ITEM_TYPE)
212 formats = self._extract_m3u8_formats(
213 '%s/vod/%s?nauth=%s&nauthsig=%s&allow_source=true'
214 % (self._USHER_BASE, item_id, access_token['token'], access_token['sig']),
216 self._prefer_source(formats)
217 info['formats'] = formats
221 class TwitchPlaylistBaseIE(TwitchBaseIE):
222 _PLAYLIST_URL = '%s/kraken/channels/%%s/videos/?offset=%%d&limit=%%d' % TwitchBaseIE._API_BASE
225 def _extract_playlist(self, channel_id):
226 info = self._download_json(
227 '%s/kraken/channels/%s' % (self._API_BASE, channel_id),
228 channel_id, 'Downloading channel info JSON')
229 channel_name = info.get('display_name') or info.get('name')
232 limit = self._PAGE_LIMIT
233 for counter in itertools.count(1):
234 response = self._download_json(
235 self._PLAYLIST_URL % (channel_id, offset, limit),
236 channel_id, 'Downloading %s videos JSON page %d' % (self._PLAYLIST_TYPE, counter))
237 page_entries = self._extract_playlist_page(response)
240 entries.extend(page_entries)
242 return self.playlist_result(
243 [self.url_result(entry) for entry in set(entries)],
244 channel_id, channel_name)
246 def _extract_playlist_page(self, response):
247 videos = response.get('videos')
248 return [video['url'] for video in videos] if videos else []
250 def _real_extract(self, url):
251 return self._extract_playlist(self._match_id(url))
254 class TwitchProfileIE(TwitchPlaylistBaseIE):
255 IE_NAME = 'twitch:profile'
256 _VALID_URL = r'%s/(?P<id>[^/]+)/profile/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
257 _PLAYLIST_TYPE = 'profile'
260 'url': 'http://www.twitch.tv/vanillatv/profile',
263 'title': 'VanillaTV',
265 'playlist_mincount': 412,
269 class TwitchPastBroadcastsIE(TwitchPlaylistBaseIE):
270 IE_NAME = 'twitch:past_broadcasts'
271 _VALID_URL = r'%s/(?P<id>[^/]+)/profile/past_broadcasts/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
272 _PLAYLIST_URL = TwitchPlaylistBaseIE._PLAYLIST_URL + '&broadcasts=true'
273 _PLAYLIST_TYPE = 'past broadcasts'
276 'url': 'http://www.twitch.tv/spamfish/profile/past_broadcasts',
281 'playlist_mincount': 54,
285 class TwitchBookmarksIE(TwitchPlaylistBaseIE):
286 IE_NAME = 'twitch:bookmarks'
287 _VALID_URL = r'%s/(?P<id>[^/]+)/profile/bookmarks/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
288 _PLAYLIST_URL = '%s/api/bookmark/?user=%%s&offset=%%d&limit=%%d' % TwitchBaseIE._API_BASE
289 _PLAYLIST_TYPE = 'bookmarks'
292 'url': 'http://www.twitch.tv/ognos/profile/bookmarks',
297 'playlist_mincount': 3,
300 def _extract_playlist_page(self, response):
302 for bookmark in response.get('bookmarks', []):
303 video = bookmark.get('video')
306 entries.append(video['url'])
310 class TwitchStreamIE(TwitchBaseIE):
311 IE_NAME = 'twitch:stream'
312 _VALID_URL = r'%s/(?P<id>[^/]+)/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
315 'url': 'http://www.twitch.tv/shroomztv',
318 'display_id': 'shroomztv',
320 'title': 're:^ShroomzTV [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
321 'description': 'H1Z1 - lonewolfing with ShroomzTV | A3 Battle Royale later - @ShroomzTV',
323 'timestamp': 1421928037,
324 'upload_date': '20150122',
325 'uploader': 'ShroomzTV',
326 'uploader_id': 'shroomztv',
331 'skip_download': True,
335 def _real_extract(self, url):
336 channel_id = self._match_id(url)
338 stream = self._download_json(
339 '%s/kraken/streams/%s' % (self._API_BASE, channel_id), channel_id,
340 'Downloading stream JSON').get('stream')
342 # Fallback on profile extraction if stream is offline
344 return self.url_result(
345 'http://www.twitch.tv/%s/profile' % channel_id,
346 'TwitchProfile', channel_id)
348 access_token = self._download_json(
349 '%s/api/channels/%s/access_token' % (self._API_BASE, channel_id), channel_id,
350 'Downloading channel access token')
353 'allow_source': 'true',
354 'p': random.randint(1000000, 10000000),
355 'player': 'twitchweb',
356 'segment_preference': '4',
357 'sig': access_token['sig'].encode('utf-8'),
358 'token': access_token['token'].encode('utf-8'),
360 formats = self._extract_m3u8_formats(
361 '%s/api/channel/hls/%s.m3u8?%s'
362 % (self._USHER_BASE, channel_id, compat_urllib_parse.urlencode(query)),
364 self._prefer_source(formats)
366 view_count = stream.get('viewers')
367 timestamp = parse_iso8601(stream.get('created_at'))
369 channel = stream['channel']
370 title = self._live_title(channel.get('display_name') or channel.get('name'))
371 description = channel.get('status')
374 for thumbnail_key, thumbnail_url in stream['preview'].items():
375 m = re.search(r'(?P<width>\d+)x(?P<height>\d+)\.jpg$', thumbnail_key)
379 'url': thumbnail_url,
380 'width': int(m.group('width')),
381 'height': int(m.group('height')),
385 'id': compat_str(stream['_id']),
386 'display_id': channel_id,
388 'description': description,
389 'thumbnails': thumbnails,
390 'uploader': channel.get('display_name'),
391 'uploader_id': channel.get('name'),
392 'timestamp': timestamp,
393 'view_count': view_count,