import itertools
import re
+import random
from .common import InfoExtractor
from ..compat import (
+ compat_str,
compat_urllib_parse,
compat_urllib_request,
)
class TwitchBaseIE(InfoExtractor):
- _VALID_URL_BASE = r'http://(?:www\.)?twitch\.tv'
+ _VALID_URL_BASE = r'https?://(?:www\.)?twitch\.tv'
_API_BASE = 'https://api.twitch.tv'
+ _USHER_BASE = 'http://usher.twitch.tv'
_LOGIN_URL = 'https://secure.twitch.tv/user/login'
+ _LOGIN_POST_URL = 'https://secure-login.twitch.tv/login'
+ _NETRC_MACHINE = 'twitch'
def _handle_error(self, response):
if not isinstance(response, dict):
expected=True)
def _download_json(self, url, video_id, note='Downloading JSON metadata'):
- response = super(TwitchBaseIE, self)._download_json(url, video_id, note)
+ 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 = compat_urllib_request.Request(url, headers=headers)
+ response = super(TwitchBaseIE, self)._download_json(request, video_id, note)
self._handle_error(response)
return response
'authenticity_token': authenticity_token,
'redirect_on_login': '',
'embed_form': 'false',
- 'mp_source_action': '',
+ 'mp_source_action': 'login-button',
'follow': '',
- 'user[login]': username,
- 'user[password]': password,
+ 'login': username,
+ 'password': password,
}
request = compat_urllib_request.Request(
- self._LOGIN_URL, compat_urllib_parse.urlencode(login_form).encode('utf-8'))
+ self._LOGIN_POST_URL, compat_urllib_parse.urlencode(login_form).encode('utf-8'))
request.add_header('Referer', self._LOGIN_URL)
response = self._download_webpage(
request, None, 'Logging in as %s' % username)
raise ExtractorError(
'Unable to login: %s' % m.group('msg').strip(), expected=True)
+ def _prefer_source(self, formats):
+ try:
+ source = next(f for f in formats if f['format_id'] == 'Source')
+ source['preference'] = 10
+ except StopIteration:
+ pass # No Source stream present
+ self._sort_formats(formats)
+
class TwitchItemBaseIE(TwitchBaseIE):
def _download_info(self, item, item_id):
class TwitchVideoIE(TwitchItemBaseIE):
IE_NAME = 'twitch:video'
- _VALID_URL = r'%s/[^/]+/b/(?P<id>[^/]+)' % TwitchBaseIE._VALID_URL_BASE
+ _VALID_URL = r'%s/[^/]+/b/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
_ITEM_TYPE = 'video'
_ITEM_SHORTCUT = 'a'
class TwitchChapterIE(TwitchItemBaseIE):
IE_NAME = 'twitch:chapter'
- _VALID_URL = r'%s/[^/]+/c/(?P<id>[^/]+)' % TwitchBaseIE._VALID_URL_BASE
+ _VALID_URL = r'%s/[^/]+/c/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
_ITEM_TYPE = 'chapter'
_ITEM_SHORTCUT = 'c'
- _TEST = {
+ _TESTS = [{
'url': 'http://www.twitch.tv/acracingleague/c/5285812',
'info_dict': {
'id': 'c5285812',
'title': 'ACRL Off Season - Sports Cars @ Nordschleife',
},
'playlist_mincount': 3,
- }
+ }, {
+ '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>[^/]+)' % TwitchBaseIE._VALID_URL_BASE
+ _VALID_URL = r'%s/[^/]+/v/(?P<id>\d+)' % TwitchBaseIE._VALID_URL_BASE
_ITEM_TYPE = 'vod'
_ITEM_SHORTCUT = 'v'
'%s/api/vods/%s/access_token' % (self._API_BASE, item_id), item_id,
'Downloading %s access token' % self._ITEM_TYPE)
formats = self._extract_m3u8_formats(
- 'http://usher.twitch.tv/vod/%s?nauth=%s&nauthsig=%s'
- % (item_id, access_token['token'], access_token['sig']),
+ '%s/vod/%s?nauth=%s&nauthsig=%s'
+ % (self._USHER_BASE, item_id, access_token['token'], access_token['sig']),
item_id, 'mp4')
+ self._prefer_source(formats)
info['formats'] = formats
return info
response = self._download_json(
self._PLAYLIST_URL % (channel_id, offset, limit),
channel_id, 'Downloading %s videos JSON page %d' % (self._PLAYLIST_TYPE, counter))
- videos = response['videos']
- if not videos:
+ page_entries = self._extract_playlist_page(response)
+ if not page_entries:
break
- entries.extend([self.url_result(video['url']) for video in videos])
+ entries.extend(page_entries)
offset += limit
- return self.playlist_result(entries, channel_id, channel_name)
+ return self.playlist_result(
+ [self.url_result(entry) for entry in set(entries)],
+ channel_id, channel_name)
+
+ 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 TwitchPastBroadcastsIE(TwitchPlaylistBaseIE):
- IE_NAME = 'twitch:profile'
+ IE_NAME = 'twitch:past_broadcasts'
_VALID_URL = r'%s/(?P<id>[^/]+)/profile/past_broadcasts/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
_PLAYLIST_URL = TwitchPlaylistBaseIE._PLAYLIST_URL + '&broadcasts=true'
_PLAYLIST_TYPE = 'past broadcasts'
},
'playlist_mincount': 54,
}
+
+
+class TwitchBookmarksIE(TwitchPlaylistBaseIE):
+ IE_NAME = 'twitch:bookmarks'
+ _VALID_URL = r'%s/(?P<id>[^/]+)/profile/bookmarks/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
+ _PLAYLIST_URL = '%s/api/bookmark/?user=%%s&offset=%%d&limit=%%d' % TwitchBaseIE._API_BASE
+ _PLAYLIST_TYPE = 'bookmarks'
+
+ _TEST = {
+ 'url': 'http://www.twitch.tv/ognos/profile/bookmarks',
+ 'info_dict': {
+ 'id': 'ognos',
+ 'title': 'Ognos',
+ },
+ 'playlist_mincount': 3,
+ }
+
+ 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 TwitchStreamIE(TwitchBaseIE):
+ IE_NAME = 'twitch:stream'
+ _VALID_URL = r'%s/(?P<id>[^/]+)/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
+
+ _TEST = {
+ 'url': 'http://www.twitch.tv/shroomztv',
+ 'info_dict': {
+ 'id': '12772022048',
+ 'display_id': 'shroomztv',
+ 'ext': 'mp4',
+ 'title': 're:^ShroomzTV [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
+ '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,
+ },
+ }
+
+ def _real_extract(self, url):
+ channel_id = self._match_id(url)
+
+ stream = self._download_json(
+ '%s/kraken/streams/%s' % (self._API_BASE, 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)
+
+ access_token = self._download_json(
+ '%s/api/channels/%s/access_token' % (self._API_BASE, channel_id), channel_id,
+ 'Downloading channel access token')
+
+ query = {
+ 'allow_source': '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,
+ }