X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fperiscope.py;h=8afe541ec3ee7be90571f4ce5cfc39436ae26007;hb=54fc90aabfb71968f28af68dfe3f7a3544cc2f0b;hp=11648a5116e3ae8e49ac4037906c5d1cdadd88e5;hpb=621d6a9516e0f9cd8c45e12904f4d4b7615e7fb8;p=youtube-dl diff --git a/youtube_dl/extractor/periscope.py b/youtube_dl/extractor/periscope.py index 11648a511..8afe541ec 100644 --- a/youtube_dl/extractor/periscope.py +++ b/youtube_dl/extractor/periscope.py @@ -1,6 +1,8 @@ # coding: utf-8 from __future__ import unicode_literals +import re + from .common import InfoExtractor from ..utils import ( parse_iso8601, @@ -8,9 +10,19 @@ from ..utils import ( ) -class PeriscopeIE(InfoExtractor): - _VALID_URL = r'https?://(?:www\.)?periscope\.tv/w/(?P[^/?#]+)' - _TEST = { +class PeriscopeBaseIE(InfoExtractor): + def _call_api(self, method, query, item_id): + return self._download_json( + 'https://api.periscope.tv/api/v2/%s' % method, + item_id, query=query) + + +class PeriscopeIE(PeriscopeBaseIE): + IE_DESC = 'Periscope' + IE_NAME = 'periscope' + _VALID_URL = r'https?://(?:www\.)?(?:periscope|pscp)\.tv/[^/]+/(?P[^/?#]+)' + # Alive example URLs can be found here http://onperiscope.com/ + _TESTS = [{ 'url': 'https://www.periscope.tv/w/aJUQnjY3MjA3ODF8NTYxMDIyMDl2zCg2pECBgwTqRpQuQD352EMPTKQjT4uqlM3cgWFA-g==', 'md5': '65b57957972e503fcbbaeed8f4fa04ca', 'info_dict': { @@ -23,40 +35,123 @@ class PeriscopeIE(InfoExtractor): 'uploader_id': '1465763', }, 'skip': 'Expires in 24 hours', - } + }, { + 'url': 'https://www.periscope.tv/w/1ZkKzPbMVggJv', + 'only_matching': True, + }, { + 'url': 'https://www.periscope.tv/bastaakanoggano/1OdKrlkZZjOJX', + 'only_matching': True, + }, { + 'url': 'https://www.periscope.tv/w/1ZkKzPbMVggJv', + 'only_matching': True, + }] - def _call_api(self, method, token): - return self._download_json( - 'https://api.periscope.tv/api/v2/%s?token=%s' % (method, token), token) + @staticmethod + def _extract_url(webpage): + mobj = re.search( + r']+src=([\'"])(?P(?:https?:)?//(?:www\.)?(?:periscope|pscp)\.tv/(?:(?!\1).)+)\1', webpage) + if mobj: + return mobj.group('url') def _real_extract(self, url): token = self._match_id(url) - replay = self._call_api('getAccessPublic', token) - video_url = replay['replay_url'] + stream = self._call_api( + 'accessVideoPublic', {'broadcast_id': token}, token) - broadcast_data = self._call_api('getBroadcastPublic', token) - broadcast = broadcast_data['broadcast'] - status = broadcast['status'] + broadcast = stream['broadcast'] + title = broadcast['status'] - uploader = broadcast.get('user_display_name') or broadcast_data.get('user', {}).get('display_name') - uploader_id = broadcast.get('user_id') or broadcast_data.get('user', {}).get('id') + uploader = broadcast.get('user_display_name') or broadcast.get('username') + uploader_id = (broadcast.get('user_id') or broadcast.get('username')) - title = '%s - %s' % (uploader, status) if uploader else status + title = '%s - %s' % (uploader, title) if uploader else title + state = broadcast.get('state').lower() + if state == 'running': + title = self._live_title(title) timestamp = parse_iso8601(broadcast.get('created_at')) thumbnails = [{ 'url': broadcast[image], } for image in ('image_url', 'image_url_small') if broadcast.get(image)] + video_urls = set() + formats = [] + for format_id in ('replay', 'rtmp', 'hls', 'https_hls', 'lhls', 'lhlsweb'): + video_url = stream.get(format_id + '_url') + if not video_url or video_url in video_urls: + continue + video_urls.add(video_url) + if format_id != 'rtmp': + formats.extend(self._extract_m3u8_formats( + video_url, token, 'mp4', + entry_protocol='m3u8_native' + if state in ('ended', 'timed_out') else 'm3u8', + m3u8_id=format_id, fatal=False)) + continue + formats.append({ + 'url': video_url, + 'ext': 'flv' if format_id == 'rtmp' else 'mp4', + }) + self._sort_formats(formats) + return { 'id': broadcast.get('id') or token, - 'url': video_url, - 'ext': 'mp4', - 'protocol': 'm3u8_native', 'title': title, 'timestamp': timestamp, 'uploader': uploader, 'uploader_id': uploader_id, 'thumbnails': thumbnails, + 'formats': formats, } + + +class PeriscopeUserIE(PeriscopeBaseIE): + _VALID_URL = r'https?://(?:www\.)?(?:periscope|pscp)\.tv/(?P[^/]+)/?$' + IE_DESC = 'Periscope user videos' + IE_NAME = 'periscope:user' + + _TEST = { + 'url': 'https://www.periscope.tv/LularoeHusbandMike/', + 'info_dict': { + 'id': 'LularoeHusbandMike', + 'title': 'LULAROE HUSBAND MIKE', + 'description': 'md5:6cf4ec8047768098da58e446e82c82f0', + }, + # Periscope only shows videos in the last 24 hours, so it's possible to + # get 0 videos + 'playlist_mincount': 0, + } + + def _real_extract(self, url): + user_name = self._match_id(url) + + webpage = self._download_webpage(url, user_name) + + data_store = self._parse_json( + unescapeHTML(self._search_regex( + r'data-store=(["\'])(?P.+?)\1', + webpage, 'data store', default='{}', group='data')), + user_name) + + user = list(data_store['UserCache']['users'].values())[0]['user'] + user_id = user['id'] + session_id = data_store['SessionToken']['public']['broadcastHistory']['token']['session_id'] + + broadcasts = self._call_api( + 'getUserBroadcastsPublic', + {'user_id': user_id, 'session_id': session_id}, + user_name)['broadcasts'] + + broadcast_ids = [ + broadcast['id'] for broadcast in broadcasts if broadcast.get('id')] + + title = user.get('display_name') or user.get('username') or user_name + description = user.get('description') + + entries = [ + self.url_result( + 'https://www.periscope.tv/%s/%s' % (user_name, broadcast_id)) + for broadcast_id in broadcast_ids] + + return self.playlist_result(entries, user_id, title, description)