X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/9f2b33881274af98a9145c533a1d295fad71521a..7b54e7278c1ff62e888689ec470f488682311356:/youtube_dl/extractor/periscope.py diff --git a/youtube_dl/extractor/periscope.py b/youtube_dl/extractor/periscope.py index 514e9b4..c02e34a 100644 --- a/youtube_dl/extractor/periscope.py +++ b/youtube_dl/extractor/periscope.py @@ -1,14 +1,70 @@ # coding: utf-8 from __future__ import unicode_literals +import re + from .common import InfoExtractor -from ..utils import parse_iso8601 +from ..utils import ( + int_or_none, + parse_iso8601, + unescapeHTML, +) + + +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) + + def _parse_broadcast_data(self, broadcast, video_id): + title = broadcast['status'] + uploader = broadcast.get('user_display_name') or broadcast.get('username') + title = '%s - %s' % (uploader, title) if uploader else title + is_live = broadcast.get('state').lower() == 'running' + + thumbnails = [{ + 'url': broadcast[image], + } for image in ('image_url', 'image_url_small') if broadcast.get(image)] + + return { + 'id': broadcast.get('id') or video_id, + 'title': self._live_title(title) if is_live else title, + 'timestamp': parse_iso8601(broadcast.get('created_at')), + 'uploader': uploader, + 'uploader_id': broadcast.get('user_id') or broadcast.get('username'), + 'thumbnails': thumbnails, + 'view_count': int_or_none(broadcast.get('total_watched')), + 'tags': broadcast.get('tags'), + 'is_live': is_live, + } + + @staticmethod + def _extract_common_format_info(broadcast): + return broadcast.get('state').lower(), int_or_none(broadcast.get('width')), int_or_none(broadcast.get('height')) + @staticmethod + def _add_width_and_height(f, width, height): + for key, val in (('width', width), ('height', height)): + if not f.get(key): + f[key] = val -class PeriscopeIE(InfoExtractor): + def _extract_pscp_m3u8_formats(self, m3u8_url, video_id, format_id, state, width, height, fatal=True): + m3u8_formats = self._extract_m3u8_formats( + m3u8_url, video_id, 'mp4', + entry_protocol='m3u8_native' + if state in ('ended', 'timed_out') else 'm3u8', + m3u8_id=format_id, fatal=fatal) + if len(m3u8_formats) == 1: + self._add_width_and_height(m3u8_formats[0], width, height) + return m3u8_formats + + +class PeriscopeIE(PeriscopeBaseIE): IE_DESC = 'Periscope' - _VALID_URL = r'https?://(?:www\.)?periscope\.tv/[^/]+/(?P[^/?#]+)' - # Alive example URLs can be found here http://onperiscope.com/ + IE_NAME = 'periscope' + _VALID_URL = r'https?://(?:www\.)?(?:periscope|pscp)\.tv/[^/]+/(?P[^/?#]+)' + # Alive example URLs can be found here https://www.periscope.tv/ _TESTS = [{ 'url': 'https://www.periscope.tv/w/aJUQnjY3MjA3ODF8NTYxMDIyMDl2zCg2pECBgwTqRpQuQD352EMPTKQjT4uqlM3cgWFA-g==', 'md5': '65b57957972e503fcbbaeed8f4fa04ca', @@ -28,54 +84,106 @@ class PeriscopeIE(InfoExtractor): }, { '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, value): - return self._download_json( - 'https://api.periscope.tv/api/v2/%s?broadcast_id=%s' % (method, value), value) + @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) - broadcast_data = self._call_api('getBroadcastPublic', token) - broadcast = broadcast_data['broadcast'] - status = broadcast['status'] + stream = self._call_api( + 'accessVideoPublic', {'broadcast_id': token}, token) - 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') + broadcast = stream['broadcast'] + info = self._parse_broadcast_data(broadcast, token) - title = '%s - %s' % (uploader, status) if uploader else status state = broadcast.get('state').lower() - if state == 'running': - title = self._live_title(title) - timestamp = parse_iso8601(broadcast.get('created_at')) + width = int_or_none(broadcast.get('width')) + height = int_or_none(broadcast.get('height')) - thumbnails = [{ - 'url': broadcast[image], - } for image in ('image_url', 'image_url_small') if broadcast.get(image)] - - stream = self._call_api('getAccessPublic', token) + def add_width_and_height(f): + for key, val in (('width', width), ('height', height)): + if not f.get(key): + f[key] = val + video_urls = set() formats = [] - for format_id in ('replay', 'rtmp', 'hls', 'https_hls'): + for format_id in ('replay', 'rtmp', 'hls', 'https_hls', 'lhls', 'lhlsweb'): video_url = stream.get(format_id + '_url') - if not video_url: + if not video_url or video_url in video_urls: + continue + video_urls.add(video_url) + if format_id != 'rtmp': + m3u8_formats = self._extract_pscp_m3u8_formats( + video_url, token, format_id, state, width, height, False) + formats.extend(m3u8_formats) continue - f = { + rtmp_format = { 'url': video_url, 'ext': 'flv' if format_id == 'rtmp' else 'mp4', } - if format_id != 'rtmp': - f['protocol'] = 'm3u8_native' if state == 'ended' else 'm3u8' - formats.append(f) + self._add_width_and_height(rtmp_format) + formats.append(rtmp_format) self._sort_formats(formats) - return { - 'id': broadcast.get('id') or token, - 'title': title, - 'timestamp': timestamp, - 'uploader': uploader, - 'uploader_id': uploader_id, - 'thumbnails': thumbnails, - 'formats': formats, - } + info['formats'] = formats + return info + + +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)