summaryrefslogtreecommitdiff
path: root/youtube_dl/extractor/periscope.py
blob: 0e362302425cbe504b33b90aa1937dc68b9e288a (plain)
    1 # coding: utf-8
    2 from __future__ import unicode_literals
    3 
    4 import re
    5 
    6 from .common import InfoExtractor
    7 from ..utils import (
    8     parse_iso8601,
    9     unescapeHTML,
   10 )
   11 
   12 
   13 class PeriscopeBaseIE(InfoExtractor):
   14     def _call_api(self, method, query, item_id):
   15         return self._download_json(
   16             'https://api.periscope.tv/api/v2/%s' % method,
   17             item_id, query=query)
   18 
   19 
   20 class PeriscopeIE(PeriscopeBaseIE):
   21     IE_DESC = 'Periscope'
   22     IE_NAME = 'periscope'
   23     _VALID_URL = r'https?://(?:www\.)?periscope\.tv/[^/]+/(?P<id>[^/?#]+)'
   24     # Alive example URLs can be found here http://onperiscope.com/
   25     _TESTS = [{
   26         'url': 'https://www.periscope.tv/w/aJUQnjY3MjA3ODF8NTYxMDIyMDl2zCg2pECBgwTqRpQuQD352EMPTKQjT4uqlM3cgWFA-g==',
   27         'md5': '65b57957972e503fcbbaeed8f4fa04ca',
   28         'info_dict': {
   29             'id': '56102209',
   30             'ext': 'mp4',
   31             'title': 'Bec Boop - ๐Ÿš โœˆ๏ธ๐Ÿ‡ฌ๐Ÿ‡ง Fly above #London in Emirates Air Line cable car at night ๐Ÿ‡ฌ๐Ÿ‡งโœˆ๏ธ๐Ÿš  #BoopScope ๐ŸŽ€๐Ÿ’—',
   32             'timestamp': 1438978559,
   33             'upload_date': '20150807',
   34             'uploader': 'Bec Boop',
   35             'uploader_id': '1465763',
   36         },
   37         'skip': 'Expires in 24 hours',
   38     }, {
   39         'url': 'https://www.periscope.tv/w/1ZkKzPbMVggJv',
   40         'only_matching': True,
   41     }, {
   42         'url': 'https://www.periscope.tv/bastaakanoggano/1OdKrlkZZjOJX',
   43         'only_matching': True,
   44     }]
   45 
   46     @staticmethod
   47     def _extract_url(webpage):
   48         mobj = re.search(
   49             r'<iframe[^>]+src=([\'"])(?P<url>(?:https?:)?//(?:www\.)?periscope\.tv/(?:(?!\1).)+)\1', webpage)
   50         if mobj:
   51             return mobj.group('url')
   52 
   53     def _real_extract(self, url):
   54         token = self._match_id(url)
   55 
   56         broadcast_data = self._call_api(
   57             'getBroadcastPublic', {'broadcast_id': token}, token)
   58         broadcast = broadcast_data['broadcast']
   59         status = broadcast['status']
   60 
   61         user = broadcast_data.get('user', {})
   62 
   63         uploader = broadcast.get('user_display_name') or user.get('display_name')
   64         uploader_id = (broadcast.get('username') or user.get('username') or
   65                        broadcast.get('user_id') or user.get('id'))
   66 
   67         title = '%s - %s' % (uploader, status) if uploader else status
   68         state = broadcast.get('state').lower()
   69         if state == 'running':
   70             title = self._live_title(title)
   71         timestamp = parse_iso8601(broadcast.get('created_at'))
   72 
   73         thumbnails = [{
   74             'url': broadcast[image],
   75         } for image in ('image_url', 'image_url_small') if broadcast.get(image)]
   76 
   77         stream = self._call_api(
   78             'getAccessPublic', {'broadcast_id': token}, token)
   79 
   80         formats = []
   81         for format_id in ('replay', 'rtmp', 'hls', 'https_hls'):
   82             video_url = stream.get(format_id + '_url')
   83             if not video_url:
   84                 continue
   85             f = {
   86                 'url': video_url,
   87                 'ext': 'flv' if format_id == 'rtmp' else 'mp4',
   88             }
   89             if format_id != 'rtmp':
   90                 f['protocol'] = 'm3u8_native' if state in ('ended', 'timed_out') else 'm3u8'
   91             formats.append(f)
   92         self._sort_formats(formats)
   93 
   94         return {
   95             'id': broadcast.get('id') or token,
   96             'title': title,
   97             'timestamp': timestamp,
   98             'uploader': uploader,
   99             'uploader_id': uploader_id,
  100             'thumbnails': thumbnails,
  101             'formats': formats,
  102         }
  103 
  104 
  105 class PeriscopeUserIE(PeriscopeBaseIE):
  106     _VALID_URL = r'https?://(?:www\.)?periscope\.tv/(?P<id>[^/]+)/?$'
  107     IE_DESC = 'Periscope user videos'
  108     IE_NAME = 'periscope:user'
  109 
  110     _TEST = {
  111         'url': 'https://www.periscope.tv/LularoeHusbandMike/',
  112         'info_dict': {
  113             'id': 'LularoeHusbandMike',
  114             'title': 'LULAROE HUSBAND MIKE',
  115             'description': 'md5:6cf4ec8047768098da58e446e82c82f0',
  116         },
  117         # Periscope only shows videos in the last 24 hours, so it's possible to
  118         # get 0 videos
  119         'playlist_mincount': 0,
  120     }
  121 
  122     def _real_extract(self, url):
  123         user_name = self._match_id(url)
  124 
  125         webpage = self._download_webpage(url, user_name)
  126 
  127         data_store = self._parse_json(
  128             unescapeHTML(self._search_regex(
  129                 r'data-store=(["\'])(?P<data>.+?)\1',
  130                 webpage, 'data store', default='{}', group='data')),
  131             user_name)
  132 
  133         user = list(data_store['UserCache']['users'].values())[0]['user']
  134         user_id = user['id']
  135         session_id = data_store['SessionToken']['public']['broadcastHistory']['token']['session_id']
  136 
  137         broadcasts = self._call_api(
  138             'getUserBroadcastsPublic',
  139             {'user_id': user_id, 'session_id': session_id},
  140             user_name)['broadcasts']
  141 
  142         broadcast_ids = [
  143             broadcast['id'] for broadcast in broadcasts if broadcast.get('id')]
  144 
  145         title = user.get('display_name') or user.get('username') or user_name
  146         description = user.get('description')
  147 
  148         entries = [
  149             self.url_result(
  150                 'https://www.periscope.tv/%s/%s' % (user_name, broadcast_id))
  151             for broadcast_id in broadcast_ids]
  152 
  153         return self.playlist_result(entries, user_id, title, description)

Generated by cgit