Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/periscope.py

   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     int_or_none,
   9     parse_iso8601,
  10     unescapeHTML,
  11 )
  12
  13
  14 class PeriscopeBaseIE(InfoExtractor):
  15     def _call_api(self, method, query, item_id):
  16         return self._download_json(
  17             'https://api.periscope.tv/api/v2/%s' % method,
  18             item_id, query=query)
  19
  20
  21 class PeriscopeIE(PeriscopeBaseIE):
  22     IE_DESC = 'Periscope'
  23     IE_NAME = 'periscope'
  24     _VALID_URL = r'https?://(?:www\.)?(?:periscope|pscp)\.tv/[^/]+/(?P<id>[^/?#]+)'
  25     # Alive example URLs can be found here http://onperiscope.com/
  26     _TESTS = [{
  27         'url': 'https://www.periscope.tv/w/aJUQnjY3MjA3ODF8NTYxMDIyMDl2zCg2pECBgwTqRpQuQD352EMPTKQjT4uqlM3cgWFA-g==',
  28         'md5': '65b57957972e503fcbbaeed8f4fa04ca',
  29         'info_dict': {
  30             'id': '56102209',
  31             'ext': 'mp4',
  32             'title': 'Bec Boop - 🚠✈️🇬🇧 Fly above #London in Emirates Air Line cable car at night 🇬🇧✈️🚠 #BoopScope 🎀💗',
  33             'timestamp': 1438978559,
  34             'upload_date': '20150807',
  35             'uploader': 'Bec Boop',
  36             'uploader_id': '1465763',
  37         },
  38         'skip': 'Expires in 24 hours',
  39     }, {
  40         'url': 'https://www.periscope.tv/w/1ZkKzPbMVggJv',
  41         'only_matching': True,
  42     }, {
  43         'url': 'https://www.periscope.tv/bastaakanoggano/1OdKrlkZZjOJX',
  44         'only_matching': True,
  45     }, {
  46         'url': 'https://www.periscope.tv/w/1ZkKzPbMVggJv',
  47         'only_matching': True,
  48     }]
  49
  50     @staticmethod
  51     def _extract_url(webpage):
  52         mobj = re.search(
  53             r'<iframe[^>]+src=([\'"])(?P<url>(?:https?:)?//(?:www\.)?(?:periscope|pscp)\.tv/(?:(?!\1).)+)\1', webpage)
  54         if mobj:
  55             return mobj.group('url')
  56
  57     def _real_extract(self, url):
  58         token = self._match_id(url)
  59
  60         stream = self._call_api(
  61             'accessVideoPublic', {'broadcast_id': token}, token)
  62
  63         broadcast = stream['broadcast']
  64         title = broadcast['status']
  65
  66         uploader = broadcast.get('user_display_name') or broadcast.get('username')
  67         uploader_id = (broadcast.get('user_id') or broadcast.get('username'))
  68
  69         title = '%s - %s' % (uploader, title) if uploader else title
  70         state = broadcast.get('state').lower()
  71         if state == 'running':
  72             title = self._live_title(title)
  73         timestamp = parse_iso8601(broadcast.get('created_at'))
  74
  75         thumbnails = [{
  76             'url': broadcast[image],
  77         } for image in ('image_url', 'image_url_small') if broadcast.get(image)]
  78
  79         width = int_or_none(broadcast.get('width'))
  80         height = int_or_none(broadcast.get('height'))
  81
  82         def add_width_and_height(f):
  83             for key, val in (('width', width), ('height', height)):
  84                 if not f.get(key):
  85                     f[key] = val
  86
  87         video_urls = set()
  88         formats = []
  89         for format_id in ('replay', 'rtmp', 'hls', 'https_hls', 'lhls', 'lhlsweb'):
  90             video_url = stream.get(format_id + '_url')
  91             if not video_url or video_url in video_urls:
  92                 continue
  93             video_urls.add(video_url)
  94             if format_id != 'rtmp':
  95                 m3u8_formats = self._extract_m3u8_formats(
  96                     video_url, token, 'mp4',
  97                     entry_protocol='m3u8_native'
  98                     if state in ('ended', 'timed_out') else 'm3u8',
  99                     m3u8_id=format_id, fatal=False)
 100                 if len(m3u8_formats) == 1:
 101                     add_width_and_height(m3u8_formats[0])
 102                 formats.extend(m3u8_formats)
 103                 continue
 104             rtmp_format = {
 105                 'url': video_url,
 106                 'ext': 'flv' if format_id == 'rtmp' else 'mp4',
 107             }
 108             add_width_and_height(rtmp_format)
 109             formats.append(rtmp_format)
 110         self._sort_formats(formats)
 111
 112         return {
 113             'id': broadcast.get('id') or token,
 114             'title': title,
 115             'timestamp': timestamp,
 116             'uploader': uploader,
 117             'uploader_id': uploader_id,
 118             'thumbnails': thumbnails,
 119             'formats': formats,
 120         }
 121
 122
 123 class PeriscopeUserIE(PeriscopeBaseIE):
 124     _VALID_URL = r'https?://(?:www\.)?(?:periscope|pscp)\.tv/(?P<id>[^/]+)/?$'
 125     IE_DESC = 'Periscope user videos'
 126     IE_NAME = 'periscope:user'
 127
 128     _TEST = {
 129         'url': 'https://www.periscope.tv/LularoeHusbandMike/',
 130         'info_dict': {
 131             'id': 'LularoeHusbandMike',
 132             'title': 'LULAROE HUSBAND MIKE',
 133             'description': 'md5:6cf4ec8047768098da58e446e82c82f0',
 134         },
 135         # Periscope only shows videos in the last 24 hours, so it's possible to
 136         # get 0 videos
 137         'playlist_mincount': 0,
 138     }
 139
 140     def _real_extract(self, url):
 141         user_name = self._match_id(url)
 142
 143         webpage = self._download_webpage(url, user_name)
 144
 145         data_store = self._parse_json(
 146             unescapeHTML(self._search_regex(
 147                 r'data-store=(["\'])(?P<data>.+?)\1',
 148                 webpage, 'data store', default='{}', group='data')),
 149             user_name)
 150
 151         user = list(data_store['UserCache']['users'].values())[0]['user']
 152         user_id = user['id']
 153         session_id = data_store['SessionToken']['public']['broadcastHistory']['token']['session_id']
 154
 155         broadcasts = self._call_api(
 156             'getUserBroadcastsPublic',
 157             {'user_id': user_id, 'session_id': session_id},
 158             user_name)['broadcasts']
 159
 160         broadcast_ids = [
 161             broadcast['id'] for broadcast in broadcasts if broadcast.get('id')]
 162
 163         title = user.get('display_name') or user.get('username') or user_name
 164         description = user.get('description')
 165
 166         entries = [
 167             self.url_result(
 168                 'https://www.periscope.tv/%s/%s' % (user_name, broadcast_id))
 169             for broadcast_id in broadcast_ids]
 170
 171         return self.playlist_result(entries, user_id, title, description)