Update upstream source from tag 'upstream/2019.01.16'

[youtubedl] / youtube_dl / extractor / periscope.py

diff --git a/youtube_dl/extractor/periscope.py b/youtube_dl/extractor/periscope.py
index 63cc764bb8eceed80893606e640466b355279a0d..8afe541ec3ee7be90571f4ce5cfc39436ae26007 100644 (file)
--- a/youtube_dl/extractor/periscope.py
+++ b/youtube_dl/extractor/periscope.py
@@ -1,13 +1,26 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import re
+
 from .common import InfoExtractor
-from ..utils import parse_iso8601
+from ..utils import (
+    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)
 
 
-class PeriscopeIE(InfoExtractor):
+class PeriscopeIE(PeriscopeBaseIE):
     IE_DESC = 'Periscope'
-    _VALID_URL = r'https?://(?:www\.)?periscope\.tv/[^/]+/(?P<id>[^/?#]+)'
+    IE_NAME = 'periscope'
+    _VALID_URL = r'https?://(?:www\.)?(?:periscope|pscp)\.tv/[^/]+/(?P<id>[^/?#]+)'
     # Alive example URLs can be found here http://onperiscope.com/
     _TESTS = [{
         'url': 'https://www.periscope.tv/w/aJUQnjY3MjA3ODF8NTYxMDIyMDl2zCg2pECBgwTqRpQuQD352EMPTKQjT4uqlM3cgWFA-g==',
@@ -28,24 +41,31 @@ 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):
-        attribute = 'token' if len(value) > 13 else 'broadcast_id'
-        return self._download_json(
-            'https://api.periscope.tv/api/v2/%s?%s=%s' % (method, attribute, value), value)
+    @staticmethod
+    def _extract_url(webpage):
+        mobj = re.search(
+            r'<iframe[^>]+src=([\'"])(?P<url>(?: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)
+
+        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)
@@ -55,20 +75,24 @@ class PeriscopeIE(InfoExtractor):
             'url': broadcast[image],
         } for image in ('image_url', 'image_url_small') if broadcast.get(image)]
 
-        stream = self._call_api('getAccessPublic', token)
-
+        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
-            f = {
+            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',
-            }
-            if format_id != 'rtmp':
-                f['protocol'] = 'm3u8_native' if state == 'ended' else 'm3u8'
-            formats.append(f)
+            })
         self._sort_formats(formats)
 
         return {
@@ -80,3 +104,54 @@ class PeriscopeIE(InfoExtractor):
             'thumbnails': thumbnails,
             'formats': formats,
         }
+
+
+class PeriscopeUserIE(PeriscopeBaseIE):
+    _VALID_URL = r'https?://(?:www\.)?(?:periscope|pscp)\.tv/(?P<id>[^/]+)/?$'
+    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<data>.+?)\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)