]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/periscope.py
debian/copyright: use spaces rather than tabs to start continuation lines.
[youtubedl] / youtube_dl / extractor / periscope.py
index c23b314e79df70e6b115fee6fb91386345d00ad4..c02e34abac8720361f94b7085e5f7fb3814df312 100644 (file)
@@ -1,18 +1,70 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import re
+
 from .common import InfoExtractor
 from ..utils import (
+    int_or_none,
     parse_iso8601,
     unescapeHTML,
 )
 
 
-class PeriscopeIE(InfoExtractor):
+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
+
+    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'
     IE_NAME = 'periscope'
-    _VALID_URL = r'https?://(?:www\.)?periscope\.tv/[^/]+/(?P<id>[^/?#]+)'
-    # Alive example URLs can be found here http://onperiscope.com/
+    _VALID_URL = r'https?://(?:www\.)?(?:periscope|pscp)\.tv/[^/]+/(?P<id>[^/?#]+)'
+    # Alive example URLs can be found here https://www.periscope.tv/
     _TESTS = [{
         'url': 'https://www.periscope.tv/w/aJUQnjY3MjA3ODF8NTYxMDIyMDl2zCg2pECBgwTqRpQuQD352EMPTKQjT4uqlM3cgWFA-g==',
         'md5': '65b57957972e503fcbbaeed8f4fa04ca',
@@ -32,64 +84,62 @@ 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'<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)
 
-        user = broadcast_data.get('user', {})
+        broadcast = stream['broadcast']
+        info = self._parse_broadcast_data(broadcast, token)
 
-        uploader = broadcast.get('user_display_name') or user.get('display_name')
-        uploader_id = (broadcast.get('username') or user.get('username') or
-                       broadcast.get('user_id') or user.get('id'))
-
-        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(InfoExtractor):
-    _VALID_URL = r'https?://www\.periscope\.tv/(?P<id>[^/]+)/?$'
+class PeriscopeUserIE(PeriscopeBaseIE):
+    _VALID_URL = r'https?://(?:www\.)?(?:periscope|pscp)\.tv/(?P<id>[^/]+)/?$'
     IE_DESC = 'Periscope user videos'
     IE_NAME = 'periscope:user'
 
@@ -106,23 +156,34 @@ class PeriscopeUserIE(InfoExtractor):
     }
 
     def _real_extract(self, url):
-        user_id = self._match_id(url)
+        user_name = self._match_id(url)
 
-        webpage = self._download_webpage(url, user_id)
+        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_id)
+            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')]
 
-        user = data_store.get('User', {}).get('user', {})
-        title = user.get('display_name') or user.get('username')
+        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_id, broadcast['id']))
-            for broadcast in data_store.get('UserBroadcastHistory', {}).get('broadcasts', [])]
+                '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)