]> jfr.im git - yt-dlp.git/blob - yt_dlp/extractor/periscope.py
[cleanup, docs] Misc cleanup
[yt-dlp.git] / yt_dlp / 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 _M3U8_HEADERS = {
16 'Referer': 'https://www.periscope.tv/'
17 }
18
19 def _call_api(self, method, query, item_id):
20 return self._download_json(
21 'https://api.periscope.tv/api/v2/%s' % method,
22 item_id, query=query)
23
24 def _parse_broadcast_data(self, broadcast, video_id):
25 title = broadcast.get('status') or 'Periscope Broadcast'
26 uploader = broadcast.get('user_display_name') or broadcast.get('username')
27 title = '%s - %s' % (uploader, title) if uploader else title
28 is_live = broadcast.get('state').lower() == 'running'
29
30 thumbnails = [{
31 'url': broadcast[image],
32 } for image in ('image_url', 'image_url_small') if broadcast.get(image)]
33
34 return {
35 'id': broadcast.get('id') or video_id,
36 'title': title,
37 'timestamp': parse_iso8601(broadcast.get('created_at')),
38 'uploader': uploader,
39 'uploader_id': broadcast.get('user_id') or broadcast.get('username'),
40 'thumbnails': thumbnails,
41 'view_count': int_or_none(broadcast.get('total_watched')),
42 'tags': broadcast.get('tags'),
43 'is_live': is_live,
44 }
45
46 @staticmethod
47 def _extract_common_format_info(broadcast):
48 return broadcast.get('state').lower(), int_or_none(broadcast.get('width')), int_or_none(broadcast.get('height'))
49
50 @staticmethod
51 def _add_width_and_height(f, width, height):
52 for key, val in (('width', width), ('height', height)):
53 if not f.get(key):
54 f[key] = val
55
56 def _extract_pscp_m3u8_formats(self, m3u8_url, video_id, format_id, state, width, height, fatal=True):
57 m3u8_formats = self._extract_m3u8_formats(
58 m3u8_url, video_id, 'mp4',
59 entry_protocol='m3u8_native'
60 if state in ('ended', 'timed_out') else 'm3u8',
61 m3u8_id=format_id, fatal=fatal, headers=self._M3U8_HEADERS)
62 if len(m3u8_formats) == 1:
63 self._add_width_and_height(m3u8_formats[0], width, height)
64 for f in m3u8_formats:
65 f.setdefault('http_headers', {}).update(self._M3U8_HEADERS)
66 return m3u8_formats
67
68
69 class PeriscopeIE(PeriscopeBaseIE):
70 IE_DESC = 'Periscope'
71 IE_NAME = 'periscope'
72 _VALID_URL = r'https?://(?:www\.)?(?:periscope|pscp)\.tv/[^/]+/(?P<id>[^/?#]+)'
73 # Alive example URLs can be found here https://www.periscope.tv/
74 _TESTS = [{
75 'url': 'https://www.periscope.tv/w/aJUQnjY3MjA3ODF8NTYxMDIyMDl2zCg2pECBgwTqRpQuQD352EMPTKQjT4uqlM3cgWFA-g==',
76 'md5': '65b57957972e503fcbbaeed8f4fa04ca',
77 'info_dict': {
78 'id': '56102209',
79 'ext': 'mp4',
80 '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’—',
81 'timestamp': 1438978559,
82 'upload_date': '20150807',
83 'uploader': 'Bec Boop',
84 'uploader_id': '1465763',
85 },
86 'skip': 'Expires in 24 hours',
87 }, {
88 'url': 'https://www.periscope.tv/w/1ZkKzPbMVggJv',
89 'only_matching': True,
90 }, {
91 'url': 'https://www.periscope.tv/bastaakanoggano/1OdKrlkZZjOJX',
92 'only_matching': True,
93 }, {
94 'url': 'https://www.periscope.tv/w/1ZkKzPbMVggJv',
95 'only_matching': True,
96 }]
97
98 @staticmethod
99 def _extract_url(webpage):
100 mobj = re.search(
101 r'<iframe[^>]+src=([\'"])(?P<url>(?:https?:)?//(?:www\.)?(?:periscope|pscp)\.tv/(?:(?!\1).)+)\1', webpage)
102 if mobj:
103 return mobj.group('url')
104
105 def _real_extract(self, url):
106 token = self._match_id(url)
107
108 stream = self._call_api(
109 'accessVideoPublic', {'broadcast_id': token}, token)
110
111 broadcast = stream['broadcast']
112 info = self._parse_broadcast_data(broadcast, token)
113
114 state = broadcast.get('state').lower()
115 width = int_or_none(broadcast.get('width'))
116 height = int_or_none(broadcast.get('height'))
117
118 def add_width_and_height(f):
119 for key, val in (('width', width), ('height', height)):
120 if not f.get(key):
121 f[key] = val
122
123 video_urls = set()
124 formats = []
125 for format_id in ('replay', 'rtmp', 'hls', 'https_hls', 'lhls', 'lhlsweb'):
126 video_url = stream.get(format_id + '_url')
127 if not video_url or video_url in video_urls:
128 continue
129 video_urls.add(video_url)
130 if format_id != 'rtmp':
131 m3u8_formats = self._extract_pscp_m3u8_formats(
132 video_url, token, format_id, state, width, height, False)
133 formats.extend(m3u8_formats)
134 continue
135 rtmp_format = {
136 'url': video_url,
137 'ext': 'flv' if format_id == 'rtmp' else 'mp4',
138 }
139 self._add_width_and_height(rtmp_format)
140 formats.append(rtmp_format)
141 self._sort_formats(formats)
142
143 info['formats'] = formats
144 return info
145
146
147 class PeriscopeUserIE(PeriscopeBaseIE):
148 _VALID_URL = r'https?://(?:www\.)?(?:periscope|pscp)\.tv/(?P<id>[^/]+)/?$'
149 IE_DESC = 'Periscope user videos'
150 IE_NAME = 'periscope:user'
151
152 _TEST = {
153 'url': 'https://www.periscope.tv/LularoeHusbandMike/',
154 'info_dict': {
155 'id': 'LularoeHusbandMike',
156 'title': 'LULAROE HUSBAND MIKE',
157 'description': 'md5:6cf4ec8047768098da58e446e82c82f0',
158 },
159 # Periscope only shows videos in the last 24 hours, so it's possible to
160 # get 0 videos
161 'playlist_mincount': 0,
162 }
163
164 def _real_extract(self, url):
165 user_name = self._match_id(url)
166
167 webpage = self._download_webpage(url, user_name)
168
169 data_store = self._parse_json(
170 unescapeHTML(self._search_regex(
171 r'data-store=(["\'])(?P<data>.+?)\1',
172 webpage, 'data store', default='{}', group='data')),
173 user_name)
174
175 user = list(data_store['UserCache']['users'].values())[0]['user']
176 user_id = user['id']
177 session_id = data_store['SessionToken']['public']['broadcastHistory']['token']['session_id']
178
179 broadcasts = self._call_api(
180 'getUserBroadcastsPublic',
181 {'user_id': user_id, 'session_id': session_id},
182 user_name)['broadcasts']
183
184 broadcast_ids = [
185 broadcast['id'] for broadcast in broadcasts if broadcast.get('id')]
186
187 title = user.get('display_name') or user.get('username') or user_name
188 description = user.get('description')
189
190 entries = [
191 self.url_result(
192 'https://www.periscope.tv/%s/%s' % (user_name, broadcast_id))
193 for broadcast_id in broadcast_ids]
194
195 return self.playlist_result(entries, user_id, title, description)