yt_dlp/extractor/twitcasting.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import itertools
   5 import re
   6
   7 from .common import InfoExtractor
   8 from ..downloader.websocket import has_websockets
   9 from ..utils import (
  10     clean_html,
  11     float_or_none,
  12     get_element_by_class,
  13     get_element_by_id,
  14     parse_duration,
  15     qualities,
  16     str_to_int,
  17     traverse_obj,
  18     try_get,
  19     unified_timestamp,
  20     urlencode_postdata,
  21     urljoin,
  22     ExtractorError,
  23 )
  24
  25
  26 class TwitCastingIE(InfoExtractor):
  27     _VALID_URL = r'https?://(?:[^/]+\.)?twitcasting\.tv/(?P<uploader_id>[^/]+)/(?:movie|twplayer)/(?P<id>\d+)'
  28     _TESTS = [{
  29         'url': 'https://twitcasting.tv/ivetesangalo/movie/2357609',
  30         'md5': '745243cad58c4681dc752490f7540d7f',
  31         'info_dict': {
  32             'id': '2357609',
  33             'ext': 'mp4',
  34             'title': 'Live #2357609',
  35             'uploader_id': 'ivetesangalo',
  36             'description': 'Twitter Oficial da cantora brasileira Ivete Sangalo.',
  37             'thumbnail': r're:^https?://.*\.jpg$',
  38             'upload_date': '20110822',
  39             'timestamp': 1314010824,
  40             'duration': 32,
  41             'view_count': int,
  42         },
  43         'params': {
  44             'skip_download': True,
  45         },
  46     }, {
  47         'url': 'https://twitcasting.tv/mttbernardini/movie/3689740',
  48         'info_dict': {
  49             'id': '3689740',
  50             'ext': 'mp4',
  51             'title': 'Live playing something #3689740',
  52             'uploader_id': 'mttbernardini',
  53             'description': 'Salve, io sono Matto (ma con la e). Questa è la mia presentazione, in quanto sono letteralmente matto (nel senso di strano), con qualcosa in più.',
  54             'thumbnail': r're:^https?://.*\.jpg$',
  55             'upload_date': '20120212',
  56             'timestamp': 1329028024,
  57             'duration': 681,
  58             'view_count': int,
  59         },
  60         'params': {
  61             'skip_download': True,
  62             'videopassword': 'abc',
  63         },
  64     }]
  65
  66     def _real_extract(self, url):
  67         uploader_id, video_id = self._match_valid_url(url).groups()
  68
  69         video_password = self.get_param('videopassword')
  70         request_data = None
  71         if video_password:
  72             request_data = urlencode_postdata({
  73                 'password': video_password,
  74             })
  75         webpage = self._download_webpage(
  76             url, video_id, data=request_data,
  77             headers={'Origin': 'https://twitcasting.tv'})
  78
  79         title = (clean_html(get_element_by_id('movietitle', webpage))
  80                  or self._html_search_meta(['og:title', 'twitter:title'], webpage, fatal=True))
  81
  82         video_js_data = {}
  83         m3u8_url = self._search_regex(
  84             r'data-movie-url=(["\'])(?P<url>(?:(?!\1).)+)\1',
  85             webpage, 'm3u8 url', group='url', default=None)
  86         if not m3u8_url:
  87             video_js_data = self._parse_json(self._search_regex(
  88                 r'data-movie-playlist=(["\'])(?P<url>(?:(?!\1).)+)',
  89                 webpage, 'movie playlist', group='url', default='[{}]'), video_id)
  90             if isinstance(video_js_data, dict):
  91                 video_js_data = list(video_js_data.values())[0]
  92             video_js_data = video_js_data[0]
  93             m3u8_url = try_get(video_js_data, lambda x: x['source']['url'])
  94
  95         stream_server_data = self._download_json(
  96             'https://twitcasting.tv/streamserver.php?target=%s&mode=client' % uploader_id, video_id,
  97             'Downloading live info', fatal=False)
  98
  99         is_live = 'data-status="online"' in webpage
 100
 101         if not traverse_obj(stream_server_data, 'llfmp4') and is_live:
 102             self.raise_login_required(method='cookies')
 103
 104         formats = []
 105         if is_live and not m3u8_url:
 106             m3u8_url = 'https://twitcasting.tv/%s/metastream.m3u8' % uploader_id
 107         if is_live and has_websockets and stream_server_data:
 108             qq = qualities(['base', 'mobilesource', 'main'])
 109             streams = traverse_obj(stream_server_data, ('llfmp4', 'streams')) or {}
 110             for mode, ws_url in streams.items():
 111                 formats.append({
 112                     'url': ws_url,
 113                     'format_id': 'ws-%s' % mode,
 114                     'ext': 'mp4',
 115                     'quality': qq(mode),
 116                     'protocol': 'websocket_frag',  # TwitCasting simply sends moof atom directly over WS
 117                 })
 118
 119         thumbnail = video_js_data.get('thumbnailUrl') or self._og_search_thumbnail(webpage)
 120         description = clean_html(get_element_by_id(
 121             'authorcomment', webpage)) or self._html_search_meta(
 122             ['description', 'og:description', 'twitter:description'], webpage)
 123         duration = float_or_none(video_js_data.get(
 124             'duration'), 1000) or parse_duration(clean_html(
 125                 get_element_by_class('tw-player-duration-time', webpage)))
 126         view_count = str_to_int(self._search_regex(
 127             r'Total\s*:\s*([\d,]+)\s*Views', webpage, 'views', None))
 128         timestamp = unified_timestamp(self._search_regex(
 129             r'data-toggle="true"[^>]+datetime="([^"]+)"',
 130             webpage, 'datetime', None))
 131
 132         if m3u8_url:
 133             formats.extend(self._extract_m3u8_formats(
 134                 m3u8_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', live=is_live))
 135         self._sort_formats(formats)
 136
 137         return {
 138             'id': video_id,
 139             'title': title,
 140             'description': description,
 141             'thumbnail': thumbnail,
 142             'timestamp': timestamp,
 143             'uploader_id': uploader_id,
 144             'duration': duration,
 145             'view_count': view_count,
 146             'formats': formats,
 147             'is_live': is_live,
 148         }
 149
 150
 151 class TwitCastingLiveIE(InfoExtractor):
 152     _VALID_URL = r'https?://(?:[^/]+\.)?twitcasting\.tv/(?P<id>[^/]+)/?(?:[#?]|$)'
 153     _TESTS = [{
 154         'url': 'https://twitcasting.tv/ivetesangalo',
 155         'only_matching': True,
 156     }]
 157
 158     def _real_extract(self, url):
 159         uploader_id = self._match_id(url)
 160         self.to_screen(
 161             'Downloading live video of user {0}. '
 162             'Pass "https://twitcasting.tv/{0}/show" to download the history'.format(uploader_id))
 163
 164         webpage = self._download_webpage(url, uploader_id)
 165         current_live = self._search_regex(
 166             (r'data-type="movie" data-id="(\d+)">',
 167              r'tw-sound-flag-open-link" data-id="(\d+)" style=',),
 168             webpage, 'current live ID', default=None)
 169         if not current_live:
 170             raise ExtractorError('The user is not currently live')
 171         return self.url_result('https://twitcasting.tv/%s/movie/%s' % (uploader_id, current_live))
 172
 173
 174 class TwitCastingUserIE(InfoExtractor):
 175     _VALID_URL = r'https?://(?:[^/]+\.)?twitcasting\.tv/(?P<id>[^/]+)/show/?(?:[#?]|$)'
 176     _TESTS = [{
 177         'url': 'https://twitcasting.tv/noriyukicas/show',
 178         'only_matching': True,
 179     }]
 180
 181     def _entries(self, uploader_id):
 182         base_url = next_url = 'https://twitcasting.tv/%s/show' % uploader_id
 183         for page_num in itertools.count(1):
 184             webpage = self._download_webpage(
 185                 next_url, uploader_id, query={'filter': 'watchable'}, note='Downloading page %d' % page_num)
 186             matches = re.finditer(
 187                 r'''(?isx)<a\s+class="tw-movie-thumbnail"\s*href="(?P<url>/[^/]+/movie/\d+)"\s*>.+?</a>''',
 188                 webpage)
 189             for mobj in matches:
 190                 yield self.url_result(urljoin(base_url, mobj.group('url')))
 191
 192             next_url = self._search_regex(
 193                 r'<a href="(/%s/show/%d-\d+)[?"]' % (re.escape(uploader_id), page_num),
 194                 webpage, 'next url', default=None)
 195             next_url = urljoin(base_url, next_url)
 196             if not next_url:
 197                 return
 198
 199     def _real_extract(self, url):
 200         uploader_id = self._match_id(url)
 201         return self.playlist_result(
 202             self._entries(uploader_id), uploader_id, '%s - Live History' % uploader_id)