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