yt_dlp/extractor/twitcasting.py

   1 import base64
   2 import itertools
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..dependencies import websockets
   7 from ..utils import (
   8     ExtractorError,
   9     UserNotLive,
  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 )
  23
  24
  25 class TwitCastingIE(InfoExtractor):
  26     _VALID_URL = r'https?://(?:[^/?#]+\.)?twitcasting\.tv/(?P<uploader_id>[^/?#]+)/(?:movie|twplayer)/(?P<id>\d+)'
  27     _M3U8_HEADERS = {
  28         'Origin': 'https://twitcasting.tv',
  29         'Referer': 'https://twitcasting.tv/',
  30     }
  31     _TESTS = [{
  32         'url': 'https://twitcasting.tv/ivetesangalo/movie/2357609',
  33         'md5': '745243cad58c4681dc752490f7540d7f',
  34         'info_dict': {
  35             'id': '2357609',
  36             'ext': 'mp4',
  37             'title': 'Live #2357609',
  38             'uploader_id': 'ivetesangalo',
  39             'description': 'Twitter Oficial da cantora brasileira Ivete Sangalo.',
  40             'thumbnail': r're:^https?://.*\.jpg$',
  41             'upload_date': '20110822',
  42             'timestamp': 1313978424,
  43             'duration': 32,
  44             'view_count': int,
  45         },
  46         'params': {
  47             'skip_download': True,
  48         },
  49     }, {
  50         'url': 'https://twitcasting.tv/mttbernardini/movie/3689740',
  51         'info_dict': {
  52             'id': '3689740',
  53             'ext': 'mp4',
  54             'title': 'Live playing something #3689740',
  55             'uploader_id': 'mttbernardini',
  56             'description': 'md5:1dc7efa2f1ab932fcd119265cebeec69',
  57             'thumbnail': r're:^https?://.*\.jpg$',
  58             'upload_date': '20120211',
  59             'timestamp': 1328995624,
  60             'duration': 681,
  61             'view_count': int,
  62         },
  63         'params': {
  64             'skip_download': True,
  65             'videopassword': 'abc',
  66         },
  67     }, {
  68         'url': 'https://twitcasting.tv/loft_heaven/movie/685979292',
  69         'info_dict': {
  70             'id': '685979292',
  71             'ext': 'mp4',
  72             'title': '【無料配信】南波一海のhear/here “ナタリー望月哲さんに聞く編集と「渋谷系狂騒曲」”',
  73             'uploader_id': 'loft_heaven',
  74             'description': 'md5:3a0c7b53019df987ce545c935538bacf',
  75             'upload_date': '20210604',
  76             'timestamp': 1622802114,
  77             'thumbnail': r're:^https?://.*\.jpg$',
  78             'duration': 6964,
  79             'view_count': int,
  80         },
  81         'params': {
  82             'skip_download': True,
  83         },
  84     }]
  85
  86     def _parse_data_movie_playlist(self, dmp, video_id):
  87         # attempt 1: parse as JSON directly
  88         try:
  89             return self._parse_json(dmp, video_id)
  90         except ExtractorError:
  91             pass
  92         # attempt 2: decode reversed base64
  93         decoded = base64.b64decode(dmp[::-1])
  94         return self._parse_json(decoded, video_id)
  95
  96     def _real_extract(self, url):
  97         uploader_id, video_id = self._match_valid_url(url).groups()
  98
  99         webpage, urlh = self._download_webpage_handle(url, video_id)
 100         video_password = self.get_param('videopassword')
 101         request_data = None
 102         if video_password:
 103             request_data = urlencode_postdata({
 104                 'password': video_password,
 105                 **self._hidden_inputs(webpage),
 106             }, encoding='utf-8')
 107             webpage, urlh = self._download_webpage_handle(
 108                 url, video_id, data=request_data,
 109                 headers={'Origin': 'https://twitcasting.tv'},
 110                 note='Trying video password')
 111         if urlh.url != url and request_data:
 112             webpage = self._download_webpage(
 113                 urlh.url, video_id, data=request_data,
 114                 headers={'Origin': 'https://twitcasting.tv'},
 115                 note='Retrying authentication')
 116         # has to check here as the first request can contain password input form even if the password is correct
 117         if re.search(r'<form\s+method="POST">\s*<input\s+[^>]+?name="password"', webpage):
 118             raise ExtractorError('This video is protected by a password, use the --video-password option', expected=True)
 119
 120         title = (clean_html(get_element_by_id('movietitle', webpage))
 121                  or self._html_search_meta(['og:title', 'twitter:title'], webpage, fatal=True))
 122
 123         video_js_data = try_get(
 124             webpage,
 125             lambda x: self._parse_data_movie_playlist(self._search_regex(
 126                 r'data-movie-playlist=\'([^\']+?)\'',
 127                 x, 'movie playlist', default=None), video_id)['2'], list)
 128
 129         thumbnail = traverse_obj(video_js_data, (0, 'thumbnailUrl')) or self._og_search_thumbnail(webpage)
 130         description = clean_html(get_element_by_id(
 131             'authorcomment', webpage)) or self._html_search_meta(
 132             ['description', 'og:description', 'twitter:description'], webpage)
 133         duration = (try_get(video_js_data, lambda x: sum(float_or_none(y.get('duration')) for y in x) / 1000)
 134                     or parse_duration(clean_html(get_element_by_class('tw-player-duration-time', webpage))))
 135         view_count = str_to_int(self._search_regex(
 136             (r'Total\s*:\s*Views\s*([\d,]+)', r'総視聴者\s*:\s*([\d,]+)\s*</'), webpage, 'views', None))
 137         timestamp = unified_timestamp(self._search_regex(
 138             r'data-toggle="true"[^>]+datetime="([^"]+)"',
 139             webpage, 'datetime', None))
 140
 141         stream_server_data = self._download_json(
 142             f'https://twitcasting.tv/streamserver.php?target={uploader_id}&mode=client', video_id,
 143             'Downloading live info', fatal=False)
 144
 145         is_live = any(f'data-{x}' in webpage for x in ['is-onlive="true"', 'live-type="live"', 'status="online"'])
 146         if not traverse_obj(stream_server_data, 'llfmp4') and is_live:
 147             self.raise_login_required(method='cookies')
 148
 149         base_dict = {
 150             'title': title,
 151             'description': description,
 152             'thumbnail': thumbnail,
 153             'timestamp': timestamp,
 154             'uploader_id': uploader_id,
 155             'duration': duration,
 156             'view_count': view_count,
 157             'is_live': is_live,
 158         }
 159
 160         def find_dmu(x):
 161             data_movie_url = self._search_regex(
 162                 r'data-movie-url=(["\'])(?P<url>(?:(?!\1).)+)\1',
 163                 x, 'm3u8 url', group='url', default=None)
 164             if data_movie_url:
 165                 return [data_movie_url]
 166
 167         m3u8_urls = (try_get(webpage, find_dmu, list)
 168                      or traverse_obj(video_js_data, (..., 'source', 'url'))
 169                      or ([f'https://twitcasting.tv/{uploader_id}/metastream.m3u8'] if is_live else None))
 170         if not m3u8_urls:
 171             raise ExtractorError('Failed to get m3u8 playlist')
 172
 173         if is_live:
 174             m3u8_url = m3u8_urls[0]
 175             formats = self._extract_m3u8_formats(
 176                 m3u8_url, video_id, ext='mp4', m3u8_id='hls',
 177                 live=True, headers=self._M3U8_HEADERS)
 178
 179             if traverse_obj(stream_server_data, ('hls', 'source')):
 180                 formats.extend(self._extract_m3u8_formats(
 181                     m3u8_url, video_id, ext='mp4', m3u8_id='source',
 182                     live=True, query={'mode': 'source'},
 183                     note='Downloading source quality m3u8',
 184                     headers=self._M3U8_HEADERS, fatal=False))
 185
 186             if websockets:
 187                 qq = qualities(['base', 'mobilesource', 'main'])
 188                 streams = traverse_obj(stream_server_data, ('llfmp4', 'streams')) or {}
 189                 for mode, ws_url in streams.items():
 190                     formats.append({
 191                         'url': ws_url,
 192                         'format_id': f'ws-{mode}',
 193                         'ext': 'mp4',
 194                         'quality': qq(mode),
 195                         'source_preference': -10,
 196                         # TwitCasting simply sends moof atom directly over WS
 197                         'protocol': 'websocket_frag',
 198                     })
 199
 200             infodict = {
 201                 'formats': formats,
 202                 '_format_sort_fields': ('source', ),
 203             }
 204         elif len(m3u8_urls) == 1:
 205             formats = self._extract_m3u8_formats(
 206                 m3u8_urls[0], video_id, 'mp4', headers=self._M3U8_HEADERS)
 207             infodict = {
 208                 # No problem here since there's only one manifest
 209                 'formats': formats,
 210                 'http_headers': self._M3U8_HEADERS,
 211             }
 212         else:
 213             infodict = {
 214                 '_type': 'multi_video',
 215                 'entries': [{
 216                     'id': f'{video_id}-{num}',
 217                     'url': m3u8_url,
 218                     'ext': 'mp4',
 219                     # Requesting the manifests here will cause download to fail.
 220                     # So use ffmpeg instead. See: https://github.com/yt-dlp/yt-dlp/issues/382
 221                     'protocol': 'm3u8',
 222                     'http_headers': self._M3U8_HEADERS,
 223                     **base_dict,
 224                 } for (num, m3u8_url) in enumerate(m3u8_urls)],
 225             }
 226
 227         return {
 228             'id': video_id,
 229             **base_dict,
 230             **infodict,
 231         }
 232
 233
 234 class TwitCastingLiveIE(InfoExtractor):
 235     _VALID_URL = r'https?://(?:[^/?#]+\.)?twitcasting\.tv/(?P<id>[^/?#]+)/?(?:[#?]|$)'
 236     _TESTS = [{
 237         'url': 'https://twitcasting.tv/ivetesangalo',
 238         'only_matching': True,
 239     }, {
 240         'url': 'https://twitcasting.tv/c:unusedlive',
 241         'expected_exception': 'UserNotLive',
 242     }]
 243
 244     def _real_extract(self, url):
 245         uploader_id = self._match_id(url)
 246         self.to_screen(
 247             f'Downloading live video of user {uploader_id}. '
 248             f'Pass "https://twitcasting.tv/{uploader_id}/show" to download the history')
 249
 250         is_live = traverse_obj(self._download_json(
 251             f'https://frontendapi.twitcasting.tv/watch/user/{uploader_id}',
 252             uploader_id, 'Checking live status', data=b'', fatal=False), ('is_live', {bool}))
 253         if is_live is False:  # only raise here if API response was as expected
 254             raise UserNotLive(video_id=uploader_id)
 255
 256         # Use /show/ page so that password-protected and members-only livestreams can be found
 257         webpage = self._download_webpage(
 258             f'https://twitcasting.tv/{uploader_id}/show/', uploader_id, 'Downloading live history')
 259         is_live = is_live or self._search_regex(
 260             r'(?s)(<span\s*class="tw-movie-thumbnail2-badge"\s*data-status="live">\s*LIVE)',
 261             webpage, 'is live?', default=False)
 262         # Current live is always the first match
 263         current_live = self._search_regex(
 264             r'(?s)<a\s+class="tw-movie-thumbnail2"\s+href="/[^/"]+/movie/(?P<video_id>\d+)"',
 265             webpage, 'current live ID', default=None, group='video_id')
 266         if not is_live or not current_live:
 267             raise UserNotLive(video_id=uploader_id)
 268
 269         return self.url_result(f'https://twitcasting.tv/{uploader_id}/movie/{current_live}', TwitCastingIE)
 270
 271
 272 class TwitCastingUserIE(InfoExtractor):
 273     _VALID_URL = r'https?://(?:[^/?#]+\.)?twitcasting\.tv/(?P<id>[^/?#]+)/(:?show|archive)/?(?:[#?]|$)'
 274     _TESTS = [{
 275         'url': 'https://twitcasting.tv/natsuiromatsuri/archive/',
 276         'info_dict': {
 277             'id': 'natsuiromatsuri',
 278             'title': 'natsuiromatsuri - Live History',
 279         },
 280         'playlist_mincount': 235,
 281     }, {
 282         'url': 'https://twitcasting.tv/noriyukicas/show',
 283         'only_matching': True,
 284     }]
 285
 286     def _entries(self, uploader_id):
 287         base_url = next_url = f'https://twitcasting.tv/{uploader_id}/show'
 288         for page_num in itertools.count(1):
 289             webpage = self._download_webpage(
 290                 next_url, uploader_id, query={'filter': 'watchable'}, note=f'Downloading page {page_num}')
 291             matches = re.finditer(
 292                 r'(?s)<a\s+class="tw-movie-thumbnail2"\s+href="(?P<url>/[^/"]+/movie/\d+)"', webpage)
 293             for mobj in matches:
 294                 yield self.url_result(urljoin(base_url, mobj.group('url')))
 295
 296             next_url = self._search_regex(
 297                 r'<a href="(/%s/show/%d-\d+)[?"]' % (re.escape(uploader_id), page_num),
 298                 webpage, 'next url', default=None)
 299             next_url = urljoin(base_url, next_url)
 300             if not next_url:
 301                 return
 302
 303     def _real_extract(self, url):
 304         uploader_id = self._match_id(url)
 305         return self.playlist_result(
 306             self._entries(uploader_id), uploader_id, f'{uploader_id} - Live History')