yt_dlp/extractor/niconico.py

   1 import datetime as dt
   2 import functools
   3 import itertools
   4 import json
   5 import re
   6 import time
   7 import urllib.parse
   8
   9 from .common import InfoExtractor, SearchInfoExtractor
  10 from ..networking import Request
  11 from ..networking.exceptions import HTTPError
  12 from ..utils import (
  13     ExtractorError,
  14     OnDemandPagedList,
  15     clean_html,
  16     float_or_none,
  17     int_or_none,
  18     join_nonempty,
  19     parse_duration,
  20     parse_iso8601,
  21     parse_resolution,
  22     qualities,
  23     remove_start,
  24     str_or_none,
  25     traverse_obj,
  26     try_get,
  27     unescapeHTML,
  28     update_url_query,
  29     url_or_none,
  30     urlencode_postdata,
  31     urljoin,
  32 )
  33
  34
  35 class NiconicoIE(InfoExtractor):
  36     IE_NAME = 'niconico'
  37     IE_DESC = 'ニコニコ動画'
  38     _GEO_COUNTRIES = ['JP']
  39     _GEO_BYPASS = False
  40
  41     _TESTS = [{
  42         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  43         'info_dict': {
  44             'id': 'sm22312215',
  45             'ext': 'mp4',
  46             'title': 'Big Buck Bunny',
  47             'thumbnail': r're:https?://.*',
  48             'uploader': 'takuya0301',
  49             'uploader_id': '2698420',
  50             'upload_date': '20131123',
  51             'timestamp': int,  # timestamp is unstable
  52             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  53             'duration': 33,
  54             'view_count': int,
  55             'comment_count': int,
  56             'genres': ['未設定'],
  57             'tags': [],
  58         },
  59         'params': {'skip_download': 'm3u8'},
  60     }, {
  61         # File downloaded with and without credentials are different, so omit
  62         # the md5 field
  63         'url': 'http://www.nicovideo.jp/watch/nm14296458',
  64         'info_dict': {
  65             'id': 'nm14296458',
  66             'ext': 'mp4',
  67             'title': '【Kagamine Rin】Dance on media【Original】take2!',
  68             'description': 'md5:9368f2b1f4178de64f2602c2f3d6cbf5',
  69             'thumbnail': r're:https?://.*',
  70             'uploader': 'りょうた',
  71             'uploader_id': '18822557',
  72             'upload_date': '20110429',
  73             'timestamp': 1304065916,
  74             'duration': 208.0,
  75             'comment_count': int,
  76             'view_count': int,
  77             'genres': ['音楽・サウンド'],
  78             'tags': ['Translation_Request', 'Kagamine_Rin', 'Rin_Original'],
  79         },
  80         'params': {'skip_download': 'm3u8'},
  81     }, {
  82         # 'video exists but is marked as "deleted"
  83         # md5 is unstable
  84         'url': 'http://www.nicovideo.jp/watch/sm10000',
  85         'info_dict': {
  86             'id': 'sm10000',
  87             'ext': 'unknown_video',
  88             'description': 'deleted',
  89             'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞',
  90             'thumbnail': r're:https?://.*',
  91             'upload_date': '20071224',
  92             'timestamp': int,  # timestamp field has different value if logged in
  93             'duration': 304,
  94             'view_count': int,
  95         },
  96         'skip': 'Requires an account',
  97     }, {
  98         'url': 'http://www.nicovideo.jp/watch/so22543406',
  99         'info_dict': {
 100             'id': '1388129933',
 101             'ext': 'mp4',
 102             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
 103             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
 104             'thumbnail': r're:https?://.*',
 105             'timestamp': 1388851200,
 106             'upload_date': '20140104',
 107             'uploader': 'アニメロチャンネル',
 108             'uploader_id': '312',
 109         },
 110         'skip': 'The viewing period of the video you were searching for has expired.',
 111     }, {
 112         # video not available via `getflv`; "old" HTML5 video
 113         'url': 'http://www.nicovideo.jp/watch/sm1151009',
 114         'info_dict': {
 115             'id': 'sm1151009',
 116             'ext': 'mp4',
 117             'title': 'マスターシステム本体内蔵のスペハリのメインテーマ（ＰＳＧ版）',
 118             'description': 'md5:f95a3d259172667b293530cc2e41ebda',
 119             'thumbnail': r're:https?://.*',
 120             'duration': 184,
 121             'timestamp': 1190835883,
 122             'upload_date': '20070926',
 123             'uploader': 'denden2',
 124             'uploader_id': '1392194',
 125             'view_count': int,
 126             'comment_count': int,
 127             'genres': ['ゲーム'],
 128             'tags': [],
 129         },
 130         'params': {'skip_download': 'm3u8'},
 131     }, {
 132         # "New" HTML5 video
 133         'url': 'http://www.nicovideo.jp/watch/sm31464864',
 134         'info_dict': {
 135             'id': 'sm31464864',
 136             'ext': 'mp4',
 137             'title': '新作TVアニメ「戦姫絶唱シンフォギアAXZ」PV 最高画質',
 138             'description': 'md5:e52974af9a96e739196b2c1ca72b5feb',
 139             'timestamp': 1498481660,
 140             'upload_date': '20170626',
 141             'uploader': 'no-namamae',
 142             'uploader_id': '40826363',
 143             'thumbnail': r're:https?://.*',
 144             'duration': 198,
 145             'view_count': int,
 146             'comment_count': int,
 147             'genres': ['アニメ'],
 148             'tags': [],
 149         },
 150         'params': {'skip_download': 'm3u8'},
 151     }, {
 152         # Video without owner
 153         'url': 'http://www.nicovideo.jp/watch/sm18238488',
 154         'info_dict': {
 155             'id': 'sm18238488',
 156             'ext': 'mp4',
 157             'title': '【実写版】ミュータントタートルズ',
 158             'description': 'md5:15df8988e47a86f9e978af2064bf6d8e',
 159             'timestamp': 1341128008,
 160             'upload_date': '20120701',
 161             'thumbnail': r're:https?://.*',
 162             'duration': 5271,
 163             'view_count': int,
 164             'comment_count': int,
 165             'genres': ['エンターテイメント'],
 166             'tags': [],
 167         },
 168         'params': {'skip_download': 'm3u8'},
 169     }, {
 170         'url': 'http://sp.nicovideo.jp/watch/sm28964488?ss_pos=1&cp_in=wt_tg',
 171         'only_matching': True,
 172     }, {
 173         'note': 'a video that is only served as an ENCRYPTED HLS.',
 174         'url': 'https://www.nicovideo.jp/watch/so38016254',
 175         'only_matching': True,
 176     }]
 177
 178     _VALID_URL = r'https?://(?:(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch|nico\.ms)/(?P<id>(?:[a-z]{2})?[0-9]+)'
 179     _NETRC_MACHINE = 'niconico'
 180     _API_HEADERS = {
 181         'X-Frontend-ID': '6',
 182         'X-Frontend-Version': '0',
 183         'X-Niconico-Language': 'en-us',
 184         'Referer': 'https://www.nicovideo.jp/',
 185         'Origin': 'https://www.nicovideo.jp',
 186     }
 187
 188     def _perform_login(self, username, password):
 189         login_ok = True
 190         login_form_strs = {
 191             'mail_tel': username,
 192             'password': password,
 193         }
 194         self._request_webpage(
 195             'https://account.nicovideo.jp/login', None,
 196             note='Acquiring Login session')
 197         page = self._download_webpage(
 198             'https://account.nicovideo.jp/login/redirector?show_button_twitter=1&site=niconico&show_button_facebook=1', None,
 199             note='Logging in', errnote='Unable to log in',
 200             data=urlencode_postdata(login_form_strs),
 201             headers={
 202                 'Referer': 'https://account.nicovideo.jp/login',
 203                 'Content-Type': 'application/x-www-form-urlencoded',
 204             })
 205         if 'oneTimePw' in page:
 206             post_url = self._search_regex(
 207                 r'<form[^>]+action=(["\'])(?P<url>.+?)\1', page, 'post url', group='url')
 208             page = self._download_webpage(
 209                 urljoin('https://account.nicovideo.jp', post_url), None,
 210                 note='Performing MFA', errnote='Unable to complete MFA',
 211                 data=urlencode_postdata({
 212                     'otp': self._get_tfa_info('6 digits code'),
 213                 }), headers={
 214                     'Content-Type': 'application/x-www-form-urlencoded',
 215                 })
 216             if 'oneTimePw' in page or 'formError' in page:
 217                 err_msg = self._html_search_regex(
 218                     r'formError["\']+>(.*?)</div>', page, 'form_error',
 219                     default='There\'s an error but the message can\'t be parsed.',
 220                     flags=re.DOTALL)
 221                 self.report_warning(f'Unable to log in: MFA challenge failed, "{err_msg}"')
 222                 return False
 223         login_ok = 'class="notice error"' not in page
 224         if not login_ok:
 225             self.report_warning('Unable to log in: bad username or password')
 226         return login_ok
 227
 228     def _get_heartbeat_info(self, info_dict):
 229         video_id, video_src_id, audio_src_id = info_dict['url'].split(':')[1].split('/')
 230         dmc_protocol = info_dict['expected_protocol']
 231
 232         api_data = (
 233             info_dict.get('_api_data')
 234             or self._parse_json(
 235                 self._html_search_regex(
 236                     'data-api-data="([^"]+)"',
 237                     self._download_webpage('https://www.nicovideo.jp/watch/' + video_id, video_id),
 238                     'API data', default='{}'),
 239                 video_id))
 240
 241         session_api_data = try_get(api_data, lambda x: x['media']['delivery']['movie']['session'])
 242         session_api_endpoint = try_get(session_api_data, lambda x: x['urls'][0])
 243
 244         def ping():
 245             tracking_id = traverse_obj(api_data, ('media', 'delivery', 'trackingId'))
 246             if tracking_id:
 247                 tracking_url = update_url_query('https://nvapi.nicovideo.jp/v1/2ab0cbaa/watch', {'t': tracking_id})
 248                 watch_request_response = self._download_json(
 249                     tracking_url, video_id,
 250                     note='Acquiring permission for downloading video', fatal=False,
 251                     headers=self._API_HEADERS)
 252                 if traverse_obj(watch_request_response, ('meta', 'status')) != 200:
 253                     self.report_warning('Failed to acquire permission for playing video. Video download may fail.')
 254
 255         yesno = lambda x: 'yes' if x else 'no'
 256
 257         if dmc_protocol == 'http':
 258             protocol = 'http'
 259             protocol_parameters = {
 260                 'http_output_download_parameters': {
 261                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 262                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 263                 },
 264             }
 265         elif dmc_protocol == 'hls':
 266             protocol = 'm3u8'
 267             segment_duration = try_get(self._configuration_arg('segment_duration'), lambda x: int(x[0])) or 6000
 268             parsed_token = self._parse_json(session_api_data['token'], video_id)
 269             encryption = traverse_obj(api_data, ('media', 'delivery', 'encryption'))
 270             protocol_parameters = {
 271                 'hls_parameters': {
 272                     'segment_duration': segment_duration,
 273                     'transfer_preset': '',
 274                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 275                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 276                 },
 277             }
 278             if 'hls_encryption' in parsed_token and encryption:
 279                 protocol_parameters['hls_parameters']['encryption'] = {
 280                     parsed_token['hls_encryption']: {
 281                         'encrypted_key': encryption['encryptedKey'],
 282                         'key_uri': encryption['keyUri'],
 283                     },
 284                 }
 285             else:
 286                 protocol = 'm3u8_native'
 287         else:
 288             raise ExtractorError(f'Unsupported DMC protocol: {dmc_protocol}')
 289
 290         session_response = self._download_json(
 291             session_api_endpoint['url'], video_id,
 292             query={'_format': 'json'},
 293             headers={'Content-Type': 'application/json'},
 294             note='Downloading JSON metadata for {}'.format(info_dict['format_id']),
 295             data=json.dumps({
 296                 'session': {
 297                     'client_info': {
 298                         'player_id': session_api_data.get('playerId'),
 299                     },
 300                     'content_auth': {
 301                         'auth_type': try_get(session_api_data, lambda x: x['authTypes'][session_api_data['protocols'][0]]),
 302                         'content_key_timeout': session_api_data.get('contentKeyTimeout'),
 303                         'service_id': 'nicovideo',
 304                         'service_user_id': session_api_data.get('serviceUserId'),
 305                     },
 306                     'content_id': session_api_data.get('contentId'),
 307                     'content_src_id_sets': [{
 308                         'content_src_ids': [{
 309                             'src_id_to_mux': {
 310                                 'audio_src_ids': [audio_src_id],
 311                                 'video_src_ids': [video_src_id],
 312                             },
 313                         }],
 314                     }],
 315                     'content_type': 'movie',
 316                     'content_uri': '',
 317                     'keep_method': {
 318                         'heartbeat': {
 319                             'lifetime': session_api_data.get('heartbeatLifetime'),
 320                         },
 321                     },
 322                     'priority': session_api_data['priority'],
 323                     'protocol': {
 324                         'name': 'http',
 325                         'parameters': {
 326                             'http_parameters': {
 327                                 'parameters': protocol_parameters,
 328                             },
 329                         },
 330                     },
 331                     'recipe_id': session_api_data.get('recipeId'),
 332                     'session_operation_auth': {
 333                         'session_operation_auth_by_signature': {
 334                             'signature': session_api_data.get('signature'),
 335                             'token': session_api_data.get('token'),
 336                         },
 337                     },
 338                     'timing_constraint': 'unlimited',
 339                 },
 340             }).encode())
 341
 342         info_dict['url'] = session_response['data']['session']['content_uri']
 343         info_dict['protocol'] = protocol
 344
 345         # get heartbeat info
 346         heartbeat_info_dict = {
 347             'url': session_api_endpoint['url'] + '/' + session_response['data']['session']['id'] + '?_format=json&_method=PUT',
 348             'data': json.dumps(session_response['data']),
 349             # interval, convert milliseconds to seconds, then halve to make a buffer.
 350             'interval': float_or_none(session_api_data.get('heartbeatLifetime'), scale=3000),
 351             'ping': ping,
 352         }
 353
 354         return info_dict, heartbeat_info_dict
 355
 356     def _extract_format_for_quality(self, video_id, audio_quality, video_quality, dmc_protocol):
 357
 358         if not audio_quality.get('isAvailable') or not video_quality.get('isAvailable'):
 359             return None
 360
 361         format_id = '-'.join(
 362             [remove_start(s['id'], 'archive_') for s in (video_quality, audio_quality)] + [dmc_protocol])
 363
 364         vid_qual_label = traverse_obj(video_quality, ('metadata', 'label'))
 365
 366         return {
 367             'url': 'niconico_dmc:{}/{}/{}'.format(video_id, video_quality['id'], audio_quality['id']),
 368             'format_id': format_id,
 369             'format_note': join_nonempty('DMC', vid_qual_label, dmc_protocol.upper(), delim=' '),
 370             'ext': 'mp4',  # Session API are used in HTML5, which always serves mp4
 371             'acodec': 'aac',
 372             'vcodec': 'h264',
 373             **traverse_obj(audio_quality, ('metadata', {
 374                 'abr': ('bitrate', {float_or_none(scale=1000)}),
 375                 'asr': ('samplingRate', {int_or_none}),
 376             })),
 377             **traverse_obj(video_quality, ('metadata', {
 378                 'vbr': ('bitrate', {float_or_none(scale=1000)}),
 379                 'height': ('resolution', 'height', {int_or_none}),
 380                 'width': ('resolution', 'width', {int_or_none}),
 381             })),
 382             'quality': -2 if 'low' in video_quality['id'] else None,
 383             'protocol': 'niconico_dmc',
 384             'expected_protocol': dmc_protocol,  # XXX: This is not a documented field
 385             'http_headers': {
 386                 'Origin': 'https://www.nicovideo.jp',
 387                 'Referer': 'https://www.nicovideo.jp/watch/' + video_id,
 388             },
 389         }
 390
 391     def _yield_dmc_formats(self, api_data, video_id):
 392         dmc_data = traverse_obj(api_data, ('media', 'delivery', 'movie'))
 393         audios = traverse_obj(dmc_data, ('audios', ..., {dict}))
 394         videos = traverse_obj(dmc_data, ('videos', ..., {dict}))
 395         protocols = traverse_obj(dmc_data, ('session', 'protocols', ..., {str}))
 396         if not all((audios, videos, protocols)):
 397             return
 398
 399         for audio_quality, video_quality, protocol in itertools.product(audios, videos, protocols):
 400             if fmt := self._extract_format_for_quality(video_id, audio_quality, video_quality, protocol):
 401                 yield fmt
 402
 403     def _yield_dms_formats(self, api_data, video_id):
 404         fmt_filter = lambda _, v: v['isAvailable'] and v['id']
 405         videos = traverse_obj(api_data, ('media', 'domand', 'videos', fmt_filter))
 406         audios = traverse_obj(api_data, ('media', 'domand', 'audios', fmt_filter))
 407         access_key = traverse_obj(api_data, ('media', 'domand', 'accessRightKey', {str}))
 408         track_id = traverse_obj(api_data, ('client', 'watchTrackId', {str}))
 409         if not all((videos, audios, access_key, track_id)):
 410             return
 411
 412         dms_m3u8_url = self._download_json(
 413             f'https://nvapi.nicovideo.jp/v1/watch/{video_id}/access-rights/hls', video_id,
 414             data=json.dumps({
 415                 'outputs': list(itertools.product((v['id'] for v in videos), (a['id'] for a in audios))),
 416             }).encode(), query={'actionTrackId': track_id}, headers={
 417                 'x-access-right-key': access_key,
 418                 'x-frontend-id': 6,
 419                 'x-frontend-version': 0,
 420                 'x-request-with': 'https://www.nicovideo.jp',
 421             })['data']['contentUrl']
 422         # Getting all audio formats results in duplicate video formats which we filter out later
 423         dms_fmts = self._extract_m3u8_formats(dms_m3u8_url, video_id, 'mp4')
 424
 425         # m3u8 extraction does not provide audio bitrates, so extract from the API data and fix
 426         for audio_fmt in traverse_obj(dms_fmts, lambda _, v: v['vcodec'] == 'none'):
 427             yield {
 428                 **audio_fmt,
 429                 **traverse_obj(audios, (lambda _, v: audio_fmt['format_id'].startswith(v['id']), {
 430                     'format_id': ('id', {str}),
 431                     'abr': ('bitRate', {float_or_none(scale=1000)}),
 432                     'asr': ('samplingRate', {int_or_none}),
 433                 }), get_all=False),
 434                 'acodec': 'aac',
 435             }
 436
 437         # Sort before removing dupes to keep the format dicts with the lowest tbr
 438         video_fmts = sorted((fmt for fmt in dms_fmts if fmt['vcodec'] != 'none'), key=lambda f: f['tbr'])
 439         self._remove_duplicate_formats(video_fmts)
 440         # Calculate the true vbr/tbr by subtracting the lowest abr
 441         min_abr = min(traverse_obj(audios, (..., 'bitRate', {float_or_none})), default=0) / 1000
 442         for video_fmt in video_fmts:
 443             video_fmt['tbr'] -= min_abr
 444             video_fmt['format_id'] = f'video-{video_fmt["tbr"]:.0f}'
 445             yield video_fmt
 446
 447     def _real_extract(self, url):
 448         video_id = self._match_id(url)
 449
 450         try:
 451             webpage, handle = self._download_webpage_handle(
 452                 'https://www.nicovideo.jp/watch/' + video_id, video_id)
 453             if video_id.startswith('so'):
 454                 video_id = self._match_id(handle.url)
 455
 456             api_data = traverse_obj(
 457                 self._parse_json(self._html_search_meta('server-response', webpage) or '', video_id),
 458                 ('data', 'response', {dict}))
 459             if not api_data:
 460                 raise ExtractorError('Server response data not found')
 461         except ExtractorError as e:
 462             try:
 463                 api_data = self._download_json(
 464                     f'https://www.nicovideo.jp/api/watch/v3/{video_id}?_frontendId=6&_frontendVersion=0&actionTrackId=AAAAAAAAAA_{round(time.time() * 1000)}', video_id,
 465                     note='Downloading API JSON', errnote='Unable to fetch data')['data']
 466             except ExtractorError:
 467                 if not isinstance(e.cause, HTTPError):
 468                     raise
 469                 webpage = e.cause.response.read().decode('utf-8', 'replace')
 470                 error_msg = self._html_search_regex(
 471                     r'(?s)<section\s+class="(?:(?:ErrorMessage|WatchExceptionPage-message)\s*)+">(.+?)</section>',
 472                     webpage, 'error reason', default=None)
 473                 if not error_msg:
 474                     raise
 475                 raise ExtractorError(clean_html(error_msg), expected=True)
 476
 477         availability = self._availability(**(traverse_obj(api_data, ('payment', 'video', {
 478             'needs_premium': ('isPremium', {bool}),
 479             'needs_subscription': ('isAdmission', {bool}),
 480         })) or {'needs_auth': True}))
 481         formats = [*self._yield_dmc_formats(api_data, video_id),
 482                    *self._yield_dms_formats(api_data, video_id)]
 483         if not formats:
 484             fail_msg = clean_html(self._html_search_regex(
 485                 r'<p[^>]+\bclass="fail-message"[^>]*>(?P<msg>.+?)</p>',
 486                 webpage, 'fail message', default=None, group='msg'))
 487             if fail_msg:
 488                 self.to_screen(f'Niconico said: {fail_msg}')
 489             if fail_msg and 'された地域と同じ地域からのみ視聴できます。' in fail_msg:
 490                 availability = None
 491                 self.raise_geo_restricted(countries=self._GEO_COUNTRIES, metadata_available=True)
 492             elif availability == 'premium_only':
 493                 self.raise_login_required('This video requires premium', metadata_available=True)
 494             elif availability == 'subscriber_only':
 495                 self.raise_login_required('This video is for members only', metadata_available=True)
 496             elif availability == 'needs_auth':
 497                 self.raise_login_required(metadata_available=False)
 498
 499         # Start extracting information
 500         tags = None
 501         if webpage:
 502             # use og:video:tag (not logged in)
 503             og_video_tags = re.finditer(r'<meta\s+property="og:video:tag"\s*content="(.*?)">', webpage)
 504             tags = list(filter(None, (clean_html(x.group(1)) for x in og_video_tags)))
 505             if not tags:
 506                 # use keywords and split with comma (not logged in)
 507                 kwds = self._html_search_meta('keywords', webpage, default=None)
 508                 if kwds:
 509                     tags = [x for x in kwds.split(',') if x]
 510         if not tags:
 511             # find in json (logged in)
 512             tags = traverse_obj(api_data, ('tag', 'items', ..., 'name'))
 513
 514         thumb_prefs = qualities(['url', 'middleUrl', 'largeUrl', 'player', 'ogp'])
 515
 516         def get_video_info(*items, get_first=True, **kwargs):
 517             return traverse_obj(api_data, ('video', *items), get_all=not get_first, **kwargs)
 518
 519         return {
 520             'id': video_id,
 521             '_api_data': api_data,
 522             'title': get_video_info(('originalTitle', 'title')) or self._og_search_title(webpage, default=None),
 523             'formats': formats,
 524             'availability': availability,
 525             'thumbnails': [{
 526                 'id': key,
 527                 'url': url,
 528                 'ext': 'jpg',
 529                 'preference': thumb_prefs(key),
 530                 **parse_resolution(url, lenient=True),
 531             } for key, url in (get_video_info('thumbnail') or {}).items() if url],
 532             'description': clean_html(get_video_info('description')),
 533             'uploader': traverse_obj(api_data, ('owner', 'nickname'), ('channel', 'name'), ('community', 'name')),
 534             'uploader_id': str_or_none(traverse_obj(api_data, ('owner', 'id'), ('channel', 'id'), ('community', 'id'))),
 535             'timestamp': parse_iso8601(get_video_info('registeredAt')) or parse_iso8601(
 536                 self._html_search_meta('video:release_date', webpage, 'date published', default=None)),
 537             'channel': traverse_obj(api_data, ('channel', 'name'), ('community', 'name')),
 538             'channel_id': traverse_obj(api_data, ('channel', 'id'), ('community', 'id')),
 539             'view_count': int_or_none(get_video_info('count', 'view')),
 540             'tags': tags,
 541             'genre': traverse_obj(api_data, ('genre', 'label'), ('genre', 'key')),
 542             'comment_count': get_video_info('count', 'comment', expected_type=int),
 543             'duration': (
 544                 parse_duration(self._html_search_meta('video:duration', webpage, 'video duration', default=None))
 545                 or get_video_info('duration')),
 546             'webpage_url': url_or_none(url) or f'https://www.nicovideo.jp/watch/{video_id}',
 547             'subtitles': self.extract_subtitles(video_id, api_data),
 548         }
 549
 550     def _get_subtitles(self, video_id, api_data):
 551         comments_info = traverse_obj(api_data, ('comment', 'nvComment', {dict})) or {}
 552         if not comments_info.get('server'):
 553             return
 554
 555         danmaku = traverse_obj(self._download_json(
 556             f'{comments_info["server"]}/v1/threads', video_id, data=json.dumps({
 557                 'additionals': {},
 558                 'params': comments_info.get('params'),
 559                 'threadKey': comments_info.get('threadKey'),
 560             }).encode(), fatal=False,
 561             headers={
 562                 'Referer': 'https://www.nicovideo.jp/',
 563                 'Origin': 'https://www.nicovideo.jp',
 564                 'Content-Type': 'text/plain;charset=UTF-8',
 565                 'x-client-os-type': 'others',
 566                 'x-frontend-id': '6',
 567                 'x-frontend-version': '0',
 568             },
 569             note='Downloading comments', errnote='Failed to download comments'),
 570             ('data', 'threads', ..., 'comments', ...))
 571
 572         return {
 573             'comments': [{
 574                 'ext': 'json',
 575                 'data': json.dumps(danmaku),
 576             }],
 577         }
 578
 579
 580 class NiconicoPlaylistBaseIE(InfoExtractor):
 581     _PAGE_SIZE = 100
 582
 583     _API_HEADERS = {
 584         'X-Frontend-ID': '6',
 585         'X-Frontend-Version': '0',
 586         'X-Niconico-Language': 'en-us',
 587     }
 588
 589     def _call_api(self, list_id, resource, query):
 590         raise NotImplementedError('Must be implemented in subclasses')
 591
 592     @staticmethod
 593     def _parse_owner(item):
 594         return {
 595             'uploader': traverse_obj(item, ('owner', 'name')),
 596             'uploader_id': traverse_obj(item, ('owner', 'id')),
 597         }
 598
 599     def _fetch_page(self, list_id, page):
 600         page += 1
 601         resp = self._call_api(list_id, f'page {page}', {
 602             'page': page,
 603             'pageSize': self._PAGE_SIZE,
 604         })
 605         # this is needed to support both mylist and user
 606         for video in traverse_obj(resp, ('items', ..., ('video', None))) or []:
 607             video_id = video.get('id')
 608             if not video_id:
 609                 # skip {"video": {"id": "blablabla", ...}}
 610                 continue
 611             count = video.get('count') or {}
 612             get_count = lambda x: int_or_none(count.get(x))
 613             yield {
 614                 '_type': 'url',
 615                 'id': video_id,
 616                 'title': video.get('title'),
 617                 'url': f'https://www.nicovideo.jp/watch/{video_id}',
 618                 'description': video.get('shortDescription'),
 619                 'duration': int_or_none(video.get('duration')),
 620                 'view_count': get_count('view'),
 621                 'comment_count': get_count('comment'),
 622                 'thumbnail': traverse_obj(video, ('thumbnail', ('nHdUrl', 'largeUrl', 'listingUrl', 'url'))),
 623                 'ie_key': NiconicoIE.ie_key(),
 624                 **self._parse_owner(video),
 625             }
 626
 627     def _entries(self, list_id):
 628         return OnDemandPagedList(functools.partial(self._fetch_page, list_id), self._PAGE_SIZE)
 629
 630
 631 class NiconicoPlaylistIE(NiconicoPlaylistBaseIE):
 632     IE_NAME = 'niconico:playlist'
 633     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp|nico\.ms)/(?:user/\d+/)?(?:my/)?mylist/(?:#/)?(?P<id>\d+)'
 634
 635     _TESTS = [{
 636         'url': 'http://www.nicovideo.jp/mylist/27411728',
 637         'info_dict': {
 638             'id': '27411728',
 639             'title': 'AKB48のオールナイトニッポン',
 640             'description': 'md5:d89694c5ded4b6c693dea2db6e41aa08',
 641             'uploader': 'のっく',
 642             'uploader_id': '805442',
 643         },
 644         'playlist_mincount': 291,
 645     }, {
 646         'url': 'https://www.nicovideo.jp/user/805442/mylist/27411728',
 647         'only_matching': True,
 648     }, {
 649         'url': 'https://www.nicovideo.jp/my/mylist/#/68048635',
 650         'only_matching': True,
 651     }]
 652
 653     def _call_api(self, list_id, resource, query):
 654         return self._download_json(
 655             f'https://nvapi.nicovideo.jp/v2/mylists/{list_id}', list_id,
 656             f'Downloading {resource}', query=query,
 657             headers=self._API_HEADERS)['data']['mylist']
 658
 659     def _real_extract(self, url):
 660         list_id = self._match_id(url)
 661         mylist = self._call_api(list_id, 'list', {
 662             'pageSize': 1,
 663         })
 664         return self.playlist_result(
 665             self._entries(list_id), list_id,
 666             mylist.get('name'), mylist.get('description'), **self._parse_owner(mylist))
 667
 668
 669 class NiconicoSeriesIE(InfoExtractor):
 670     IE_NAME = 'niconico:series'
 671     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp(?:/user/\d+)?|nico\.ms)/series/(?P<id>\d+)'
 672
 673     _TESTS = [{
 674         'url': 'https://www.nicovideo.jp/user/44113208/series/110226',
 675         'info_dict': {
 676             'id': '110226',
 677             'title': 'ご立派ァ！のシリーズ',
 678         },
 679         'playlist_mincount': 10,
 680     }, {
 681         'url': 'https://www.nicovideo.jp/series/12312/',
 682         'info_dict': {
 683             'id': '12312',
 684             'title': 'バトルスピリッツ　お勧めカード紹介(調整中)',
 685         },
 686         'playlist_mincount': 103,
 687     }, {
 688         'url': 'https://nico.ms/series/203559',
 689         'only_matching': True,
 690     }]
 691
 692     def _real_extract(self, url):
 693         list_id = self._match_id(url)
 694         webpage = self._download_webpage(url, list_id)
 695
 696         title = self._search_regex(
 697             (r'<title>「(.+)（全',
 698              r'<div class="TwitterShareButton"\s+data-text="(.+)\s+https:'),
 699             webpage, 'title', fatal=False)
 700         if title:
 701             title = unescapeHTML(title)
 702         json_data = next(self._yield_json_ld(webpage, None, fatal=False))
 703         return self.playlist_from_matches(
 704             traverse_obj(json_data, ('itemListElement', ..., 'url')), list_id, title, ie=NiconicoIE)
 705
 706
 707 class NiconicoHistoryIE(NiconicoPlaylistBaseIE):
 708     IE_NAME = 'niconico:history'
 709     IE_DESC = 'NicoNico user history or likes. Requires cookies.'
 710     _VALID_URL = r'https?://(?:www\.|sp\.)?nicovideo\.jp/my/(?P<id>history(?:/like)?)'
 711
 712     _TESTS = [{
 713         'note': 'PC page, with /video',
 714         'url': 'https://www.nicovideo.jp/my/history/video',
 715         'only_matching': True,
 716     }, {
 717         'note': 'PC page, without /video',
 718         'url': 'https://www.nicovideo.jp/my/history',
 719         'only_matching': True,
 720     }, {
 721         'note': 'mobile page, with /video',
 722         'url': 'https://sp.nicovideo.jp/my/history/video',
 723         'only_matching': True,
 724     }, {
 725         'note': 'mobile page, without /video',
 726         'url': 'https://sp.nicovideo.jp/my/history',
 727         'only_matching': True,
 728     }, {
 729         'note': 'PC page',
 730         'url': 'https://www.nicovideo.jp/my/history/like',
 731         'only_matching': True,
 732     }, {
 733         'note': 'Mobile page',
 734         'url': 'https://sp.nicovideo.jp/my/history/like',
 735         'only_matching': True,
 736     }]
 737
 738     def _call_api(self, list_id, resource, query):
 739         path = 'likes' if list_id == 'history/like' else 'watch/history'
 740         return self._download_json(
 741             f'https://nvapi.nicovideo.jp/v1/users/me/{path}', list_id,
 742             f'Downloading {resource}', query=query, headers=self._API_HEADERS)['data']
 743
 744     def _real_extract(self, url):
 745         list_id = self._match_id(url)
 746         try:
 747             mylist = self._call_api(list_id, 'list', {'pageSize': 1})
 748         except ExtractorError as e:
 749             if isinstance(e.cause, HTTPError) and e.cause.status == 401:
 750                 self.raise_login_required('You have to be logged in to get your history')
 751             raise
 752         return self.playlist_result(self._entries(list_id), list_id, **self._parse_owner(mylist))
 753
 754
 755 class NicovideoSearchBaseIE(InfoExtractor):
 756     _SEARCH_TYPE = 'search'
 757
 758     def _entries(self, url, item_id, query=None, note='Downloading page %(page)s'):
 759         query = query or {}
 760         pages = [query['page']] if 'page' in query else itertools.count(1)
 761         for page_num in pages:
 762             query['page'] = str(page_num)
 763             webpage = self._download_webpage(url, item_id, query=query, note=note % {'page': page_num})
 764             results = re.findall(r'(?<=data-video-id=)["\']?(?P<videoid>.*?)(?=["\'])', webpage)
 765             for item in results:
 766                 yield self.url_result(f'https://www.nicovideo.jp/watch/{item}', 'Niconico', item)
 767             if not results:
 768                 break
 769
 770     def _search_results(self, query):
 771         return self._entries(
 772             self._proto_relative_url(f'//www.nicovideo.jp/{self._SEARCH_TYPE}/{query}'), query)
 773
 774
 775 class NicovideoSearchIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 776     IE_DESC = 'Nico video search'
 777     IE_NAME = 'nicovideo:search'
 778     _SEARCH_KEY = 'nicosearch'
 779
 780
 781 class NicovideoSearchURLIE(NicovideoSearchBaseIE):
 782     IE_NAME = f'{NicovideoSearchIE.IE_NAME}_url'
 783     IE_DESC = 'Nico video search URLs'
 784     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/search/(?P<id>[^?#&]+)?'
 785     _TESTS = [{
 786         'url': 'http://www.nicovideo.jp/search/sm9',
 787         'info_dict': {
 788             'id': 'sm9',
 789             'title': 'sm9',
 790         },
 791         'playlist_mincount': 40,
 792     }, {
 793         'url': 'https://www.nicovideo.jp/search/sm9?sort=h&order=d&end=2020-12-31&start=2020-01-01',
 794         'info_dict': {
 795             'id': 'sm9',
 796             'title': 'sm9',
 797         },
 798         'playlist_count': 31,
 799     }]
 800
 801     def _real_extract(self, url):
 802         query = self._match_id(url)
 803         return self.playlist_result(self._entries(url, query), query, query)
 804
 805
 806 class NicovideoSearchDateIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 807     IE_DESC = 'Nico video search, newest first'
 808     IE_NAME = f'{NicovideoSearchIE.IE_NAME}:date'
 809     _SEARCH_KEY = 'nicosearchdate'
 810     _TESTS = [{
 811         'url': 'nicosearchdateall:a',
 812         'info_dict': {
 813             'id': 'a',
 814             'title': 'a',
 815         },
 816         'playlist_mincount': 1610,
 817     }]
 818
 819     _START_DATE = dt.date(2007, 1, 1)
 820     _RESULTS_PER_PAGE = 32
 821     _MAX_PAGES = 50
 822
 823     def _entries(self, url, item_id, start_date=None, end_date=None):
 824         start_date, end_date = start_date or self._START_DATE, end_date or dt.datetime.now().date()
 825
 826         # If the last page has a full page of videos, we need to break down the query interval further
 827         last_page_len = len(list(self._get_entries_for_date(
 828             url, item_id, start_date, end_date, self._MAX_PAGES,
 829             note=f'Checking number of videos from {start_date} to {end_date}')))
 830         if (last_page_len == self._RESULTS_PER_PAGE and start_date != end_date):
 831             midpoint = start_date + ((end_date - start_date) // 2)
 832             yield from self._entries(url, item_id, midpoint, end_date)
 833             yield from self._entries(url, item_id, start_date, midpoint)
 834         else:
 835             self.to_screen(f'{item_id}: Downloading results from {start_date} to {end_date}')
 836             yield from self._get_entries_for_date(
 837                 url, item_id, start_date, end_date, note='    Downloading page %(page)s')
 838
 839     def _get_entries_for_date(self, url, item_id, start_date, end_date=None, page_num=None, note=None):
 840         query = {
 841             'start': str(start_date),
 842             'end': str(end_date or start_date),
 843             'sort': 'f',
 844             'order': 'd',
 845         }
 846         if page_num:
 847             query['page'] = str(page_num)
 848
 849         yield from super()._entries(url, item_id, query=query, note=note)
 850
 851
 852 class NicovideoTagURLIE(NicovideoSearchBaseIE):
 853     IE_NAME = 'niconico:tag'
 854     IE_DESC = 'NicoNico video tag URLs'
 855     _SEARCH_TYPE = 'tag'
 856     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/tag/(?P<id>[^?#&]+)?'
 857     _TESTS = [{
 858         'url': 'https://www.nicovideo.jp/tag/ドキュメンタリー淫夢',
 859         'info_dict': {
 860             'id': 'ドキュメンタリー淫夢',
 861             'title': 'ドキュメンタリー淫夢',
 862         },
 863         'playlist_mincount': 400,
 864     }]
 865
 866     def _real_extract(self, url):
 867         query = self._match_id(url)
 868         return self.playlist_result(self._entries(url, query), query, query)
 869
 870
 871 class NiconicoUserIE(InfoExtractor):
 872     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/user/(?P<id>\d+)(?:/video)?/?(?:$|[#?])'
 873     _TEST = {
 874         'url': 'https://www.nicovideo.jp/user/419948',
 875         'info_dict': {
 876             'id': '419948',
 877         },
 878         'playlist_mincount': 101,
 879     }
 880     _API_URL = 'https://nvapi.nicovideo.jp/v2/users/%s/videos?sortKey=registeredAt&sortOrder=desc&pageSize=%s&page=%s'
 881     _PAGE_SIZE = 100
 882
 883     _API_HEADERS = {
 884         'X-Frontend-ID': '6',
 885         'X-Frontend-Version': '0',
 886     }
 887
 888     def _entries(self, list_id):
 889         total_count = 1
 890         count = page_num = 0
 891         while count < total_count:
 892             json_parsed = self._download_json(
 893                 self._API_URL % (list_id, self._PAGE_SIZE, page_num + 1), list_id,
 894                 headers=self._API_HEADERS,
 895                 note='Downloading JSON metadata%s' % (f' page {page_num}' if page_num else ''))
 896             if not page_num:
 897                 total_count = int_or_none(json_parsed['data'].get('totalCount'))
 898             for entry in json_parsed['data']['items']:
 899                 count += 1
 900                 yield self.url_result(
 901                     f'https://www.nicovideo.jp/watch/{entry["essential"]["id"]}', ie=NiconicoIE)
 902             page_num += 1
 903
 904     def _real_extract(self, url):
 905         list_id = self._match_id(url)
 906         return self.playlist_result(self._entries(list_id), list_id)
 907
 908
 909 class NiconicoLiveIE(InfoExtractor):
 910     IE_NAME = 'niconico:live'
 911     IE_DESC = 'ニコニコ生放送'
 912     _VALID_URL = r'https?://(?:sp\.)?live2?\.nicovideo\.jp/(?:watch|gate)/(?P<id>lv\d+)'
 913     _TESTS = [{
 914         'note': 'this test case includes invisible characters for title, pasting them as-is',
 915         'url': 'https://live.nicovideo.jp/watch/lv339533123',
 916         'info_dict': {
 917             'id': 'lv339533123',
 918             'title': '激辛ペヤング食べます\u202a( ;ᯅ; )\u202c（歌枠オーディション参加中）',
 919             'view_count': 1526,
 920             'comment_count': 1772,
 921             'description': '初めましてもかって言います❕\nのんびり自由に適当に暮らしてます',
 922             'uploader': 'もか',
 923             'channel': 'ゲストさんのコミュニティ',
 924             'channel_id': 'co5776900',
 925             'channel_url': 'https://com.nicovideo.jp/community/co5776900',
 926             'timestamp': 1670677328,
 927             'is_live': True,
 928         },
 929         'skip': 'livestream',
 930     }, {
 931         'url': 'https://live2.nicovideo.jp/watch/lv339533123',
 932         'only_matching': True,
 933     }, {
 934         'url': 'https://sp.live.nicovideo.jp/watch/lv339533123',
 935         'only_matching': True,
 936     }, {
 937         'url': 'https://sp.live2.nicovideo.jp/watch/lv339533123',
 938         'only_matching': True,
 939     }]
 940
 941     _KNOWN_LATENCY = ('high', 'low')
 942
 943     def _real_extract(self, url):
 944         video_id = self._match_id(url)
 945         webpage, urlh = self._download_webpage_handle(f'https://live.nicovideo.jp/watch/{video_id}', video_id)
 946
 947         embedded_data = self._parse_json(unescapeHTML(self._search_regex(
 948             r'<script\s+id="embedded-data"\s*data-props="(.+?)"', webpage, 'embedded data')), video_id)
 949
 950         ws_url = traverse_obj(embedded_data, ('site', 'relive', 'webSocketUrl'))
 951         if not ws_url:
 952             raise ExtractorError('The live hasn\'t started yet or already ended.', expected=True)
 953         ws_url = update_url_query(ws_url, {
 954             'frontend_id': traverse_obj(embedded_data, ('site', 'frontendId')) or '9',
 955         })
 956
 957         hostname = remove_start(urllib.parse.urlparse(urlh.url).hostname, 'sp.')
 958         latency = try_get(self._configuration_arg('latency'), lambda x: x[0])
 959         if latency not in self._KNOWN_LATENCY:
 960             latency = 'high'
 961
 962         ws = self._request_webpage(
 963             Request(ws_url, headers={'Origin': f'https://{hostname}'}),
 964             video_id=video_id, note='Connecting to WebSocket server')
 965
 966         self.write_debug('[debug] Sending HLS server request')
 967         ws.send(json.dumps({
 968             'type': 'startWatching',
 969             'data': {
 970                 'stream': {
 971                     'quality': 'abr',
 972                     'protocol': 'hls+fmp4',
 973                     'latency': latency,
 974                     'chasePlay': False,
 975                 },
 976                 'room': {
 977                     'protocol': 'webSocket',
 978                     'commentable': True,
 979                 },
 980                 'reconnect': False,
 981             },
 982         }))
 983
 984         while True:
 985             recv = ws.recv()
 986             if not recv:
 987                 continue
 988             data = json.loads(recv)
 989             if not isinstance(data, dict):
 990                 continue
 991             if data.get('type') == 'stream':
 992                 m3u8_url = data['data']['uri']
 993                 qualities = data['data']['availableQualities']
 994                 break
 995             elif data.get('type') == 'disconnect':
 996                 self.write_debug(recv)
 997                 raise ExtractorError('Disconnected at middle of extraction')
 998             elif data.get('type') == 'error':
 999                 self.write_debug(recv)
1000                 message = traverse_obj(data, ('body', 'code')) or recv
1001                 raise ExtractorError(message)
1002             elif self.get_param('verbose', False):
1003                 if len(recv) > 100:
1004                     recv = recv[:100] + '...'
1005                 self.write_debug(f'Server said: {recv}')
1006
1007         title = traverse_obj(embedded_data, ('program', 'title')) or self._html_search_meta(
1008             ('og:title', 'twitter:title'), webpage, 'live title', fatal=False)
1009
1010         raw_thumbs = traverse_obj(embedded_data, ('program', 'thumbnail')) or {}
1011         thumbnails = []
1012         for name, value in raw_thumbs.items():
1013             if not isinstance(value, dict):
1014                 thumbnails.append({
1015                     'id': name,
1016                     'url': value,
1017                     **parse_resolution(value, lenient=True),
1018                 })
1019                 continue
1020
1021             for k, img_url in value.items():
1022                 res = parse_resolution(k, lenient=True) or parse_resolution(img_url, lenient=True)
1023                 width, height = res.get('width'), res.get('height')
1024
1025                 thumbnails.append({
1026                     'id': f'{name}_{width}x{height}',
1027                     'url': img_url,
1028                     **res,
1029                 })
1030
1031         formats = self._extract_m3u8_formats(m3u8_url, video_id, ext='mp4', live=True)
1032         for fmt, q in zip(formats, reversed(qualities[1:])):
1033             fmt.update({
1034                 'format_id': q,
1035                 'protocol': 'niconico_live',
1036                 'ws': ws,
1037                 'video_id': video_id,
1038                 'live_latency': latency,
1039                 'origin': hostname,
1040             })
1041
1042         return {
1043             'id': video_id,
1044             'title': title,
1045             **traverse_obj(embedded_data, {
1046                 'view_count': ('program', 'statistics', 'watchCount'),
1047                 'comment_count': ('program', 'statistics', 'commentCount'),
1048                 'uploader': ('program', 'supplier', 'name'),
1049                 'channel': ('socialGroup', 'name'),
1050                 'channel_id': ('socialGroup', 'id'),
1051                 'channel_url': ('socialGroup', 'socialGroupPageUrl'),
1052             }),
1053             'description': clean_html(traverse_obj(embedded_data, ('program', 'description'))),
1054             'timestamp': int_or_none(traverse_obj(embedded_data, ('program', 'openTime'))),
1055             'is_live': True,
1056             'thumbnails': thumbnails,
1057             'formats': formats,
1058         }