yt_dlp/extractor/niconico.py

   1 import datetime
   2 import functools
   3 import itertools
   4 import json
   5 import re
   6 import time
   7
   8 from urllib.parse import urlparse
   9
  10 from .common import InfoExtractor, SearchInfoExtractor
  11 from ..networking import Request
  12 from ..networking.exceptions import HTTPError
  13 from ..utils import (
  14     ExtractorError,
  15     OnDemandPagedList,
  16     clean_html,
  17     float_or_none,
  18     int_or_none,
  19     join_nonempty,
  20     parse_duration,
  21     parse_iso8601,
  22     parse_resolution,
  23     qualities,
  24     remove_start,
  25     str_or_none,
  26     traverse_obj,
  27     try_get,
  28     unescapeHTML,
  29     update_url_query,
  30     url_or_none,
  31     urlencode_postdata,
  32     urljoin,
  33 )
  34
  35
  36 class NiconicoIE(InfoExtractor):
  37     IE_NAME = 'niconico'
  38     IE_DESC = 'ニコニコ動画'
  39
  40     _TESTS = [{
  41         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  42         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  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             'expected_protocol': str,
  59         },
  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             'expected_protocol': str,
  80         },
  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         'md5': 'f95a3d259172667b293530cc2e41ebda',
 115         'info_dict': {
 116             'id': 'sm1151009',
 117             'ext': 'mp4',
 118             '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）',
 119             'description': 'md5:f95a3d259172667b293530cc2e41ebda',
 120             'thumbnail': r're:https?://.*',
 121             'duration': 184,
 122             'timestamp': 1190835883,
 123             'upload_date': '20070926',
 124             'uploader': 'denden2',
 125             'uploader_id': '1392194',
 126             'view_count': int,
 127             'comment_count': int,
 128             'genres': ['ゲーム'],
 129             'tags': [],
 130             'expected_protocol': str,
 131         },
 132     }, {
 133         # "New" HTML5 video
 134         # md5 is unstable
 135         'url': 'http://www.nicovideo.jp/watch/sm31464864',
 136         'info_dict': {
 137             'id': 'sm31464864',
 138             'ext': 'mp4',
 139             'title': '新作TVアニメ「戦姫絶唱シンフォギアAXZ」PV 最高画質',
 140             'description': 'md5:e52974af9a96e739196b2c1ca72b5feb',
 141             'timestamp': 1498481660,
 142             'upload_date': '20170626',
 143             'uploader': 'no-namamae',
 144             'uploader_id': '40826363',
 145             'thumbnail': r're:https?://.*',
 146             'duration': 198,
 147             'view_count': int,
 148             'comment_count': int,
 149             'genres': ['アニメ'],
 150             'tags': [],
 151             'expected_protocol': str,
 152         },
 153     }, {
 154         # Video without owner
 155         'url': 'http://www.nicovideo.jp/watch/sm18238488',
 156         'md5': 'd265680a1f92bdcbbd2a507fc9e78a9e',
 157         'info_dict': {
 158             'id': 'sm18238488',
 159             'ext': 'mp4',
 160             'title': '【実写版】ミュータントタートルズ',
 161             'description': 'md5:15df8988e47a86f9e978af2064bf6d8e',
 162             'timestamp': 1341128008,
 163             'upload_date': '20120701',
 164             'uploader': None,
 165             'uploader_id': None,
 166             'thumbnail': r're:https?://.*',
 167             'duration': 5271,
 168             'view_count': int,
 169             'comment_count': int,
 170             'genres': ['エンターテイメント'],
 171             'tags': [],
 172             'expected_protocol': str,
 173         },
 174     }, {
 175         'url': 'http://sp.nicovideo.jp/watch/sm28964488?ss_pos=1&cp_in=wt_tg',
 176         'only_matching': True,
 177     }, {
 178         'note': 'a video that is only served as an ENCRYPTED HLS.',
 179         'url': 'https://www.nicovideo.jp/watch/so38016254',
 180         'only_matching': True,
 181     }]
 182
 183     _VALID_URL = r'https?://(?:(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch|nico\.ms)/(?P<id>(?:[a-z]{2})?[0-9]+)'
 184     _NETRC_MACHINE = 'niconico'
 185     _API_HEADERS = {
 186         'X-Frontend-ID': '6',
 187         'X-Frontend-Version': '0',
 188         'X-Niconico-Language': 'en-us',
 189         'Referer': 'https://www.nicovideo.jp/',
 190         'Origin': 'https://www.nicovideo.jp',
 191     }
 192
 193     def _perform_login(self, username, password):
 194         login_ok = True
 195         login_form_strs = {
 196             'mail_tel': username,
 197             'password': password,
 198         }
 199         self._request_webpage(
 200             'https://account.nicovideo.jp/login', None,
 201             note='Acquiring Login session')
 202         page = self._download_webpage(
 203             'https://account.nicovideo.jp/login/redirector?show_button_twitter=1&site=niconico&show_button_facebook=1', None,
 204             note='Logging in', errnote='Unable to log in',
 205             data=urlencode_postdata(login_form_strs),
 206             headers={
 207                 'Referer': 'https://account.nicovideo.jp/login',
 208                 'Content-Type': 'application/x-www-form-urlencoded',
 209             })
 210         if 'oneTimePw' in page:
 211             post_url = self._search_regex(
 212                 r'<form[^>]+action=(["\'])(?P<url>.+?)\1', page, 'post url', group='url')
 213             page = self._download_webpage(
 214                 urljoin('https://account.nicovideo.jp', post_url), None,
 215                 note='Performing MFA', errnote='Unable to complete MFA',
 216                 data=urlencode_postdata({
 217                     'otp': self._get_tfa_info('6 digits code')
 218                 }), headers={
 219                     'Content-Type': 'application/x-www-form-urlencoded',
 220                 })
 221             if 'oneTimePw' in page or 'formError' in page:
 222                 err_msg = self._html_search_regex(
 223                     r'formError["\']+>(.*?)</div>', page, 'form_error',
 224                     default='There\'s an error but the message can\'t be parsed.',
 225                     flags=re.DOTALL)
 226                 self.report_warning(f'Unable to log in: MFA challenge failed, "{err_msg}"')
 227                 return False
 228         login_ok = 'class="notice error"' not in page
 229         if not login_ok:
 230             self.report_warning('Unable to log in: bad username or password')
 231         return login_ok
 232
 233     def _get_heartbeat_info(self, info_dict):
 234         video_id, video_src_id, audio_src_id = info_dict['url'].split(':')[1].split('/')
 235         dmc_protocol = info_dict['expected_protocol']
 236
 237         api_data = (
 238             info_dict.get('_api_data')
 239             or self._parse_json(
 240                 self._html_search_regex(
 241                     'data-api-data="([^"]+)"',
 242                     self._download_webpage('https://www.nicovideo.jp/watch/' + video_id, video_id),
 243                     'API data', default='{}'),
 244                 video_id))
 245
 246         session_api_data = try_get(api_data, lambda x: x['media']['delivery']['movie']['session'])
 247         session_api_endpoint = try_get(session_api_data, lambda x: x['urls'][0])
 248
 249         def ping():
 250             tracking_id = traverse_obj(api_data, ('media', 'delivery', 'trackingId'))
 251             if tracking_id:
 252                 tracking_url = update_url_query('https://nvapi.nicovideo.jp/v1/2ab0cbaa/watch', {'t': tracking_id})
 253                 watch_request_response = self._download_json(
 254                     tracking_url, video_id,
 255                     note='Acquiring permission for downloading video', fatal=False,
 256                     headers=self._API_HEADERS)
 257                 if traverse_obj(watch_request_response, ('meta', 'status')) != 200:
 258                     self.report_warning('Failed to acquire permission for playing video. Video download may fail.')
 259
 260         yesno = lambda x: 'yes' if x else 'no'
 261
 262         if dmc_protocol == 'http':
 263             protocol = 'http'
 264             protocol_parameters = {
 265                 'http_output_download_parameters': {
 266                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 267                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 268                 }
 269             }
 270         elif dmc_protocol == 'hls':
 271             protocol = 'm3u8'
 272             segment_duration = try_get(self._configuration_arg('segment_duration'), lambda x: int(x[0])) or 6000
 273             parsed_token = self._parse_json(session_api_data['token'], video_id)
 274             encryption = traverse_obj(api_data, ('media', 'delivery', 'encryption'))
 275             protocol_parameters = {
 276                 'hls_parameters': {
 277                     'segment_duration': segment_duration,
 278                     'transfer_preset': '',
 279                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 280                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 281                 }
 282             }
 283             if 'hls_encryption' in parsed_token and encryption:
 284                 protocol_parameters['hls_parameters']['encryption'] = {
 285                     parsed_token['hls_encryption']: {
 286                         'encrypted_key': encryption['encryptedKey'],
 287                         'key_uri': encryption['keyUri'],
 288                     }
 289                 }
 290             else:
 291                 protocol = 'm3u8_native'
 292         else:
 293             raise ExtractorError(f'Unsupported DMC protocol: {dmc_protocol}')
 294
 295         session_response = self._download_json(
 296             session_api_endpoint['url'], video_id,
 297             query={'_format': 'json'},
 298             headers={'Content-Type': 'application/json'},
 299             note='Downloading JSON metadata for %s' % info_dict['format_id'],
 300             data=json.dumps({
 301                 'session': {
 302                     'client_info': {
 303                         'player_id': session_api_data.get('playerId'),
 304                     },
 305                     'content_auth': {
 306                         'auth_type': try_get(session_api_data, lambda x: x['authTypes'][session_api_data['protocols'][0]]),
 307                         'content_key_timeout': session_api_data.get('contentKeyTimeout'),
 308                         'service_id': 'nicovideo',
 309                         'service_user_id': session_api_data.get('serviceUserId')
 310                     },
 311                     'content_id': session_api_data.get('contentId'),
 312                     'content_src_id_sets': [{
 313                         'content_src_ids': [{
 314                             'src_id_to_mux': {
 315                                 'audio_src_ids': [audio_src_id],
 316                                 'video_src_ids': [video_src_id],
 317                             }
 318                         }]
 319                     }],
 320                     'content_type': 'movie',
 321                     'content_uri': '',
 322                     'keep_method': {
 323                         'heartbeat': {
 324                             'lifetime': session_api_data.get('heartbeatLifetime')
 325                         }
 326                     },
 327                     'priority': session_api_data['priority'],
 328                     'protocol': {
 329                         'name': 'http',
 330                         'parameters': {
 331                             'http_parameters': {
 332                                 'parameters': protocol_parameters
 333                             }
 334                         }
 335                     },
 336                     'recipe_id': session_api_data.get('recipeId'),
 337                     'session_operation_auth': {
 338                         'session_operation_auth_by_signature': {
 339                             'signature': session_api_data.get('signature'),
 340                             'token': session_api_data.get('token'),
 341                         }
 342                     },
 343                     'timing_constraint': 'unlimited'
 344                 }
 345             }).encode())
 346
 347         info_dict['url'] = session_response['data']['session']['content_uri']
 348         info_dict['protocol'] = protocol
 349
 350         # get heartbeat info
 351         heartbeat_info_dict = {
 352             'url': session_api_endpoint['url'] + '/' + session_response['data']['session']['id'] + '?_format=json&_method=PUT',
 353             'data': json.dumps(session_response['data']),
 354             # interval, convert milliseconds to seconds, then halve to make a buffer.
 355             'interval': float_or_none(session_api_data.get('heartbeatLifetime'), scale=3000),
 356             'ping': ping
 357         }
 358
 359         return info_dict, heartbeat_info_dict
 360
 361     def _extract_format_for_quality(self, video_id, audio_quality, video_quality, dmc_protocol):
 362
 363         if not audio_quality.get('isAvailable') or not video_quality.get('isAvailable'):
 364             return None
 365
 366         format_id = '-'.join(
 367             [remove_start(s['id'], 'archive_') for s in (video_quality, audio_quality)] + [dmc_protocol])
 368
 369         vid_qual_label = traverse_obj(video_quality, ('metadata', 'label'))
 370
 371         return {
 372             'url': 'niconico_dmc:%s/%s/%s' % (video_id, video_quality['id'], audio_quality['id']),
 373             'format_id': format_id,
 374             'format_note': join_nonempty('DMC', vid_qual_label, dmc_protocol.upper(), delim=' '),
 375             'ext': 'mp4',  # Session API are used in HTML5, which always serves mp4
 376             'acodec': 'aac',
 377             'vcodec': 'h264',
 378             **traverse_obj(audio_quality, ('metadata', {
 379                 'abr': ('bitrate', {functools.partial(float_or_none, scale=1000)}),
 380                 'asr': ('samplingRate', {int_or_none}),
 381             })),
 382             **traverse_obj(video_quality, ('metadata', {
 383                 'vbr': ('bitrate', {functools.partial(float_or_none, scale=1000)}),
 384                 'height': ('resolution', 'height', {int_or_none}),
 385                 'width': ('resolution', 'width', {int_or_none}),
 386             })),
 387             'quality': -2 if 'low' in video_quality['id'] else None,
 388             'protocol': 'niconico_dmc',
 389             'expected_protocol': dmc_protocol,  # XXX: This is not a documented field
 390             'http_headers': {
 391                 'Origin': 'https://www.nicovideo.jp',
 392                 'Referer': 'https://www.nicovideo.jp/watch/' + video_id,
 393             }
 394         }
 395
 396     def _yield_dmc_formats(self, api_data, video_id):
 397         dmc_data = traverse_obj(api_data, ('media', 'delivery', 'movie'))
 398         audios = traverse_obj(dmc_data, ('audios', ..., {dict}))
 399         videos = traverse_obj(dmc_data, ('videos', ..., {dict}))
 400         protocols = traverse_obj(dmc_data, ('session', 'protocols', ..., {str}))
 401         if not all((audios, videos, protocols)):
 402             return
 403
 404         for audio_quality, video_quality, protocol in itertools.product(audios, videos, protocols):
 405             if fmt := self._extract_format_for_quality(video_id, audio_quality, video_quality, protocol):
 406                 yield fmt
 407
 408     def _yield_dms_formats(self, api_data, video_id):
 409         fmt_filter = lambda _, v: v['isAvailable'] and v['id']
 410         videos = traverse_obj(api_data, ('media', 'domand', 'videos', fmt_filter))
 411         audios = traverse_obj(api_data, ('media', 'domand', 'audios', fmt_filter))
 412         access_key = traverse_obj(api_data, ('media', 'domand', 'accessRightKey', {str}))
 413         track_id = traverse_obj(api_data, ('client', 'watchTrackId', {str}))
 414         if not all((videos, audios, access_key, track_id)):
 415             return
 416
 417         dms_m3u8_url = self._download_json(
 418             f'https://nvapi.nicovideo.jp/v1/watch/{video_id}/access-rights/hls', video_id,
 419             data=json.dumps({
 420                 'outputs': list(itertools.product((v['id'] for v in videos), (a['id'] for a in audios)))
 421             }).encode(), query={'actionTrackId': track_id}, headers={
 422                 'x-access-right-key': access_key,
 423                 'x-frontend-id': 6,
 424                 'x-frontend-version': 0,
 425                 'x-request-with': 'https://www.nicovideo.jp',
 426             })['data']['contentUrl']
 427         # Getting all audio formats results in duplicate video formats which we filter out later
 428         dms_fmts = self._extract_m3u8_formats(dms_m3u8_url, video_id)
 429
 430         # m3u8 extraction does not provide audio bitrates, so extract from the API data and fix
 431         for audio_fmt in traverse_obj(dms_fmts, lambda _, v: v['vcodec'] == 'none'):
 432             yield {
 433                 **audio_fmt,
 434                 **traverse_obj(audios, (lambda _, v: audio_fmt['format_id'].startswith(v['id']), {
 435                     'format_id': ('id', {str}),
 436                     'abr': ('bitRate', {functools.partial(float_or_none, scale=1000)}),
 437                     'asr': ('samplingRate', {int_or_none}),
 438                 }), get_all=False),
 439                 'acodec': 'aac',
 440                 'ext': 'm4a',
 441             }
 442
 443         # Sort before removing dupes to keep the format dicts with the lowest tbr
 444         video_fmts = sorted((fmt for fmt in dms_fmts if fmt['vcodec'] != 'none'), key=lambda f: f['tbr'])
 445         self._remove_duplicate_formats(video_fmts)
 446         # Calculate the true vbr/tbr by subtracting the lowest abr
 447         min_abr = min(traverse_obj(audios, (..., 'bitRate', {float_or_none})), default=0) / 1000
 448         for video_fmt in video_fmts:
 449             video_fmt['tbr'] -= min_abr
 450             video_fmt['format_id'] = f'video-{video_fmt["tbr"]:.0f}'
 451             yield video_fmt
 452
 453     def _real_extract(self, url):
 454         video_id = self._match_id(url)
 455
 456         try:
 457             webpage, handle = self._download_webpage_handle(
 458                 'https://www.nicovideo.jp/watch/' + video_id, video_id)
 459             if video_id.startswith('so'):
 460                 video_id = self._match_id(handle.url)
 461
 462             api_data = self._parse_json(self._html_search_regex(
 463                 'data-api-data="([^"]+)"', webpage,
 464                 'API data', default='{}'), video_id)
 465         except ExtractorError as e:
 466             try:
 467                 api_data = self._download_json(
 468                     'https://www.nicovideo.jp/api/watch/v3/%s?_frontendId=6&_frontendVersion=0&actionTrackId=AAAAAAAAAA_%d' % (video_id, round(time.time() * 1000)), video_id,
 469                     note='Downloading API JSON', errnote='Unable to fetch data')['data']
 470             except ExtractorError:
 471                 if not isinstance(e.cause, HTTPError):
 472                     raise
 473                 webpage = e.cause.response.read().decode('utf-8', 'replace')
 474                 error_msg = self._html_search_regex(
 475                     r'(?s)<section\s+class="(?:(?:ErrorMessage|WatchExceptionPage-message)\s*)+">(.+?)</section>',
 476                     webpage, 'error reason', default=None)
 477                 if not error_msg:
 478                     raise
 479                 raise ExtractorError(clean_html(error_msg), expected=True)
 480
 481         club_joined = traverse_obj(api_data, ('channel', 'viewer', 'follow', 'isFollowed', {bool}))
 482         if club_joined is None:
 483             fail_msg = self._html_search_regex(
 484                 r'<p[^>]+\bclass="fail-message"[^>]*>(?P<msg>.+?)</p>',
 485                 webpage, 'fail message', default=None, group='msg')
 486             if fail_msg:
 487                 self.raise_login_required(clean_html(fail_msg), metadata_available=True)
 488         elif not club_joined:
 489             self.raise_login_required('This video is for members only', metadata_available=True)
 490
 491         # Start extracting information
 492         tags = None
 493         if webpage:
 494             # use og:video:tag (not logged in)
 495             og_video_tags = re.finditer(r'<meta\s+property="og:video:tag"\s*content="(.*?)">', webpage)
 496             tags = list(filter(None, (clean_html(x.group(1)) for x in og_video_tags)))
 497             if not tags:
 498                 # use keywords and split with comma (not logged in)
 499                 kwds = self._html_search_meta('keywords', webpage, default=None)
 500                 if kwds:
 501                     tags = [x for x in kwds.split(',') if x]
 502         if not tags:
 503             # find in json (logged in)
 504             tags = traverse_obj(api_data, ('tag', 'items', ..., 'name'))
 505
 506         thumb_prefs = qualities(['url', 'middleUrl', 'largeUrl', 'player', 'ogp'])
 507
 508         def get_video_info(*items, get_first=True, **kwargs):
 509             return traverse_obj(api_data, ('video', *items), get_all=not get_first, **kwargs)
 510
 511         return {
 512             'id': video_id,
 513             '_api_data': api_data,
 514             'title': get_video_info(('originalTitle', 'title')) or self._og_search_title(webpage, default=None),
 515             'formats': [*self._yield_dmc_formats(api_data, video_id),
 516                         *self._yield_dms_formats(api_data, video_id)],
 517             'thumbnails': [{
 518                 'id': key,
 519                 'url': url,
 520                 'ext': 'jpg',
 521                 'preference': thumb_prefs(key),
 522                 **parse_resolution(url, lenient=True),
 523             } for key, url in (get_video_info('thumbnail') or {}).items() if url],
 524             'description': clean_html(get_video_info('description')),
 525             'uploader': traverse_obj(api_data, ('owner', 'nickname'), ('channel', 'name'), ('community', 'name')),
 526             'uploader_id': str_or_none(traverse_obj(api_data, ('owner', 'id'), ('channel', 'id'), ('community', 'id'))),
 527             'timestamp': parse_iso8601(get_video_info('registeredAt')) or parse_iso8601(
 528                 self._html_search_meta('video:release_date', webpage, 'date published', default=None)),
 529             'channel': traverse_obj(api_data, ('channel', 'name'), ('community', 'name')),
 530             'channel_id': traverse_obj(api_data, ('channel', 'id'), ('community', 'id')),
 531             'view_count': int_or_none(get_video_info('count', 'view')),
 532             'tags': tags,
 533             'genre': traverse_obj(api_data, ('genre', 'label'), ('genre', 'key')),
 534             'comment_count': get_video_info('count', 'comment', expected_type=int),
 535             'duration': (
 536                 parse_duration(self._html_search_meta('video:duration', webpage, 'video duration', default=None))
 537                 or get_video_info('duration')),
 538             'webpage_url': url_or_none(url) or f'https://www.nicovideo.jp/watch/{video_id}',
 539             'subtitles': self.extract_subtitles(video_id, api_data),
 540         }
 541
 542     def _get_subtitles(self, video_id, api_data):
 543         comments_info = traverse_obj(api_data, ('comment', 'nvComment', {dict})) or {}
 544         if not comments_info.get('server'):
 545             return
 546
 547         danmaku = traverse_obj(self._download_json(
 548             f'{comments_info["server"]}/v1/threads', video_id, data=json.dumps({
 549                 'additionals': {},
 550                 'params': comments_info.get('params'),
 551                 'threadKey': comments_info.get('threadKey'),
 552             }).encode(), fatal=False,
 553             headers={
 554                 'Referer': 'https://www.nicovideo.jp/',
 555                 'Origin': 'https://www.nicovideo.jp',
 556                 'Content-Type': 'text/plain;charset=UTF-8',
 557                 'x-client-os-type': 'others',
 558                 'x-frontend-id': '6',
 559                 'x-frontend-version': '0',
 560             },
 561             note='Downloading comments', errnote='Failed to download comments'),
 562             ('data', 'threads', ..., 'comments', ...))
 563
 564         return {
 565             'comments': [{
 566                 'ext': 'json',
 567                 'data': json.dumps(danmaku),
 568             }],
 569         }
 570
 571
 572 class NiconicoPlaylistBaseIE(InfoExtractor):
 573     _PAGE_SIZE = 100
 574
 575     _API_HEADERS = {
 576         'X-Frontend-ID': '6',
 577         'X-Frontend-Version': '0',
 578         'X-Niconico-Language': 'en-us'
 579     }
 580
 581     def _call_api(self, list_id, resource, query):
 582         raise NotImplementedError('Must be implemented in subclasses')
 583
 584     @staticmethod
 585     def _parse_owner(item):
 586         return {
 587             'uploader': traverse_obj(item, ('owner', 'name')),
 588             'uploader_id': traverse_obj(item, ('owner', 'id')),
 589         }
 590
 591     def _fetch_page(self, list_id, page):
 592         page += 1
 593         resp = self._call_api(list_id, 'page %d' % page, {
 594             'page': page,
 595             'pageSize': self._PAGE_SIZE,
 596         })
 597         # this is needed to support both mylist and user
 598         for video in traverse_obj(resp, ('items', ..., ('video', None))) or []:
 599             video_id = video.get('id')
 600             if not video_id:
 601                 # skip {"video": {"id": "blablabla", ...}}
 602                 continue
 603             count = video.get('count') or {}
 604             get_count = lambda x: int_or_none(count.get(x))
 605             yield {
 606                 '_type': 'url',
 607                 'id': video_id,
 608                 'title': video.get('title'),
 609                 'url': f'https://www.nicovideo.jp/watch/{video_id}',
 610                 'description': video.get('shortDescription'),
 611                 'duration': int_or_none(video.get('duration')),
 612                 'view_count': get_count('view'),
 613                 'comment_count': get_count('comment'),
 614                 'thumbnail': traverse_obj(video, ('thumbnail', ('nHdUrl', 'largeUrl', 'listingUrl', 'url'))),
 615                 'ie_key': NiconicoIE.ie_key(),
 616                 **self._parse_owner(video),
 617             }
 618
 619     def _entries(self, list_id):
 620         return OnDemandPagedList(functools.partial(self._fetch_page, list_id), self._PAGE_SIZE)
 621
 622
 623 class NiconicoPlaylistIE(NiconicoPlaylistBaseIE):
 624     IE_NAME = 'niconico:playlist'
 625     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp|nico\.ms)/(?:user/\d+/)?(?:my/)?mylist/(?:#/)?(?P<id>\d+)'
 626
 627     _TESTS = [{
 628         'url': 'http://www.nicovideo.jp/mylist/27411728',
 629         'info_dict': {
 630             'id': '27411728',
 631             'title': 'AKB48のオールナイトニッポン',
 632             'description': 'md5:d89694c5ded4b6c693dea2db6e41aa08',
 633             'uploader': 'のっく',
 634             'uploader_id': '805442',
 635         },
 636         'playlist_mincount': 291,
 637     }, {
 638         'url': 'https://www.nicovideo.jp/user/805442/mylist/27411728',
 639         'only_matching': True,
 640     }, {
 641         'url': 'https://www.nicovideo.jp/my/mylist/#/68048635',
 642         'only_matching': True,
 643     }]
 644
 645     def _call_api(self, list_id, resource, query):
 646         return self._download_json(
 647             f'https://nvapi.nicovideo.jp/v2/mylists/{list_id}', list_id,
 648             f'Downloading {resource}', query=query,
 649             headers=self._API_HEADERS)['data']['mylist']
 650
 651     def _real_extract(self, url):
 652         list_id = self._match_id(url)
 653         mylist = self._call_api(list_id, 'list', {
 654             'pageSize': 1,
 655         })
 656         return self.playlist_result(
 657             self._entries(list_id), list_id,
 658             mylist.get('name'), mylist.get('description'), **self._parse_owner(mylist))
 659
 660
 661 class NiconicoSeriesIE(InfoExtractor):
 662     IE_NAME = 'niconico:series'
 663     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp(?:/user/\d+)?|nico\.ms)/series/(?P<id>\d+)'
 664
 665     _TESTS = [{
 666         'url': 'https://www.nicovideo.jp/user/44113208/series/110226',
 667         'info_dict': {
 668             'id': '110226',
 669             'title': 'ご立派ァ！のシリーズ',
 670         },
 671         'playlist_mincount': 10,
 672     }, {
 673         'url': 'https://www.nicovideo.jp/series/12312/',
 674         'info_dict': {
 675             'id': '12312',
 676             'title': 'バトルスピリッツ　お勧めカード紹介(調整中)',
 677         },
 678         'playlist_mincount': 103,
 679     }, {
 680         'url': 'https://nico.ms/series/203559',
 681         'only_matching': True,
 682     }]
 683
 684     def _real_extract(self, url):
 685         list_id = self._match_id(url)
 686         webpage = self._download_webpage(url, list_id)
 687
 688         title = self._search_regex(
 689             (r'<title>「(.+)（全',
 690              r'<div class="TwitterShareButton"\s+data-text="(.+)\s+https:'),
 691             webpage, 'title', fatal=False)
 692         if title:
 693             title = unescapeHTML(title)
 694         json_data = next(self._yield_json_ld(webpage, None, fatal=False))
 695         return self.playlist_from_matches(
 696             traverse_obj(json_data, ('itemListElement', ..., 'url')), list_id, title, ie=NiconicoIE)
 697
 698
 699 class NiconicoHistoryIE(NiconicoPlaylistBaseIE):
 700     IE_NAME = 'niconico:history'
 701     IE_DESC = 'NicoNico user history or likes. Requires cookies.'
 702     _VALID_URL = r'https?://(?:www\.|sp\.)?nicovideo\.jp/my/(?P<id>history(?:/like)?)'
 703
 704     _TESTS = [{
 705         'note': 'PC page, with /video',
 706         'url': 'https://www.nicovideo.jp/my/history/video',
 707         'only_matching': True,
 708     }, {
 709         'note': 'PC page, without /video',
 710         'url': 'https://www.nicovideo.jp/my/history',
 711         'only_matching': True,
 712     }, {
 713         'note': 'mobile page, with /video',
 714         'url': 'https://sp.nicovideo.jp/my/history/video',
 715         'only_matching': True,
 716     }, {
 717         'note': 'mobile page, without /video',
 718         'url': 'https://sp.nicovideo.jp/my/history',
 719         'only_matching': True,
 720     }, {
 721         'note': 'PC page',
 722         'url': 'https://www.nicovideo.jp/my/history/like',
 723         'only_matching': True,
 724     }, {
 725         'note': 'Mobile page',
 726         'url': 'https://sp.nicovideo.jp/my/history/like',
 727         'only_matching': True,
 728     }]
 729
 730     def _call_api(self, list_id, resource, query):
 731         path = 'likes' if list_id == 'history/like' else 'watch/history'
 732         return self._download_json(
 733             f'https://nvapi.nicovideo.jp/v1/users/me/{path}', list_id,
 734             f'Downloading {resource}', query=query, headers=self._API_HEADERS)['data']
 735
 736     def _real_extract(self, url):
 737         list_id = self._match_id(url)
 738         try:
 739             mylist = self._call_api(list_id, 'list', {'pageSize': 1})
 740         except ExtractorError as e:
 741             if isinstance(e.cause, HTTPError) and e.cause.status == 401:
 742                 self.raise_login_required('You have to be logged in to get your history')
 743             raise
 744         return self.playlist_result(self._entries(list_id), list_id, **self._parse_owner(mylist))
 745
 746
 747 class NicovideoSearchBaseIE(InfoExtractor):
 748     _SEARCH_TYPE = 'search'
 749
 750     def _entries(self, url, item_id, query=None, note='Downloading page %(page)s'):
 751         query = query or {}
 752         pages = [query['page']] if 'page' in query else itertools.count(1)
 753         for page_num in pages:
 754             query['page'] = str(page_num)
 755             webpage = self._download_webpage(url, item_id, query=query, note=note % {'page': page_num})
 756             results = re.findall(r'(?<=data-video-id=)["\']?(?P<videoid>.*?)(?=["\'])', webpage)
 757             for item in results:
 758                 yield self.url_result(f'https://www.nicovideo.jp/watch/{item}', 'Niconico', item)
 759             if not results:
 760                 break
 761
 762     def _search_results(self, query):
 763         return self._entries(
 764             self._proto_relative_url(f'//www.nicovideo.jp/{self._SEARCH_TYPE}/{query}'), query)
 765
 766
 767 class NicovideoSearchIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 768     IE_DESC = 'Nico video search'
 769     IE_NAME = 'nicovideo:search'
 770     _SEARCH_KEY = 'nicosearch'
 771
 772
 773 class NicovideoSearchURLIE(NicovideoSearchBaseIE):
 774     IE_NAME = f'{NicovideoSearchIE.IE_NAME}_url'
 775     IE_DESC = 'Nico video search URLs'
 776     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/search/(?P<id>[^?#&]+)?'
 777     _TESTS = [{
 778         'url': 'http://www.nicovideo.jp/search/sm9',
 779         'info_dict': {
 780             'id': 'sm9',
 781             'title': 'sm9'
 782         },
 783         'playlist_mincount': 40,
 784     }, {
 785         'url': 'https://www.nicovideo.jp/search/sm9?sort=h&order=d&end=2020-12-31&start=2020-01-01',
 786         'info_dict': {
 787             'id': 'sm9',
 788             'title': 'sm9'
 789         },
 790         'playlist_count': 31,
 791     }]
 792
 793     def _real_extract(self, url):
 794         query = self._match_id(url)
 795         return self.playlist_result(self._entries(url, query), query, query)
 796
 797
 798 class NicovideoSearchDateIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 799     IE_DESC = 'Nico video search, newest first'
 800     IE_NAME = f'{NicovideoSearchIE.IE_NAME}:date'
 801     _SEARCH_KEY = 'nicosearchdate'
 802     _TESTS = [{
 803         'url': 'nicosearchdateall:a',
 804         'info_dict': {
 805             'id': 'a',
 806             'title': 'a'
 807         },
 808         'playlist_mincount': 1610,
 809     }]
 810
 811     _START_DATE = datetime.date(2007, 1, 1)
 812     _RESULTS_PER_PAGE = 32
 813     _MAX_PAGES = 50
 814
 815     def _entries(self, url, item_id, start_date=None, end_date=None):
 816         start_date, end_date = start_date or self._START_DATE, end_date or datetime.datetime.now().date()
 817
 818         # If the last page has a full page of videos, we need to break down the query interval further
 819         last_page_len = len(list(self._get_entries_for_date(
 820             url, item_id, start_date, end_date, self._MAX_PAGES,
 821             note=f'Checking number of videos from {start_date} to {end_date}')))
 822         if (last_page_len == self._RESULTS_PER_PAGE and start_date != end_date):
 823             midpoint = start_date + ((end_date - start_date) // 2)
 824             yield from self._entries(url, item_id, midpoint, end_date)
 825             yield from self._entries(url, item_id, start_date, midpoint)
 826         else:
 827             self.to_screen(f'{item_id}: Downloading results from {start_date} to {end_date}')
 828             yield from self._get_entries_for_date(
 829                 url, item_id, start_date, end_date, note='    Downloading page %(page)s')
 830
 831     def _get_entries_for_date(self, url, item_id, start_date, end_date=None, page_num=None, note=None):
 832         query = {
 833             'start': str(start_date),
 834             'end': str(end_date or start_date),
 835             'sort': 'f',
 836             'order': 'd',
 837         }
 838         if page_num:
 839             query['page'] = str(page_num)
 840
 841         yield from super()._entries(url, item_id, query=query, note=note)
 842
 843
 844 class NicovideoTagURLIE(NicovideoSearchBaseIE):
 845     IE_NAME = 'niconico:tag'
 846     IE_DESC = 'NicoNico video tag URLs'
 847     _SEARCH_TYPE = 'tag'
 848     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/tag/(?P<id>[^?#&]+)?'
 849     _TESTS = [{
 850         'url': 'https://www.nicovideo.jp/tag/ドキュメンタリー淫夢',
 851         'info_dict': {
 852             'id': 'ドキュメンタリー淫夢',
 853             'title': 'ドキュメンタリー淫夢'
 854         },
 855         'playlist_mincount': 400,
 856     }]
 857
 858     def _real_extract(self, url):
 859         query = self._match_id(url)
 860         return self.playlist_result(self._entries(url, query), query, query)
 861
 862
 863 class NiconicoUserIE(InfoExtractor):
 864     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/user/(?P<id>\d+)/?(?:$|[#?])'
 865     _TEST = {
 866         'url': 'https://www.nicovideo.jp/user/419948',
 867         'info_dict': {
 868             'id': '419948',
 869         },
 870         'playlist_mincount': 101,
 871     }
 872     _API_URL = "https://nvapi.nicovideo.jp/v1/users/%s/videos?sortKey=registeredAt&sortOrder=desc&pageSize=%s&page=%s"
 873     _PAGE_SIZE = 100
 874
 875     _API_HEADERS = {
 876         'X-Frontend-ID': '6',
 877         'X-Frontend-Version': '0'
 878     }
 879
 880     def _entries(self, list_id):
 881         total_count = 1
 882         count = page_num = 0
 883         while count < total_count:
 884             json_parsed = self._download_json(
 885                 self._API_URL % (list_id, self._PAGE_SIZE, page_num + 1), list_id,
 886                 headers=self._API_HEADERS,
 887                 note='Downloading JSON metadata%s' % (' page %d' % page_num if page_num else ''))
 888             if not page_num:
 889                 total_count = int_or_none(json_parsed['data'].get('totalCount'))
 890             for entry in json_parsed["data"]["items"]:
 891                 count += 1
 892                 yield self.url_result('https://www.nicovideo.jp/watch/%s' % entry['id'])
 893             page_num += 1
 894
 895     def _real_extract(self, url):
 896         list_id = self._match_id(url)
 897         return self.playlist_result(self._entries(list_id), list_id, ie=NiconicoIE.ie_key())
 898
 899
 900 class NiconicoLiveIE(InfoExtractor):
 901     IE_NAME = 'niconico:live'
 902     IE_DESC = 'ニコニコ生放送'
 903     _VALID_URL = r'https?://(?:sp\.)?live2?\.nicovideo\.jp/(?:watch|gate)/(?P<id>lv\d+)'
 904     _TESTS = [{
 905         'note': 'this test case includes invisible characters for title, pasting them as-is',
 906         'url': 'https://live.nicovideo.jp/watch/lv339533123',
 907         'info_dict': {
 908             'id': 'lv339533123',
 909             'title': '激辛ペヤング食べます‪( ;ᯅ; )‬（歌枠オーディション参加中）',
 910             'view_count': 1526,
 911             'comment_count': 1772,
 912             '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',
 913             'uploader': 'もか',
 914             'channel': 'ゲストさんのコミュニティ',
 915             'channel_id': 'co5776900',
 916             'channel_url': 'https://com.nicovideo.jp/community/co5776900',
 917             'timestamp': 1670677328,
 918             'is_live': True,
 919         },
 920         'skip': 'livestream',
 921     }, {
 922         'url': 'https://live2.nicovideo.jp/watch/lv339533123',
 923         'only_matching': True,
 924     }, {
 925         'url': 'https://sp.live.nicovideo.jp/watch/lv339533123',
 926         'only_matching': True,
 927     }, {
 928         'url': 'https://sp.live2.nicovideo.jp/watch/lv339533123',
 929         'only_matching': True,
 930     }]
 931
 932     _KNOWN_LATENCY = ('high', 'low')
 933
 934     def _real_extract(self, url):
 935         video_id = self._match_id(url)
 936         webpage, urlh = self._download_webpage_handle(f'https://live.nicovideo.jp/watch/{video_id}', video_id)
 937
 938         embedded_data = self._parse_json(unescapeHTML(self._search_regex(
 939             r'<script\s+id="embedded-data"\s*data-props="(.+?)"', webpage, 'embedded data')), video_id)
 940
 941         ws_url = traverse_obj(embedded_data, ('site', 'relive', 'webSocketUrl'))
 942         if not ws_url:
 943             raise ExtractorError('The live hasn\'t started yet or already ended.', expected=True)
 944         ws_url = update_url_query(ws_url, {
 945             'frontend_id': traverse_obj(embedded_data, ('site', 'frontendId')) or '9',
 946         })
 947
 948         hostname = remove_start(urlparse(urlh.url).hostname, 'sp.')
 949         latency = try_get(self._configuration_arg('latency'), lambda x: x[0])
 950         if latency not in self._KNOWN_LATENCY:
 951             latency = 'high'
 952
 953         ws = self._request_webpage(
 954             Request(ws_url, headers={'Origin': f'https://{hostname}'}),
 955             video_id=video_id, note='Connecting to WebSocket server')
 956
 957         self.write_debug('[debug] Sending HLS server request')
 958         ws.send(json.dumps({
 959             'type': 'startWatching',
 960             'data': {
 961                 'stream': {
 962                     'quality': 'abr',
 963                     'protocol': 'hls+fmp4',
 964                     'latency': latency,
 965                     'chasePlay': False
 966                 },
 967                 'room': {
 968                     'protocol': 'webSocket',
 969                     'commentable': True
 970                 },
 971                 'reconnect': False,
 972             }
 973         }))
 974
 975         while True:
 976             recv = ws.recv()
 977             if not recv:
 978                 continue
 979             data = json.loads(recv)
 980             if not isinstance(data, dict):
 981                 continue
 982             if data.get('type') == 'stream':
 983                 m3u8_url = data['data']['uri']
 984                 qualities = data['data']['availableQualities']
 985                 break
 986             elif data.get('type') == 'disconnect':
 987                 self.write_debug(recv)
 988                 raise ExtractorError('Disconnected at middle of extraction')
 989             elif data.get('type') == 'error':
 990                 self.write_debug(recv)
 991                 message = traverse_obj(data, ('body', 'code')) or recv
 992                 raise ExtractorError(message)
 993             elif self.get_param('verbose', False):
 994                 if len(recv) > 100:
 995                     recv = recv[:100] + '...'
 996                 self.write_debug('Server said: %s' % recv)
 997
 998         title = traverse_obj(embedded_data, ('program', 'title')) or self._html_search_meta(
 999             ('og:title', 'twitter:title'), webpage, 'live title', fatal=False)
1000
1001         raw_thumbs = traverse_obj(embedded_data, ('program', 'thumbnail')) or {}
1002         thumbnails = []
1003         for name, value in raw_thumbs.items():
1004             if not isinstance(value, dict):
1005                 thumbnails.append({
1006                     'id': name,
1007                     'url': value,
1008                     **parse_resolution(value, lenient=True),
1009                 })
1010                 continue
1011
1012             for k, img_url in value.items():
1013                 res = parse_resolution(k, lenient=True) or parse_resolution(img_url, lenient=True)
1014                 width, height = res.get('width'), res.get('height')
1015
1016                 thumbnails.append({
1017                     'id': f'{name}_{width}x{height}',
1018                     'url': img_url,
1019                     **res,
1020                 })
1021
1022         formats = self._extract_m3u8_formats(m3u8_url, video_id, ext='mp4', live=True)
1023         for fmt, q in zip(formats, reversed(qualities[1:])):
1024             fmt.update({
1025                 'format_id': q,
1026                 'protocol': 'niconico_live',
1027                 'ws': ws,
1028                 'video_id': video_id,
1029                 'live_latency': latency,
1030                 'origin': hostname,
1031             })
1032
1033         return {
1034             'id': video_id,
1035             'title': title,
1036             **traverse_obj(embedded_data, {
1037                 'view_count': ('program', 'statistics', 'watchCount'),
1038                 'comment_count': ('program', 'statistics', 'commentCount'),
1039                 'uploader': ('program', 'supplier', 'name'),
1040                 'channel': ('socialGroup', 'name'),
1041                 'channel_id': ('socialGroup', 'id'),
1042                 'channel_url': ('socialGroup', 'socialGroupPageUrl'),
1043             }),
1044             'description': clean_html(traverse_obj(embedded_data, ('program', 'description'))),
1045             'timestamp': int_or_none(traverse_obj(embedded_data, ('program', 'openTime'))),
1046             'is_live': True,
1047             'thumbnails': thumbnails,
1048             'formats': formats,
1049         }