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         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  44         'info_dict': {
  45             'id': 'sm22312215',
  46             'ext': 'mp4',
  47             'title': 'Big Buck Bunny',
  48             'thumbnail': r're:https?://.*',
  49             'uploader': 'takuya0301',
  50             'uploader_id': '2698420',
  51             'upload_date': '20131123',
  52             'timestamp': int,  # timestamp is unstable
  53             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  54             'duration': 33,
  55             'view_count': int,
  56             'comment_count': int,
  57             'genres': ['未設定'],
  58             'tags': [],
  59             'expected_protocol': str,
  60         },
  61     }, {
  62         # File downloaded with and without credentials are different, so omit
  63         # the md5 field
  64         'url': 'http://www.nicovideo.jp/watch/nm14296458',
  65         'info_dict': {
  66             'id': 'nm14296458',
  67             'ext': 'mp4',
  68             'title': '【Kagamine Rin】Dance on media【Original】take2!',
  69             'description': 'md5:9368f2b1f4178de64f2602c2f3d6cbf5',
  70             'thumbnail': r're:https?://.*',
  71             'uploader': 'りょうた',
  72             'uploader_id': '18822557',
  73             'upload_date': '20110429',
  74             'timestamp': 1304065916,
  75             'duration': 208.0,
  76             'comment_count': int,
  77             'view_count': int,
  78             'genres': ['音楽・サウンド'],
  79             'tags': ['Translation_Request', 'Kagamine_Rin', 'Rin_Original'],
  80             'expected_protocol': str,
  81         },
  82     }, {
  83         # 'video exists but is marked as "deleted"
  84         # md5 is unstable
  85         'url': 'http://www.nicovideo.jp/watch/sm10000',
  86         'info_dict': {
  87             'id': 'sm10000',
  88             'ext': 'unknown_video',
  89             'description': 'deleted',
  90             'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞',
  91             'thumbnail': r're:https?://.*',
  92             'upload_date': '20071224',
  93             'timestamp': int,  # timestamp field has different value if logged in
  94             'duration': 304,
  95             'view_count': int,
  96         },
  97         'skip': 'Requires an account',
  98     }, {
  99         'url': 'http://www.nicovideo.jp/watch/so22543406',
 100         'info_dict': {
 101             'id': '1388129933',
 102             'ext': 'mp4',
 103             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
 104             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
 105             'thumbnail': r're:https?://.*',
 106             'timestamp': 1388851200,
 107             'upload_date': '20140104',
 108             'uploader': 'アニメロチャンネル',
 109             'uploader_id': '312',
 110         },
 111         'skip': 'The viewing period of the video you were searching for has expired.',
 112     }, {
 113         # video not available via `getflv`; "old" HTML5 video
 114         'url': 'http://www.nicovideo.jp/watch/sm1151009',
 115         'md5': 'f95a3d259172667b293530cc2e41ebda',
 116         'info_dict': {
 117             'id': 'sm1151009',
 118             'ext': 'mp4',
 119             '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）',
 120             'description': 'md5:f95a3d259172667b293530cc2e41ebda',
 121             'thumbnail': r're:https?://.*',
 122             'duration': 184,
 123             'timestamp': 1190835883,
 124             'upload_date': '20070926',
 125             'uploader': 'denden2',
 126             'uploader_id': '1392194',
 127             'view_count': int,
 128             'comment_count': int,
 129             'genres': ['ゲーム'],
 130             'tags': [],
 131             'expected_protocol': str,
 132         },
 133     }, {
 134         # "New" HTML5 video
 135         # md5 is unstable
 136         'url': 'http://www.nicovideo.jp/watch/sm31464864',
 137         'info_dict': {
 138             'id': 'sm31464864',
 139             'ext': 'mp4',
 140             'title': '新作TVアニメ「戦姫絶唱シンフォギアAXZ」PV 最高画質',
 141             'description': 'md5:e52974af9a96e739196b2c1ca72b5feb',
 142             'timestamp': 1498481660,
 143             'upload_date': '20170626',
 144             'uploader': 'no-namamae',
 145             'uploader_id': '40826363',
 146             'thumbnail': r're:https?://.*',
 147             'duration': 198,
 148             'view_count': int,
 149             'comment_count': int,
 150             'genres': ['アニメ'],
 151             'tags': [],
 152             'expected_protocol': str,
 153         },
 154     }, {
 155         # Video without owner
 156         'url': 'http://www.nicovideo.jp/watch/sm18238488',
 157         'md5': 'd265680a1f92bdcbbd2a507fc9e78a9e',
 158         'info_dict': {
 159             'id': 'sm18238488',
 160             'ext': 'mp4',
 161             'title': '【実写版】ミュータントタートルズ',
 162             'description': 'md5:15df8988e47a86f9e978af2064bf6d8e',
 163             'timestamp': 1341128008,
 164             'upload_date': '20120701',
 165             'thumbnail': r're:https?://.*',
 166             'duration': 5271,
 167             'view_count': int,
 168             'comment_count': int,
 169             'genres': ['エンターテイメント'],
 170             'tags': [],
 171             'expected_protocol': str,
 172         },
 173     }, {
 174         'url': 'http://sp.nicovideo.jp/watch/sm28964488?ss_pos=1&cp_in=wt_tg',
 175         'only_matching': True,
 176     }, {
 177         'note': 'a video that is only served as an ENCRYPTED HLS.',
 178         'url': 'https://www.nicovideo.jp/watch/so38016254',
 179         'only_matching': True,
 180     }]
 181
 182     _VALID_URL = r'https?://(?:(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch|nico\.ms)/(?P<id>(?:[a-z]{2})?[0-9]+)'
 183     _NETRC_MACHINE = 'niconico'
 184     _API_HEADERS = {
 185         'X-Frontend-ID': '6',
 186         'X-Frontend-Version': '0',
 187         'X-Niconico-Language': 'en-us',
 188         'Referer': 'https://www.nicovideo.jp/',
 189         'Origin': 'https://www.nicovideo.jp',
 190     }
 191
 192     def _perform_login(self, username, password):
 193         login_ok = True
 194         login_form_strs = {
 195             'mail_tel': username,
 196             'password': password,
 197         }
 198         self._request_webpage(
 199             'https://account.nicovideo.jp/login', None,
 200             note='Acquiring Login session')
 201         page = self._download_webpage(
 202             'https://account.nicovideo.jp/login/redirector?show_button_twitter=1&site=niconico&show_button_facebook=1', None,
 203             note='Logging in', errnote='Unable to log in',
 204             data=urlencode_postdata(login_form_strs),
 205             headers={
 206                 'Referer': 'https://account.nicovideo.jp/login',
 207                 'Content-Type': 'application/x-www-form-urlencoded',
 208             })
 209         if 'oneTimePw' in page:
 210             post_url = self._search_regex(
 211                 r'<form[^>]+action=(["\'])(?P<url>.+?)\1', page, 'post url', group='url')
 212             page = self._download_webpage(
 213                 urljoin('https://account.nicovideo.jp', post_url), None,
 214                 note='Performing MFA', errnote='Unable to complete MFA',
 215                 data=urlencode_postdata({
 216                     'otp': self._get_tfa_info('6 digits code')
 217                 }), headers={
 218                     'Content-Type': 'application/x-www-form-urlencoded',
 219                 })
 220             if 'oneTimePw' in page or 'formError' in page:
 221                 err_msg = self._html_search_regex(
 222                     r'formError["\']+>(.*?)</div>', page, 'form_error',
 223                     default='There\'s an error but the message can\'t be parsed.',
 224                     flags=re.DOTALL)
 225                 self.report_warning(f'Unable to log in: MFA challenge failed, "{err_msg}"')
 226                 return False
 227         login_ok = 'class="notice error"' not in page
 228         if not login_ok:
 229             self.report_warning('Unable to log in: bad username or password')
 230         return login_ok
 231
 232     def _get_heartbeat_info(self, info_dict):
 233         video_id, video_src_id, audio_src_id = info_dict['url'].split(':')[1].split('/')
 234         dmc_protocol = info_dict['expected_protocol']
 235
 236         api_data = (
 237             info_dict.get('_api_data')
 238             or self._parse_json(
 239                 self._html_search_regex(
 240                     'data-api-data="([^"]+)"',
 241                     self._download_webpage('https://www.nicovideo.jp/watch/' + video_id, video_id),
 242                     'API data', default='{}'),
 243                 video_id))
 244
 245         session_api_data = try_get(api_data, lambda x: x['media']['delivery']['movie']['session'])
 246         session_api_endpoint = try_get(session_api_data, lambda x: x['urls'][0])
 247
 248         def ping():
 249             tracking_id = traverse_obj(api_data, ('media', 'delivery', 'trackingId'))
 250             if tracking_id:
 251                 tracking_url = update_url_query('https://nvapi.nicovideo.jp/v1/2ab0cbaa/watch', {'t': tracking_id})
 252                 watch_request_response = self._download_json(
 253                     tracking_url, video_id,
 254                     note='Acquiring permission for downloading video', fatal=False,
 255                     headers=self._API_HEADERS)
 256                 if traverse_obj(watch_request_response, ('meta', 'status')) != 200:
 257                     self.report_warning('Failed to acquire permission for playing video. Video download may fail.')
 258
 259         yesno = lambda x: 'yes' if x else 'no'
 260
 261         if dmc_protocol == 'http':
 262             protocol = 'http'
 263             protocol_parameters = {
 264                 'http_output_download_parameters': {
 265                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 266                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 267                 }
 268             }
 269         elif dmc_protocol == 'hls':
 270             protocol = 'm3u8'
 271             segment_duration = try_get(self._configuration_arg('segment_duration'), lambda x: int(x[0])) or 6000
 272             parsed_token = self._parse_json(session_api_data['token'], video_id)
 273             encryption = traverse_obj(api_data, ('media', 'delivery', 'encryption'))
 274             protocol_parameters = {
 275                 'hls_parameters': {
 276                     'segment_duration': segment_duration,
 277                     'transfer_preset': '',
 278                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 279                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 280                 }
 281             }
 282             if 'hls_encryption' in parsed_token and encryption:
 283                 protocol_parameters['hls_parameters']['encryption'] = {
 284                     parsed_token['hls_encryption']: {
 285                         'encrypted_key': encryption['encryptedKey'],
 286                         'key_uri': encryption['keyUri'],
 287                     }
 288                 }
 289             else:
 290                 protocol = 'm3u8_native'
 291         else:
 292             raise ExtractorError(f'Unsupported DMC protocol: {dmc_protocol}')
 293
 294         session_response = self._download_json(
 295             session_api_endpoint['url'], video_id,
 296             query={'_format': 'json'},
 297             headers={'Content-Type': 'application/json'},
 298             note='Downloading JSON metadata for %s' % info_dict['format_id'],
 299             data=json.dumps({
 300                 'session': {
 301                     'client_info': {
 302                         'player_id': session_api_data.get('playerId'),
 303                     },
 304                     'content_auth': {
 305                         'auth_type': try_get(session_api_data, lambda x: x['authTypes'][session_api_data['protocols'][0]]),
 306                         'content_key_timeout': session_api_data.get('contentKeyTimeout'),
 307                         'service_id': 'nicovideo',
 308                         'service_user_id': session_api_data.get('serviceUserId')
 309                     },
 310                     'content_id': session_api_data.get('contentId'),
 311                     'content_src_id_sets': [{
 312                         'content_src_ids': [{
 313                             'src_id_to_mux': {
 314                                 'audio_src_ids': [audio_src_id],
 315                                 'video_src_ids': [video_src_id],
 316                             }
 317                         }]
 318                     }],
 319                     'content_type': 'movie',
 320                     'content_uri': '',
 321                     'keep_method': {
 322                         'heartbeat': {
 323                             'lifetime': session_api_data.get('heartbeatLifetime')
 324                         }
 325                     },
 326                     'priority': session_api_data['priority'],
 327                     'protocol': {
 328                         'name': 'http',
 329                         'parameters': {
 330                             'http_parameters': {
 331                                 'parameters': protocol_parameters
 332                             }
 333                         }
 334                     },
 335                     'recipe_id': session_api_data.get('recipeId'),
 336                     'session_operation_auth': {
 337                         'session_operation_auth_by_signature': {
 338                             'signature': session_api_data.get('signature'),
 339                             'token': session_api_data.get('token'),
 340                         }
 341                     },
 342                     'timing_constraint': 'unlimited'
 343                 }
 344             }).encode())
 345
 346         info_dict['url'] = session_response['data']['session']['content_uri']
 347         info_dict['protocol'] = protocol
 348
 349         # get heartbeat info
 350         heartbeat_info_dict = {
 351             'url': session_api_endpoint['url'] + '/' + session_response['data']['session']['id'] + '?_format=json&_method=PUT',
 352             'data': json.dumps(session_response['data']),
 353             # interval, convert milliseconds to seconds, then halve to make a buffer.
 354             'interval': float_or_none(session_api_data.get('heartbeatLifetime'), scale=3000),
 355             'ping': ping
 356         }
 357
 358         return info_dict, heartbeat_info_dict
 359
 360     def _extract_format_for_quality(self, video_id, audio_quality, video_quality, dmc_protocol):
 361
 362         if not audio_quality.get('isAvailable') or not video_quality.get('isAvailable'):
 363             return None
 364
 365         format_id = '-'.join(
 366             [remove_start(s['id'], 'archive_') for s in (video_quality, audio_quality)] + [dmc_protocol])
 367
 368         vid_qual_label = traverse_obj(video_quality, ('metadata', 'label'))
 369
 370         return {
 371             'url': 'niconico_dmc:%s/%s/%s' % (video_id, video_quality['id'], audio_quality['id']),
 372             'format_id': format_id,
 373             'format_note': join_nonempty('DMC', vid_qual_label, dmc_protocol.upper(), delim=' '),
 374             'ext': 'mp4',  # Session API are used in HTML5, which always serves mp4
 375             'acodec': 'aac',
 376             'vcodec': 'h264',
 377             **traverse_obj(audio_quality, ('metadata', {
 378                 'abr': ('bitrate', {functools.partial(float_or_none, scale=1000)}),
 379                 'asr': ('samplingRate', {int_or_none}),
 380             })),
 381             **traverse_obj(video_quality, ('metadata', {
 382                 'vbr': ('bitrate', {functools.partial(float_or_none, scale=1000)}),
 383                 'height': ('resolution', 'height', {int_or_none}),
 384                 'width': ('resolution', 'width', {int_or_none}),
 385             })),
 386             'quality': -2 if 'low' in video_quality['id'] else None,
 387             'protocol': 'niconico_dmc',
 388             'expected_protocol': dmc_protocol,  # XXX: This is not a documented field
 389             'http_headers': {
 390                 'Origin': 'https://www.nicovideo.jp',
 391                 'Referer': 'https://www.nicovideo.jp/watch/' + video_id,
 392             }
 393         }
 394
 395     def _yield_dmc_formats(self, api_data, video_id):
 396         dmc_data = traverse_obj(api_data, ('media', 'delivery', 'movie'))
 397         audios = traverse_obj(dmc_data, ('audios', ..., {dict}))
 398         videos = traverse_obj(dmc_data, ('videos', ..., {dict}))
 399         protocols = traverse_obj(dmc_data, ('session', 'protocols', ..., {str}))
 400         if not all((audios, videos, protocols)):
 401             return
 402
 403         for audio_quality, video_quality, protocol in itertools.product(audios, videos, protocols):
 404             if fmt := self._extract_format_for_quality(video_id, audio_quality, video_quality, protocol):
 405                 yield fmt
 406
 407     def _yield_dms_formats(self, api_data, video_id):
 408         fmt_filter = lambda _, v: v['isAvailable'] and v['id']
 409         videos = traverse_obj(api_data, ('media', 'domand', 'videos', fmt_filter))
 410         audios = traverse_obj(api_data, ('media', 'domand', 'audios', fmt_filter))
 411         access_key = traverse_obj(api_data, ('media', 'domand', 'accessRightKey', {str}))
 412         track_id = traverse_obj(api_data, ('client', 'watchTrackId', {str}))
 413         if not all((videos, audios, access_key, track_id)):
 414             return
 415
 416         dms_m3u8_url = self._download_json(
 417             f'https://nvapi.nicovideo.jp/v1/watch/{video_id}/access-rights/hls', video_id,
 418             data=json.dumps({
 419                 'outputs': list(itertools.product((v['id'] for v in videos), (a['id'] for a in audios)))
 420             }).encode(), query={'actionTrackId': track_id}, headers={
 421                 'x-access-right-key': access_key,
 422                 'x-frontend-id': 6,
 423                 'x-frontend-version': 0,
 424                 'x-request-with': 'https://www.nicovideo.jp',
 425             })['data']['contentUrl']
 426         # Getting all audio formats results in duplicate video formats which we filter out later
 427         dms_fmts = self._extract_m3u8_formats(dms_m3u8_url, video_id)
 428
 429         # m3u8 extraction does not provide audio bitrates, so extract from the API data and fix
 430         for audio_fmt in traverse_obj(dms_fmts, lambda _, v: v['vcodec'] == 'none'):
 431             yield {
 432                 **audio_fmt,
 433                 **traverse_obj(audios, (lambda _, v: audio_fmt['format_id'].startswith(v['id']), {
 434                     'format_id': ('id', {str}),
 435                     'abr': ('bitRate', {functools.partial(float_or_none, scale=1000)}),
 436                     'asr': ('samplingRate', {int_or_none}),
 437                 }), get_all=False),
 438                 'acodec': 'aac',
 439                 'ext': 'm4a',
 440             }
 441
 442         # Sort before removing dupes to keep the format dicts with the lowest tbr
 443         video_fmts = sorted((fmt for fmt in dms_fmts if fmt['vcodec'] != 'none'), key=lambda f: f['tbr'])
 444         self._remove_duplicate_formats(video_fmts)
 445         # Calculate the true vbr/tbr by subtracting the lowest abr
 446         min_abr = min(traverse_obj(audios, (..., 'bitRate', {float_or_none})), default=0) / 1000
 447         for video_fmt in video_fmts:
 448             video_fmt['tbr'] -= min_abr
 449             video_fmt['format_id'] = f'video-{video_fmt["tbr"]:.0f}'
 450             yield video_fmt
 451
 452     def _real_extract(self, url):
 453         video_id = self._match_id(url)
 454
 455         try:
 456             webpage, handle = self._download_webpage_handle(
 457                 'https://www.nicovideo.jp/watch/' + video_id, video_id)
 458             if video_id.startswith('so'):
 459                 video_id = self._match_id(handle.url)
 460
 461             api_data = self._parse_json(self._html_search_regex(
 462                 'data-api-data="([^"]+)"', webpage,
 463                 'API data', default='{}'), video_id)
 464         except ExtractorError as e:
 465             try:
 466                 api_data = self._download_json(
 467                     'https://www.nicovideo.jp/api/watch/v3/%s?_frontendId=6&_frontendVersion=0&actionTrackId=AAAAAAAAAA_%d' % (video_id, round(time.time() * 1000)), video_id,
 468                     note='Downloading API JSON', errnote='Unable to fetch data')['data']
 469             except ExtractorError:
 470                 if not isinstance(e.cause, HTTPError):
 471                     raise
 472                 webpage = e.cause.response.read().decode('utf-8', 'replace')
 473                 error_msg = self._html_search_regex(
 474                     r'(?s)<section\s+class="(?:(?:ErrorMessage|WatchExceptionPage-message)\s*)+">(.+?)</section>',
 475                     webpage, 'error reason', default=None)
 476                 if not error_msg:
 477                     raise
 478                 raise ExtractorError(clean_html(error_msg), expected=True)
 479
 480         availability = self._availability(**(traverse_obj(api_data, ('payment', 'video', {
 481             'needs_premium': ('isPremium', {bool}),
 482             'needs_subscription': ('isAdmission', {bool}),
 483         })) or {'needs_auth': True}))
 484         formats = [*self._yield_dmc_formats(api_data, video_id),
 485                    *self._yield_dms_formats(api_data, video_id)]
 486         if not formats:
 487             fail_msg = clean_html(self._html_search_regex(
 488                 r'<p[^>]+\bclass="fail-message"[^>]*>(?P<msg>.+?)</p>',
 489                 webpage, 'fail message', default=None, group='msg'))
 490             if fail_msg:
 491                 self.to_screen(f'Niconico said: {fail_msg}')
 492             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:
 493                 availability = None
 494                 self.raise_geo_restricted(countries=self._GEO_COUNTRIES, metadata_available=True)
 495             elif availability == 'premium_only':
 496                 self.raise_login_required('This video requires premium', metadata_available=True)
 497             elif availability == 'subscriber_only':
 498                 self.raise_login_required('This video is for members only', metadata_available=True)
 499             elif availability == 'needs_auth':
 500                 self.raise_login_required(metadata_available=False)
 501
 502         # Start extracting information
 503         tags = None
 504         if webpage:
 505             # use og:video:tag (not logged in)
 506             og_video_tags = re.finditer(r'<meta\s+property="og:video:tag"\s*content="(.*?)">', webpage)
 507             tags = list(filter(None, (clean_html(x.group(1)) for x in og_video_tags)))
 508             if not tags:
 509                 # use keywords and split with comma (not logged in)
 510                 kwds = self._html_search_meta('keywords', webpage, default=None)
 511                 if kwds:
 512                     tags = [x for x in kwds.split(',') if x]
 513         if not tags:
 514             # find in json (logged in)
 515             tags = traverse_obj(api_data, ('tag', 'items', ..., 'name'))
 516
 517         thumb_prefs = qualities(['url', 'middleUrl', 'largeUrl', 'player', 'ogp'])
 518
 519         def get_video_info(*items, get_first=True, **kwargs):
 520             return traverse_obj(api_data, ('video', *items), get_all=not get_first, **kwargs)
 521
 522         return {
 523             'id': video_id,
 524             '_api_data': api_data,
 525             'title': get_video_info(('originalTitle', 'title')) or self._og_search_title(webpage, default=None),
 526             'formats': formats,
 527             'availability': availability,
 528             'thumbnails': [{
 529                 'id': key,
 530                 'url': url,
 531                 'ext': 'jpg',
 532                 'preference': thumb_prefs(key),
 533                 **parse_resolution(url, lenient=True),
 534             } for key, url in (get_video_info('thumbnail') or {}).items() if url],
 535             'description': clean_html(get_video_info('description')),
 536             'uploader': traverse_obj(api_data, ('owner', 'nickname'), ('channel', 'name'), ('community', 'name')),
 537             'uploader_id': str_or_none(traverse_obj(api_data, ('owner', 'id'), ('channel', 'id'), ('community', 'id'))),
 538             'timestamp': parse_iso8601(get_video_info('registeredAt')) or parse_iso8601(
 539                 self._html_search_meta('video:release_date', webpage, 'date published', default=None)),
 540             'channel': traverse_obj(api_data, ('channel', 'name'), ('community', 'name')),
 541             'channel_id': traverse_obj(api_data, ('channel', 'id'), ('community', 'id')),
 542             'view_count': int_or_none(get_video_info('count', 'view')),
 543             'tags': tags,
 544             'genre': traverse_obj(api_data, ('genre', 'label'), ('genre', 'key')),
 545             'comment_count': get_video_info('count', 'comment', expected_type=int),
 546             'duration': (
 547                 parse_duration(self._html_search_meta('video:duration', webpage, 'video duration', default=None))
 548                 or get_video_info('duration')),
 549             'webpage_url': url_or_none(url) or f'https://www.nicovideo.jp/watch/{video_id}',
 550             'subtitles': self.extract_subtitles(video_id, api_data),
 551         }
 552
 553     def _get_subtitles(self, video_id, api_data):
 554         comments_info = traverse_obj(api_data, ('comment', 'nvComment', {dict})) or {}
 555         if not comments_info.get('server'):
 556             return
 557
 558         danmaku = traverse_obj(self._download_json(
 559             f'{comments_info["server"]}/v1/threads', video_id, data=json.dumps({
 560                 'additionals': {},
 561                 'params': comments_info.get('params'),
 562                 'threadKey': comments_info.get('threadKey'),
 563             }).encode(), fatal=False,
 564             headers={
 565                 'Referer': 'https://www.nicovideo.jp/',
 566                 'Origin': 'https://www.nicovideo.jp',
 567                 'Content-Type': 'text/plain;charset=UTF-8',
 568                 'x-client-os-type': 'others',
 569                 'x-frontend-id': '6',
 570                 'x-frontend-version': '0',
 571             },
 572             note='Downloading comments', errnote='Failed to download comments'),
 573             ('data', 'threads', ..., 'comments', ...))
 574
 575         return {
 576             'comments': [{
 577                 'ext': 'json',
 578                 'data': json.dumps(danmaku),
 579             }],
 580         }
 581
 582
 583 class NiconicoPlaylistBaseIE(InfoExtractor):
 584     _PAGE_SIZE = 100
 585
 586     _API_HEADERS = {
 587         'X-Frontend-ID': '6',
 588         'X-Frontend-Version': '0',
 589         'X-Niconico-Language': 'en-us'
 590     }
 591
 592     def _call_api(self, list_id, resource, query):
 593         raise NotImplementedError('Must be implemented in subclasses')
 594
 595     @staticmethod
 596     def _parse_owner(item):
 597         return {
 598             'uploader': traverse_obj(item, ('owner', 'name')),
 599             'uploader_id': traverse_obj(item, ('owner', 'id')),
 600         }
 601
 602     def _fetch_page(self, list_id, page):
 603         page += 1
 604         resp = self._call_api(list_id, 'page %d' % page, {
 605             'page': page,
 606             'pageSize': self._PAGE_SIZE,
 607         })
 608         # this is needed to support both mylist and user
 609         for video in traverse_obj(resp, ('items', ..., ('video', None))) or []:
 610             video_id = video.get('id')
 611             if not video_id:
 612                 # skip {"video": {"id": "blablabla", ...}}
 613                 continue
 614             count = video.get('count') or {}
 615             get_count = lambda x: int_or_none(count.get(x))
 616             yield {
 617                 '_type': 'url',
 618                 'id': video_id,
 619                 'title': video.get('title'),
 620                 'url': f'https://www.nicovideo.jp/watch/{video_id}',
 621                 'description': video.get('shortDescription'),
 622                 'duration': int_or_none(video.get('duration')),
 623                 'view_count': get_count('view'),
 624                 'comment_count': get_count('comment'),
 625                 'thumbnail': traverse_obj(video, ('thumbnail', ('nHdUrl', 'largeUrl', 'listingUrl', 'url'))),
 626                 'ie_key': NiconicoIE.ie_key(),
 627                 **self._parse_owner(video),
 628             }
 629
 630     def _entries(self, list_id):
 631         return OnDemandPagedList(functools.partial(self._fetch_page, list_id), self._PAGE_SIZE)
 632
 633
 634 class NiconicoPlaylistIE(NiconicoPlaylistBaseIE):
 635     IE_NAME = 'niconico:playlist'
 636     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp|nico\.ms)/(?:user/\d+/)?(?:my/)?mylist/(?:#/)?(?P<id>\d+)'
 637
 638     _TESTS = [{
 639         'url': 'http://www.nicovideo.jp/mylist/27411728',
 640         'info_dict': {
 641             'id': '27411728',
 642             'title': 'AKB48のオールナイトニッポン',
 643             'description': 'md5:d89694c5ded4b6c693dea2db6e41aa08',
 644             'uploader': 'のっく',
 645             'uploader_id': '805442',
 646         },
 647         'playlist_mincount': 291,
 648     }, {
 649         'url': 'https://www.nicovideo.jp/user/805442/mylist/27411728',
 650         'only_matching': True,
 651     }, {
 652         'url': 'https://www.nicovideo.jp/my/mylist/#/68048635',
 653         'only_matching': True,
 654     }]
 655
 656     def _call_api(self, list_id, resource, query):
 657         return self._download_json(
 658             f'https://nvapi.nicovideo.jp/v2/mylists/{list_id}', list_id,
 659             f'Downloading {resource}', query=query,
 660             headers=self._API_HEADERS)['data']['mylist']
 661
 662     def _real_extract(self, url):
 663         list_id = self._match_id(url)
 664         mylist = self._call_api(list_id, 'list', {
 665             'pageSize': 1,
 666         })
 667         return self.playlist_result(
 668             self._entries(list_id), list_id,
 669             mylist.get('name'), mylist.get('description'), **self._parse_owner(mylist))
 670
 671
 672 class NiconicoSeriesIE(InfoExtractor):
 673     IE_NAME = 'niconico:series'
 674     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp(?:/user/\d+)?|nico\.ms)/series/(?P<id>\d+)'
 675
 676     _TESTS = [{
 677         'url': 'https://www.nicovideo.jp/user/44113208/series/110226',
 678         'info_dict': {
 679             'id': '110226',
 680             'title': 'ご立派ァ！のシリーズ',
 681         },
 682         'playlist_mincount': 10,
 683     }, {
 684         'url': 'https://www.nicovideo.jp/series/12312/',
 685         'info_dict': {
 686             'id': '12312',
 687             'title': 'バトルスピリッツ　お勧めカード紹介(調整中)',
 688         },
 689         'playlist_mincount': 103,
 690     }, {
 691         'url': 'https://nico.ms/series/203559',
 692         'only_matching': True,
 693     }]
 694
 695     def _real_extract(self, url):
 696         list_id = self._match_id(url)
 697         webpage = self._download_webpage(url, list_id)
 698
 699         title = self._search_regex(
 700             (r'<title>「(.+)（全',
 701              r'<div class="TwitterShareButton"\s+data-text="(.+)\s+https:'),
 702             webpage, 'title', fatal=False)
 703         if title:
 704             title = unescapeHTML(title)
 705         json_data = next(self._yield_json_ld(webpage, None, fatal=False))
 706         return self.playlist_from_matches(
 707             traverse_obj(json_data, ('itemListElement', ..., 'url')), list_id, title, ie=NiconicoIE)
 708
 709
 710 class NiconicoHistoryIE(NiconicoPlaylistBaseIE):
 711     IE_NAME = 'niconico:history'
 712     IE_DESC = 'NicoNico user history or likes. Requires cookies.'
 713     _VALID_URL = r'https?://(?:www\.|sp\.)?nicovideo\.jp/my/(?P<id>history(?:/like)?)'
 714
 715     _TESTS = [{
 716         'note': 'PC page, with /video',
 717         'url': 'https://www.nicovideo.jp/my/history/video',
 718         'only_matching': True,
 719     }, {
 720         'note': 'PC page, without /video',
 721         'url': 'https://www.nicovideo.jp/my/history',
 722         'only_matching': True,
 723     }, {
 724         'note': 'mobile page, with /video',
 725         'url': 'https://sp.nicovideo.jp/my/history/video',
 726         'only_matching': True,
 727     }, {
 728         'note': 'mobile page, without /video',
 729         'url': 'https://sp.nicovideo.jp/my/history',
 730         'only_matching': True,
 731     }, {
 732         'note': 'PC page',
 733         'url': 'https://www.nicovideo.jp/my/history/like',
 734         'only_matching': True,
 735     }, {
 736         'note': 'Mobile page',
 737         'url': 'https://sp.nicovideo.jp/my/history/like',
 738         'only_matching': True,
 739     }]
 740
 741     def _call_api(self, list_id, resource, query):
 742         path = 'likes' if list_id == 'history/like' else 'watch/history'
 743         return self._download_json(
 744             f'https://nvapi.nicovideo.jp/v1/users/me/{path}', list_id,
 745             f'Downloading {resource}', query=query, headers=self._API_HEADERS)['data']
 746
 747     def _real_extract(self, url):
 748         list_id = self._match_id(url)
 749         try:
 750             mylist = self._call_api(list_id, 'list', {'pageSize': 1})
 751         except ExtractorError as e:
 752             if isinstance(e.cause, HTTPError) and e.cause.status == 401:
 753                 self.raise_login_required('You have to be logged in to get your history')
 754             raise
 755         return self.playlist_result(self._entries(list_id), list_id, **self._parse_owner(mylist))
 756
 757
 758 class NicovideoSearchBaseIE(InfoExtractor):
 759     _SEARCH_TYPE = 'search'
 760
 761     def _entries(self, url, item_id, query=None, note='Downloading page %(page)s'):
 762         query = query or {}
 763         pages = [query['page']] if 'page' in query else itertools.count(1)
 764         for page_num in pages:
 765             query['page'] = str(page_num)
 766             webpage = self._download_webpage(url, item_id, query=query, note=note % {'page': page_num})
 767             results = re.findall(r'(?<=data-video-id=)["\']?(?P<videoid>.*?)(?=["\'])', webpage)
 768             for item in results:
 769                 yield self.url_result(f'https://www.nicovideo.jp/watch/{item}', 'Niconico', item)
 770             if not results:
 771                 break
 772
 773     def _search_results(self, query):
 774         return self._entries(
 775             self._proto_relative_url(f'//www.nicovideo.jp/{self._SEARCH_TYPE}/{query}'), query)
 776
 777
 778 class NicovideoSearchIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 779     IE_DESC = 'Nico video search'
 780     IE_NAME = 'nicovideo:search'
 781     _SEARCH_KEY = 'nicosearch'
 782
 783
 784 class NicovideoSearchURLIE(NicovideoSearchBaseIE):
 785     IE_NAME = f'{NicovideoSearchIE.IE_NAME}_url'
 786     IE_DESC = 'Nico video search URLs'
 787     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/search/(?P<id>[^?#&]+)?'
 788     _TESTS = [{
 789         'url': 'http://www.nicovideo.jp/search/sm9',
 790         'info_dict': {
 791             'id': 'sm9',
 792             'title': 'sm9'
 793         },
 794         'playlist_mincount': 40,
 795     }, {
 796         'url': 'https://www.nicovideo.jp/search/sm9?sort=h&order=d&end=2020-12-31&start=2020-01-01',
 797         'info_dict': {
 798             'id': 'sm9',
 799             'title': 'sm9'
 800         },
 801         'playlist_count': 31,
 802     }]
 803
 804     def _real_extract(self, url):
 805         query = self._match_id(url)
 806         return self.playlist_result(self._entries(url, query), query, query)
 807
 808
 809 class NicovideoSearchDateIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 810     IE_DESC = 'Nico video search, newest first'
 811     IE_NAME = f'{NicovideoSearchIE.IE_NAME}:date'
 812     _SEARCH_KEY = 'nicosearchdate'
 813     _TESTS = [{
 814         'url': 'nicosearchdateall:a',
 815         'info_dict': {
 816             'id': 'a',
 817             'title': 'a'
 818         },
 819         'playlist_mincount': 1610,
 820     }]
 821
 822     _START_DATE = dt.date(2007, 1, 1)
 823     _RESULTS_PER_PAGE = 32
 824     _MAX_PAGES = 50
 825
 826     def _entries(self, url, item_id, start_date=None, end_date=None):
 827         start_date, end_date = start_date or self._START_DATE, end_date or dt.datetime.now().date()
 828
 829         # If the last page has a full page of videos, we need to break down the query interval further
 830         last_page_len = len(list(self._get_entries_for_date(
 831             url, item_id, start_date, end_date, self._MAX_PAGES,
 832             note=f'Checking number of videos from {start_date} to {end_date}')))
 833         if (last_page_len == self._RESULTS_PER_PAGE and start_date != end_date):
 834             midpoint = start_date + ((end_date - start_date) // 2)
 835             yield from self._entries(url, item_id, midpoint, end_date)
 836             yield from self._entries(url, item_id, start_date, midpoint)
 837         else:
 838             self.to_screen(f'{item_id}: Downloading results from {start_date} to {end_date}')
 839             yield from self._get_entries_for_date(
 840                 url, item_id, start_date, end_date, note='    Downloading page %(page)s')
 841
 842     def _get_entries_for_date(self, url, item_id, start_date, end_date=None, page_num=None, note=None):
 843         query = {
 844             'start': str(start_date),
 845             'end': str(end_date or start_date),
 846             'sort': 'f',
 847             'order': 'd',
 848         }
 849         if page_num:
 850             query['page'] = str(page_num)
 851
 852         yield from super()._entries(url, item_id, query=query, note=note)
 853
 854
 855 class NicovideoTagURLIE(NicovideoSearchBaseIE):
 856     IE_NAME = 'niconico:tag'
 857     IE_DESC = 'NicoNico video tag URLs'
 858     _SEARCH_TYPE = 'tag'
 859     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/tag/(?P<id>[^?#&]+)?'
 860     _TESTS = [{
 861         'url': 'https://www.nicovideo.jp/tag/ドキュメンタリー淫夢',
 862         'info_dict': {
 863             'id': 'ドキュメンタリー淫夢',
 864             'title': 'ドキュメンタリー淫夢'
 865         },
 866         'playlist_mincount': 400,
 867     }]
 868
 869     def _real_extract(self, url):
 870         query = self._match_id(url)
 871         return self.playlist_result(self._entries(url, query), query, query)
 872
 873
 874 class NiconicoUserIE(InfoExtractor):
 875     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/user/(?P<id>\d+)/?(?:$|[#?])'
 876     _TEST = {
 877         'url': 'https://www.nicovideo.jp/user/419948',
 878         'info_dict': {
 879             'id': '419948',
 880         },
 881         'playlist_mincount': 101,
 882     }
 883     _API_URL = "https://nvapi.nicovideo.jp/v1/users/%s/videos?sortKey=registeredAt&sortOrder=desc&pageSize=%s&page=%s"
 884     _PAGE_SIZE = 100
 885
 886     _API_HEADERS = {
 887         'X-Frontend-ID': '6',
 888         'X-Frontend-Version': '0'
 889     }
 890
 891     def _entries(self, list_id):
 892         total_count = 1
 893         count = page_num = 0
 894         while count < total_count:
 895             json_parsed = self._download_json(
 896                 self._API_URL % (list_id, self._PAGE_SIZE, page_num + 1), list_id,
 897                 headers=self._API_HEADERS,
 898                 note='Downloading JSON metadata%s' % (' page %d' % page_num if page_num else ''))
 899             if not page_num:
 900                 total_count = int_or_none(json_parsed['data'].get('totalCount'))
 901             for entry in json_parsed["data"]["items"]:
 902                 count += 1
 903                 yield self.url_result('https://www.nicovideo.jp/watch/%s' % entry['id'])
 904             page_num += 1
 905
 906     def _real_extract(self, url):
 907         list_id = self._match_id(url)
 908         return self.playlist_result(self._entries(list_id), list_id, ie=NiconicoIE.ie_key())
 909
 910
 911 class NiconicoLiveIE(InfoExtractor):
 912     IE_NAME = 'niconico:live'
 913     IE_DESC = 'ニコニコ生放送'
 914     _VALID_URL = r'https?://(?:sp\.)?live2?\.nicovideo\.jp/(?:watch|gate)/(?P<id>lv\d+)'
 915     _TESTS = [{
 916         'note': 'this test case includes invisible characters for title, pasting them as-is',
 917         'url': 'https://live.nicovideo.jp/watch/lv339533123',
 918         'info_dict': {
 919             'id': 'lv339533123',
 920             'title': '激辛ペヤング食べます‪( ;ᯅ; )‬（歌枠オーディション参加中）',
 921             'view_count': 1526,
 922             'comment_count': 1772,
 923             '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',
 924             'uploader': 'もか',
 925             'channel': 'ゲストさんのコミュニティ',
 926             'channel_id': 'co5776900',
 927             'channel_url': 'https://com.nicovideo.jp/community/co5776900',
 928             'timestamp': 1670677328,
 929             'is_live': True,
 930         },
 931         'skip': 'livestream',
 932     }, {
 933         'url': 'https://live2.nicovideo.jp/watch/lv339533123',
 934         'only_matching': True,
 935     }, {
 936         'url': 'https://sp.live.nicovideo.jp/watch/lv339533123',
 937         'only_matching': True,
 938     }, {
 939         'url': 'https://sp.live2.nicovideo.jp/watch/lv339533123',
 940         'only_matching': True,
 941     }]
 942
 943     _KNOWN_LATENCY = ('high', 'low')
 944
 945     def _real_extract(self, url):
 946         video_id = self._match_id(url)
 947         webpage, urlh = self._download_webpage_handle(f'https://live.nicovideo.jp/watch/{video_id}', video_id)
 948
 949         embedded_data = self._parse_json(unescapeHTML(self._search_regex(
 950             r'<script\s+id="embedded-data"\s*data-props="(.+?)"', webpage, 'embedded data')), video_id)
 951
 952         ws_url = traverse_obj(embedded_data, ('site', 'relive', 'webSocketUrl'))
 953         if not ws_url:
 954             raise ExtractorError('The live hasn\'t started yet or already ended.', expected=True)
 955         ws_url = update_url_query(ws_url, {
 956             'frontend_id': traverse_obj(embedded_data, ('site', 'frontendId')) or '9',
 957         })
 958
 959         hostname = remove_start(urllib.parse.urlparse(urlh.url).hostname, 'sp.')
 960         latency = try_get(self._configuration_arg('latency'), lambda x: x[0])
 961         if latency not in self._KNOWN_LATENCY:
 962             latency = 'high'
 963
 964         ws = self._request_webpage(
 965             Request(ws_url, headers={'Origin': f'https://{hostname}'}),
 966             video_id=video_id, note='Connecting to WebSocket server')
 967
 968         self.write_debug('[debug] Sending HLS server request')
 969         ws.send(json.dumps({
 970             'type': 'startWatching',
 971             'data': {
 972                 'stream': {
 973                     'quality': 'abr',
 974                     'protocol': 'hls+fmp4',
 975                     'latency': latency,
 976                     'chasePlay': False
 977                 },
 978                 'room': {
 979                     'protocol': 'webSocket',
 980                     'commentable': True
 981                 },
 982                 'reconnect': False,
 983             }
 984         }))
 985
 986         while True:
 987             recv = ws.recv()
 988             if not recv:
 989                 continue
 990             data = json.loads(recv)
 991             if not isinstance(data, dict):
 992                 continue
 993             if data.get('type') == 'stream':
 994                 m3u8_url = data['data']['uri']
 995                 qualities = data['data']['availableQualities']
 996                 break
 997             elif data.get('type') == 'disconnect':
 998                 self.write_debug(recv)
 999                 raise ExtractorError('Disconnected at middle of extraction')
1000             elif data.get('type') == 'error':
1001                 self.write_debug(recv)
1002                 message = traverse_obj(data, ('body', 'code')) or recv
1003                 raise ExtractorError(message)
1004             elif self.get_param('verbose', False):
1005                 if len(recv) > 100:
1006                     recv = recv[:100] + '...'
1007                 self.write_debug('Server said: %s' % recv)
1008
1009         title = traverse_obj(embedded_data, ('program', 'title')) or self._html_search_meta(
1010             ('og:title', 'twitter:title'), webpage, 'live title', fatal=False)
1011
1012         raw_thumbs = traverse_obj(embedded_data, ('program', 'thumbnail')) or {}
1013         thumbnails = []
1014         for name, value in raw_thumbs.items():
1015             if not isinstance(value, dict):
1016                 thumbnails.append({
1017                     'id': name,
1018                     'url': value,
1019                     **parse_resolution(value, lenient=True),
1020                 })
1021                 continue
1022
1023             for k, img_url in value.items():
1024                 res = parse_resolution(k, lenient=True) or parse_resolution(img_url, lenient=True)
1025                 width, height = res.get('width'), res.get('height')
1026
1027                 thumbnails.append({
1028                     'id': f'{name}_{width}x{height}',
1029                     'url': img_url,
1030                     **res,
1031                 })
1032
1033         formats = self._extract_m3u8_formats(m3u8_url, video_id, ext='mp4', live=True)
1034         for fmt, q in zip(formats, reversed(qualities[1:])):
1035             fmt.update({
1036                 'format_id': q,
1037                 'protocol': 'niconico_live',
1038                 'ws': ws,
1039                 'video_id': video_id,
1040                 'live_latency': latency,
1041                 'origin': hostname,
1042             })
1043
1044         return {
1045             'id': video_id,
1046             'title': title,
1047             **traverse_obj(embedded_data, {
1048                 'view_count': ('program', 'statistics', 'watchCount'),
1049                 'comment_count': ('program', 'statistics', 'commentCount'),
1050                 'uploader': ('program', 'supplier', 'name'),
1051                 'channel': ('socialGroup', 'name'),
1052                 'channel_id': ('socialGroup', 'id'),
1053                 'channel_url': ('socialGroup', 'socialGroupPageUrl'),
1054             }),
1055             'description': clean_html(traverse_obj(embedded_data, ('program', 'description'))),
1056             'timestamp': int_or_none(traverse_obj(embedded_data, ('program', 'openTime'))),
1057             'is_live': True,
1058             'thumbnails': thumbnails,
1059             'formats': formats,
1060         }