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 ..dependencies import websockets
  12 from ..networking.exceptions import HTTPError
  13 from ..utils import (
  14     ExtractorError,
  15     OnDemandPagedList,
  16     WebSocketsWrapper,
  17     bug_reports_message,
  18     clean_html,
  19     float_or_none,
  20     int_or_none,
  21     join_nonempty,
  22     parse_duration,
  23     parse_filesize,
  24     parse_iso8601,
  25     parse_resolution,
  26     qualities,
  27     remove_start,
  28     str_or_none,
  29     traverse_obj,
  30     try_get,
  31     unescapeHTML,
  32     update_url_query,
  33     url_or_none,
  34     urlencode_postdata,
  35     urljoin,
  36 )
  37
  38
  39 class NiconicoIE(InfoExtractor):
  40     IE_NAME = 'niconico'
  41     IE_DESC = 'ニコニコ動画'
  42
  43     _TESTS = [{
  44         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  45         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  46         'info_dict': {
  47             'id': 'sm22312215',
  48             'ext': 'mp4',
  49             'title': 'Big Buck Bunny',
  50             'thumbnail': r're:https?://.*',
  51             'uploader': 'takuya0301',
  52             'uploader_id': '2698420',
  53             'upload_date': '20131123',
  54             'timestamp': int,  # timestamp is unstable
  55             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  56             'duration': 33,
  57             'view_count': int,
  58             'comment_count': int,
  59         },
  60         'skip': 'Requires an account',
  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': 'swf',
  68             'title': '【鏡音リン】Dance on media【オリジナル】take2!',
  69             'description': 'md5:689f066d74610b3b22e0f1739add0f58',
  70             'thumbnail': r're:https?://.*',
  71             'uploader': 'りょうた',
  72             'uploader_id': '18822557',
  73             'upload_date': '20110429',
  74             'timestamp': 1304065916,
  75             'duration': 209,
  76         },
  77         'skip': 'Requires an account',
  78     }, {
  79         # 'video exists but is marked as "deleted"
  80         # md5 is unstable
  81         'url': 'http://www.nicovideo.jp/watch/sm10000',
  82         'info_dict': {
  83             'id': 'sm10000',
  84             'ext': 'unknown_video',
  85             'description': 'deleted',
  86             'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞',
  87             'thumbnail': r're:https?://.*',
  88             'upload_date': '20071224',
  89             'timestamp': int,  # timestamp field has different value if logged in
  90             'duration': 304,
  91             'view_count': int,
  92         },
  93         'skip': 'Requires an account',
  94     }, {
  95         'url': 'http://www.nicovideo.jp/watch/so22543406',
  96         'info_dict': {
  97             'id': '1388129933',
  98             'ext': 'mp4',
  99             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
 100             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
 101             'thumbnail': r're:https?://.*',
 102             'timestamp': 1388851200,
 103             'upload_date': '20140104',
 104             'uploader': 'アニメロチャンネル',
 105             'uploader_id': '312',
 106         },
 107         'skip': 'The viewing period of the video you were searching for has expired.',
 108     }, {
 109         # video not available via `getflv`; "old" HTML5 video
 110         'url': 'http://www.nicovideo.jp/watch/sm1151009',
 111         'md5': '8fa81c364eb619d4085354eab075598a',
 112         'info_dict': {
 113             'id': 'sm1151009',
 114             'ext': 'mp4',
 115             '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）',
 116             'description': 'md5:6ee077e0581ff5019773e2e714cdd0b7',
 117             'thumbnail': r're:https?://.*',
 118             'duration': 184,
 119             'timestamp': 1190868283,
 120             'upload_date': '20070927',
 121             'uploader': 'denden2',
 122             'uploader_id': '1392194',
 123             'view_count': int,
 124             'comment_count': int,
 125         },
 126         'skip': 'Requires an account',
 127     }, {
 128         # "New" HTML5 video
 129         # md5 is unstable
 130         'url': 'http://www.nicovideo.jp/watch/sm31464864',
 131         'info_dict': {
 132             'id': 'sm31464864',
 133             'ext': 'mp4',
 134             'title': '新作TVアニメ「戦姫絶唱シンフォギアAXZ」PV 最高画質',
 135             'description': 'md5:e52974af9a96e739196b2c1ca72b5feb',
 136             'timestamp': 1498514060,
 137             'upload_date': '20170626',
 138             'uploader': 'ゲスト',
 139             'uploader_id': '40826363',
 140             'thumbnail': r're:https?://.*',
 141             'duration': 198,
 142             'view_count': int,
 143             'comment_count': int,
 144         },
 145         'skip': 'Requires an account',
 146     }, {
 147         # Video without owner
 148         'url': 'http://www.nicovideo.jp/watch/sm18238488',
 149         'md5': 'd265680a1f92bdcbbd2a507fc9e78a9e',
 150         'info_dict': {
 151             'id': 'sm18238488',
 152             'ext': 'mp4',
 153             'title': '【実写版】ミュータントタートルズ',
 154             'description': 'md5:15df8988e47a86f9e978af2064bf6d8e',
 155             'timestamp': 1341160408,
 156             'upload_date': '20120701',
 157             'uploader': None,
 158             'uploader_id': None,
 159             'thumbnail': r're:https?://.*',
 160             'duration': 5271,
 161             'view_count': int,
 162             'comment_count': int,
 163         },
 164         'skip': 'Requires an account',
 165     }, {
 166         'url': 'http://sp.nicovideo.jp/watch/sm28964488?ss_pos=1&cp_in=wt_tg',
 167         'only_matching': True,
 168     }, {
 169         'note': 'a video that is only served as an ENCRYPTED HLS.',
 170         'url': 'https://www.nicovideo.jp/watch/so38016254',
 171         'only_matching': True,
 172     }]
 173
 174     _VALID_URL = r'https?://(?:(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch|nico\.ms)/(?P<id>(?:[a-z]{2})?[0-9]+)'
 175     _NETRC_MACHINE = 'niconico'
 176     _COMMENT_API_ENDPOINTS = (
 177         'https://nvcomment.nicovideo.jp/legacy/api.json',
 178         'https://nmsg.nicovideo.jp/api.json',)
 179     _API_HEADERS = {
 180         'X-Frontend-ID': '6',
 181         'X-Frontend-Version': '0',
 182         'X-Niconico-Language': 'en-us',
 183         'Referer': 'https://www.nicovideo.jp/',
 184         'Origin': 'https://www.nicovideo.jp',
 185     }
 186
 187     def _perform_login(self, username, password):
 188         login_ok = True
 189         login_form_strs = {
 190             'mail_tel': username,
 191             'password': password,
 192         }
 193         self._request_webpage(
 194             'https://account.nicovideo.jp/login', None,
 195             note='Acquiring Login session')
 196         page = self._download_webpage(
 197             'https://account.nicovideo.jp/login/redirector?show_button_twitter=1&site=niconico&show_button_facebook=1', None,
 198             note='Logging in', errnote='Unable to log in',
 199             data=urlencode_postdata(login_form_strs),
 200             headers={
 201                 'Referer': 'https://account.nicovideo.jp/login',
 202                 'Content-Type': 'application/x-www-form-urlencoded',
 203             })
 204         if 'oneTimePw' in page:
 205             post_url = self._search_regex(
 206                 r'<form[^>]+action=(["\'])(?P<url>.+?)\1', page, 'post url', group='url')
 207             page = self._download_webpage(
 208                 urljoin('https://account.nicovideo.jp', post_url), None,
 209                 note='Performing MFA', errnote='Unable to complete MFA',
 210                 data=urlencode_postdata({
 211                     'otp': self._get_tfa_info('6 digits code')
 212                 }), headers={
 213                     'Content-Type': 'application/x-www-form-urlencoded',
 214                 })
 215             if 'oneTimePw' in page or 'formError' in page:
 216                 err_msg = self._html_search_regex(
 217                     r'formError["\']+>(.*?)</div>', page, 'form_error',
 218                     default='There\'s an error but the message can\'t be parsed.',
 219                     flags=re.DOTALL)
 220                 self.report_warning(f'Unable to log in: MFA challenge failed, "{err_msg}"')
 221                 return False
 222         login_ok = 'class="notice error"' not in page
 223         if not login_ok:
 224             self.report_warning('Unable to log in: bad username or password')
 225         return login_ok
 226
 227     def _get_heartbeat_info(self, info_dict):
 228         video_id, video_src_id, audio_src_id = info_dict['url'].split(':')[1].split('/')
 229         dmc_protocol = info_dict['expected_protocol']
 230
 231         api_data = (
 232             info_dict.get('_api_data')
 233             or self._parse_json(
 234                 self._html_search_regex(
 235                     'data-api-data="([^"]+)"',
 236                     self._download_webpage('https://www.nicovideo.jp/watch/' + video_id, video_id),
 237                     'API data', default='{}'),
 238                 video_id))
 239
 240         session_api_data = try_get(api_data, lambda x: x['media']['delivery']['movie']['session'])
 241         session_api_endpoint = try_get(session_api_data, lambda x: x['urls'][0])
 242
 243         def ping():
 244             tracking_id = traverse_obj(api_data, ('media', 'delivery', 'trackingId'))
 245             if tracking_id:
 246                 tracking_url = update_url_query('https://nvapi.nicovideo.jp/v1/2ab0cbaa/watch', {'t': tracking_id})
 247                 watch_request_response = self._download_json(
 248                     tracking_url, video_id,
 249                     note='Acquiring permission for downloading video', fatal=False,
 250                     headers=self._API_HEADERS)
 251                 if traverse_obj(watch_request_response, ('meta', 'status')) != 200:
 252                     self.report_warning('Failed to acquire permission for playing video. Video download may fail.')
 253
 254         yesno = lambda x: 'yes' if x else 'no'
 255
 256         if dmc_protocol == 'http':
 257             protocol = 'http'
 258             protocol_parameters = {
 259                 'http_output_download_parameters': {
 260                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 261                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 262                 }
 263             }
 264         elif dmc_protocol == 'hls':
 265             protocol = 'm3u8'
 266             segment_duration = try_get(self._configuration_arg('segment_duration'), lambda x: int(x[0])) or 6000
 267             parsed_token = self._parse_json(session_api_data['token'], video_id)
 268             encryption = traverse_obj(api_data, ('media', 'delivery', 'encryption'))
 269             protocol_parameters = {
 270                 'hls_parameters': {
 271                     'segment_duration': segment_duration,
 272                     'transfer_preset': '',
 273                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 274                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 275                 }
 276             }
 277             if 'hls_encryption' in parsed_token and encryption:
 278                 protocol_parameters['hls_parameters']['encryption'] = {
 279                     parsed_token['hls_encryption']: {
 280                         'encrypted_key': encryption['encryptedKey'],
 281                         'key_uri': encryption['keyUri'],
 282                     }
 283                 }
 284             else:
 285                 protocol = 'm3u8_native'
 286         else:
 287             raise ExtractorError(f'Unsupported DMC protocol: {dmc_protocol}')
 288
 289         session_response = self._download_json(
 290             session_api_endpoint['url'], video_id,
 291             query={'_format': 'json'},
 292             headers={'Content-Type': 'application/json'},
 293             note='Downloading JSON metadata for %s' % info_dict['format_id'],
 294             data=json.dumps({
 295                 'session': {
 296                     'client_info': {
 297                         'player_id': session_api_data.get('playerId'),
 298                     },
 299                     'content_auth': {
 300                         'auth_type': try_get(session_api_data, lambda x: x['authTypes'][session_api_data['protocols'][0]]),
 301                         'content_key_timeout': session_api_data.get('contentKeyTimeout'),
 302                         'service_id': 'nicovideo',
 303                         'service_user_id': session_api_data.get('serviceUserId')
 304                     },
 305                     'content_id': session_api_data.get('contentId'),
 306                     'content_src_id_sets': [{
 307                         'content_src_ids': [{
 308                             'src_id_to_mux': {
 309                                 'audio_src_ids': [audio_src_id],
 310                                 'video_src_ids': [video_src_id],
 311                             }
 312                         }]
 313                     }],
 314                     'content_type': 'movie',
 315                     'content_uri': '',
 316                     'keep_method': {
 317                         'heartbeat': {
 318                             'lifetime': session_api_data.get('heartbeatLifetime')
 319                         }
 320                     },
 321                     'priority': session_api_data['priority'],
 322                     'protocol': {
 323                         'name': 'http',
 324                         'parameters': {
 325                             'http_parameters': {
 326                                 'parameters': protocol_parameters
 327                             }
 328                         }
 329                     },
 330                     'recipe_id': session_api_data.get('recipeId'),
 331                     'session_operation_auth': {
 332                         'session_operation_auth_by_signature': {
 333                             'signature': session_api_data.get('signature'),
 334                             'token': session_api_data.get('token'),
 335                         }
 336                     },
 337                     'timing_constraint': 'unlimited'
 338                 }
 339             }).encode())
 340
 341         info_dict['url'] = session_response['data']['session']['content_uri']
 342         info_dict['protocol'] = protocol
 343
 344         # get heartbeat info
 345         heartbeat_info_dict = {
 346             'url': session_api_endpoint['url'] + '/' + session_response['data']['session']['id'] + '?_format=json&_method=PUT',
 347             'data': json.dumps(session_response['data']),
 348             # interval, convert milliseconds to seconds, then halve to make a buffer.
 349             'interval': float_or_none(session_api_data.get('heartbeatLifetime'), scale=3000),
 350             'ping': ping
 351         }
 352
 353         return info_dict, heartbeat_info_dict
 354
 355     def _extract_format_for_quality(self, video_id, audio_quality, video_quality, dmc_protocol):
 356
 357         if not audio_quality.get('isAvailable') or not video_quality.get('isAvailable'):
 358             return None
 359
 360         def extract_video_quality(video_quality):
 361             return parse_filesize('%sB' % self._search_regex(
 362                 r'\| ([0-9]*\.?[0-9]*[MK])', video_quality, 'vbr', default=''))
 363
 364         format_id = '-'.join(
 365             [remove_start(s['id'], 'archive_') for s in (video_quality, audio_quality)] + [dmc_protocol])
 366
 367         vid_qual_label = traverse_obj(video_quality, ('metadata', 'label'))
 368         vid_quality = traverse_obj(video_quality, ('metadata', 'bitrate'))
 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             'abr': float_or_none(traverse_obj(audio_quality, ('metadata', 'bitrate')), 1000),
 378             'vbr': float_or_none(vid_quality if vid_quality > 0 else extract_video_quality(vid_qual_label), 1000),
 379             'height': traverse_obj(video_quality, ('metadata', 'resolution', 'height')),
 380             'width': traverse_obj(video_quality, ('metadata', 'resolution', 'width')),
 381             'quality': -2 if 'low' in video_quality['id'] else None,
 382             'protocol': 'niconico_dmc',
 383             'expected_protocol': dmc_protocol,  # XXX: This is not a documented field
 384             'http_headers': {
 385                 'Origin': 'https://www.nicovideo.jp',
 386                 'Referer': 'https://www.nicovideo.jp/watch/' + video_id,
 387             }
 388         }
 389
 390     def _real_extract(self, url):
 391         video_id = self._match_id(url)
 392
 393         try:
 394             webpage, handle = self._download_webpage_handle(
 395                 'https://www.nicovideo.jp/watch/' + video_id, video_id)
 396             if video_id.startswith('so'):
 397                 video_id = self._match_id(handle.url)
 398
 399             api_data = self._parse_json(self._html_search_regex(
 400                 'data-api-data="([^"]+)"', webpage,
 401                 'API data', default='{}'), video_id)
 402         except ExtractorError as e:
 403             try:
 404                 api_data = self._download_json(
 405                     'https://www.nicovideo.jp/api/watch/v3/%s?_frontendId=6&_frontendVersion=0&actionTrackId=AAAAAAAAAA_%d' % (video_id, round(time.time() * 1000)), video_id,
 406                     note='Downloading API JSON', errnote='Unable to fetch data')['data']
 407             except ExtractorError:
 408                 if not isinstance(e.cause, HTTPError):
 409                     raise
 410                 webpage = e.cause.response.read().decode('utf-8', 'replace')
 411                 error_msg = self._html_search_regex(
 412                     r'(?s)<section\s+class="(?:(?:ErrorMessage|WatchExceptionPage-message)\s*)+">(.+?)</section>',
 413                     webpage, 'error reason', default=None)
 414                 if not error_msg:
 415                     raise
 416                 raise ExtractorError(re.sub(r'\s+', ' ', error_msg), expected=True)
 417
 418         formats = []
 419
 420         def get_video_info(*items, get_first=True, **kwargs):
 421             return traverse_obj(api_data, ('video', *items), get_all=not get_first, **kwargs)
 422
 423         quality_info = api_data['media']['delivery']['movie']
 424         session_api_data = quality_info['session']
 425         for (audio_quality, video_quality, protocol) in itertools.product(quality_info['audios'], quality_info['videos'], session_api_data['protocols']):
 426             fmt = self._extract_format_for_quality(video_id, audio_quality, video_quality, protocol)
 427             if fmt:
 428                 formats.append(fmt)
 429
 430         # Start extracting information
 431         tags = None
 432         if webpage:
 433             # use og:video:tag (not logged in)
 434             og_video_tags = re.finditer(r'<meta\s+property="og:video:tag"\s*content="(.*?)">', webpage)
 435             tags = list(filter(None, (clean_html(x.group(1)) for x in og_video_tags)))
 436             if not tags:
 437                 # use keywords and split with comma (not logged in)
 438                 kwds = self._html_search_meta('keywords', webpage, default=None)
 439                 if kwds:
 440                     tags = [x for x in kwds.split(',') if x]
 441         if not tags:
 442             # find in json (logged in)
 443             tags = traverse_obj(api_data, ('tag', 'items', ..., 'name'))
 444
 445         thumb_prefs = qualities(['url', 'middleUrl', 'largeUrl', 'player', 'ogp'])
 446
 447         return {
 448             'id': video_id,
 449             '_api_data': api_data,
 450             'title': get_video_info(('originalTitle', 'title')) or self._og_search_title(webpage, default=None),
 451             'formats': formats,
 452             'thumbnails': [{
 453                 'id': key,
 454                 'url': url,
 455                 'ext': 'jpg',
 456                 'preference': thumb_prefs(key),
 457                 **parse_resolution(url, lenient=True),
 458             } for key, url in (get_video_info('thumbnail') or {}).items() if url],
 459             'description': clean_html(get_video_info('description')),
 460             'uploader': traverse_obj(api_data, ('owner', 'nickname'), ('channel', 'name'), ('community', 'name')),
 461             'uploader_id': str_or_none(traverse_obj(api_data, ('owner', 'id'), ('channel', 'id'), ('community', 'id'))),
 462             'timestamp': parse_iso8601(get_video_info('registeredAt')) or parse_iso8601(
 463                 self._html_search_meta('video:release_date', webpage, 'date published', default=None)),
 464             'channel': traverse_obj(api_data, ('channel', 'name'), ('community', 'name')),
 465             'channel_id': traverse_obj(api_data, ('channel', 'id'), ('community', 'id')),
 466             'view_count': int_or_none(get_video_info('count', 'view')),
 467             'tags': tags,
 468             'genre': traverse_obj(api_data, ('genre', 'label'), ('genre', 'key')),
 469             'comment_count': get_video_info('count', 'comment', expected_type=int),
 470             'duration': (
 471                 parse_duration(self._html_search_meta('video:duration', webpage, 'video duration', default=None))
 472                 or get_video_info('duration')),
 473             'webpage_url': url_or_none(url) or f'https://www.nicovideo.jp/watch/{video_id}',
 474             'subtitles': self.extract_subtitles(video_id, api_data, session_api_data),
 475         }
 476
 477     def _get_subtitles(self, video_id, api_data, session_api_data):
 478         comment_user_key = traverse_obj(api_data, ('comment', 'keys', 'userKey'))
 479         user_id_str = session_api_data.get('serviceUserId')
 480
 481         thread_ids = traverse_obj(api_data, ('comment', 'threads', lambda _, v: v['isActive']))
 482         legacy_danmaku = self._extract_legacy_comments(video_id, thread_ids, user_id_str, comment_user_key) or []
 483
 484         new_comments = traverse_obj(api_data, ('comment', 'nvComment'))
 485         new_danmaku = self._extract_new_comments(
 486             new_comments.get('server'), video_id,
 487             new_comments.get('params'), new_comments.get('threadKey'))
 488
 489         if not legacy_danmaku and not new_danmaku:
 490             self.report_warning(f'Failed to get comments. {bug_reports_message()}')
 491             return
 492
 493         return {
 494             'comments': [{
 495                 'ext': 'json',
 496                 'data': json.dumps(legacy_danmaku + new_danmaku),
 497             }],
 498         }
 499
 500     def _extract_legacy_comments(self, video_id, threads, user_id, user_key):
 501         auth_data = {
 502             'user_id': user_id,
 503             'userkey': user_key,
 504         } if user_id and user_key else {'user_id': ''}
 505
 506         api_url = traverse_obj(threads, (..., 'server'), get_all=False)
 507
 508         # Request Start
 509         post_data = [{'ping': {'content': 'rs:0'}}]
 510         for i, thread in enumerate(threads):
 511             thread_id = thread['id']
 512             thread_fork = thread['fork']
 513             # Post Start (2N)
 514             post_data.append({'ping': {'content': f'ps:{i * 2}'}})
 515             post_data.append({'thread': {
 516                 'fork': thread_fork,
 517                 'language': 0,
 518                 'nicoru': 3,
 519                 'scores': 1,
 520                 'thread': thread_id,
 521                 'version': '20090904',
 522                 'with_global': 1,
 523                 **auth_data,
 524             }})
 525             # Post Final (2N)
 526             post_data.append({'ping': {'content': f'pf:{i * 2}'}})
 527
 528             # Post Start (2N+1)
 529             post_data.append({'ping': {'content': f'ps:{i * 2 + 1}'}})
 530             post_data.append({'thread_leaves': {
 531                 # format is '<bottom of minute range>-<top of minute range>:<comments per minute>,<total last comments'
 532                 # unfortunately NND limits (deletes?) comment returns this way, so you're only able to grab the last 1000 per language
 533                 'content': '0-999999:999999,999999,nicoru:999999',
 534                 'fork': thread_fork,
 535                 'language': 0,
 536                 'nicoru': 3,
 537                 'scores': 1,
 538                 'thread': thread_id,
 539                 **auth_data,
 540             }})
 541             # Post Final (2N+1)
 542             post_data.append({'ping': {'content': f'pf:{i * 2 + 1}'}})
 543         # Request Final
 544         post_data.append({'ping': {'content': 'rf:0'}})
 545
 546         return self._download_json(
 547             f'{api_url}/api.json', video_id, data=json.dumps(post_data).encode(), fatal=False,
 548             headers={
 549                 'Referer': f'https://www.nicovideo.jp/watch/{video_id}',
 550                 'Origin': 'https://www.nicovideo.jp',
 551                 'Content-Type': 'text/plain;charset=UTF-8',
 552             },
 553             note='Downloading comments', errnote=f'Failed to access endpoint {api_url}')
 554
 555     def _extract_new_comments(self, endpoint, video_id, params, thread_key):
 556         comments = self._download_json(
 557             f'{endpoint}/v1/threads', video_id, data=json.dumps({
 558                 'additionals': {},
 559                 'params': params,
 560                 'threadKey': thread_key,
 561             }).encode(), fatal=False,
 562             headers={
 563                 'Referer': 'https://www.nicovideo.jp/',
 564                 'Origin': 'https://www.nicovideo.jp',
 565                 'Content-Type': 'text/plain;charset=UTF-8',
 566                 'x-client-os-type': 'others',
 567                 'x-frontend-id': '6',
 568                 'x-frontend-version': '0',
 569             },
 570             note='Downloading comments (new)', errnote='Failed to download comments (new)')
 571         return traverse_obj(comments, ('data', 'threads', ..., 'comments', ...))
 572
 573
 574 class NiconicoPlaylistBaseIE(InfoExtractor):
 575     _PAGE_SIZE = 100
 576
 577     _API_HEADERS = {
 578         'X-Frontend-ID': '6',
 579         'X-Frontend-Version': '0',
 580         'X-Niconico-Language': 'en-us'
 581     }
 582
 583     def _call_api(self, list_id, resource, query):
 584         raise NotImplementedError('Must be implemented in subclasses')
 585
 586     @staticmethod
 587     def _parse_owner(item):
 588         return {
 589             'uploader': traverse_obj(item, ('owner', 'name')),
 590             'uploader_id': traverse_obj(item, ('owner', 'id')),
 591         }
 592
 593     def _fetch_page(self, list_id, page):
 594         page += 1
 595         resp = self._call_api(list_id, 'page %d' % page, {
 596             'page': page,
 597             'pageSize': self._PAGE_SIZE,
 598         })
 599         # this is needed to support both mylist and user
 600         for video in traverse_obj(resp, ('items', ..., ('video', None))) or []:
 601             video_id = video.get('id')
 602             if not video_id:
 603                 # skip {"video": {"id": "blablabla", ...}}
 604                 continue
 605             count = video.get('count') or {}
 606             get_count = lambda x: int_or_none(count.get(x))
 607             yield {
 608                 '_type': 'url',
 609                 'id': video_id,
 610                 'title': video.get('title'),
 611                 'url': f'https://www.nicovideo.jp/watch/{video_id}',
 612                 'description': video.get('shortDescription'),
 613                 'duration': int_or_none(video.get('duration')),
 614                 'view_count': get_count('view'),
 615                 'comment_count': get_count('comment'),
 616                 'thumbnail': traverse_obj(video, ('thumbnail', ('nHdUrl', 'largeUrl', 'listingUrl', 'url'))),
 617                 'ie_key': NiconicoIE.ie_key(),
 618                 **self._parse_owner(video),
 619             }
 620
 621     def _entries(self, list_id):
 622         return OnDemandPagedList(functools.partial(self._fetch_page, list_id), self._PAGE_SIZE)
 623
 624
 625 class NiconicoPlaylistIE(NiconicoPlaylistBaseIE):
 626     IE_NAME = 'niconico:playlist'
 627     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp|nico\.ms)/(?:user/\d+/)?(?:my/)?mylist/(?:#/)?(?P<id>\d+)'
 628
 629     _TESTS = [{
 630         'url': 'http://www.nicovideo.jp/mylist/27411728',
 631         'info_dict': {
 632             'id': '27411728',
 633             'title': 'AKB48のオールナイトニッポン',
 634             'description': 'md5:d89694c5ded4b6c693dea2db6e41aa08',
 635             'uploader': 'のっく',
 636             'uploader_id': '805442',
 637         },
 638         'playlist_mincount': 291,
 639     }, {
 640         'url': 'https://www.nicovideo.jp/user/805442/mylist/27411728',
 641         'only_matching': True,
 642     }, {
 643         'url': 'https://www.nicovideo.jp/my/mylist/#/68048635',
 644         'only_matching': True,
 645     }]
 646
 647     def _call_api(self, list_id, resource, query):
 648         return self._download_json(
 649             f'https://nvapi.nicovideo.jp/v2/mylists/{list_id}', list_id,
 650             f'Downloading {resource}', query=query,
 651             headers=self._API_HEADERS)['data']['mylist']
 652
 653     def _real_extract(self, url):
 654         list_id = self._match_id(url)
 655         mylist = self._call_api(list_id, 'list', {
 656             'pageSize': 1,
 657         })
 658         return self.playlist_result(
 659             self._entries(list_id), list_id,
 660             mylist.get('name'), mylist.get('description'), **self._parse_owner(mylist))
 661
 662
 663 class NiconicoSeriesIE(InfoExtractor):
 664     IE_NAME = 'niconico:series'
 665     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp(?:/user/\d+)?|nico\.ms)/series/(?P<id>\d+)'
 666
 667     _TESTS = [{
 668         'url': 'https://www.nicovideo.jp/user/44113208/series/110226',
 669         'info_dict': {
 670             'id': '110226',
 671             'title': 'ご立派ァ！のシリーズ',
 672         },
 673         'playlist_mincount': 10,
 674     }, {
 675         'url': 'https://www.nicovideo.jp/series/12312/',
 676         'info_dict': {
 677             'id': '12312',
 678             'title': 'バトルスピリッツ　お勧めカード紹介(調整中)',
 679         },
 680         'playlist_mincount': 103,
 681     }, {
 682         'url': 'https://nico.ms/series/203559',
 683         'only_matching': True,
 684     }]
 685
 686     def _real_extract(self, url):
 687         list_id = self._match_id(url)
 688         webpage = self._download_webpage(url, list_id)
 689
 690         title = self._search_regex(
 691             (r'<title>「(.+)（全',
 692              r'<div class="TwitterShareButton"\s+data-text="(.+)\s+https:'),
 693             webpage, 'title', fatal=False)
 694         if title:
 695             title = unescapeHTML(title)
 696         json_data = next(self._yield_json_ld(webpage, None, fatal=False))
 697         return self.playlist_from_matches(
 698             traverse_obj(json_data, ('itemListElement', ..., 'url')), list_id, title, ie=NiconicoIE)
 699
 700
 701 class NiconicoHistoryIE(NiconicoPlaylistBaseIE):
 702     IE_NAME = 'niconico:history'
 703     IE_DESC = 'NicoNico user history or likes. Requires cookies.'
 704     _VALID_URL = r'https?://(?:www\.|sp\.)?nicovideo\.jp/my/(?P<id>history(?:/like)?)'
 705
 706     _TESTS = [{
 707         'note': 'PC page, with /video',
 708         'url': 'https://www.nicovideo.jp/my/history/video',
 709         'only_matching': True,
 710     }, {
 711         'note': 'PC page, without /video',
 712         'url': 'https://www.nicovideo.jp/my/history',
 713         'only_matching': True,
 714     }, {
 715         'note': 'mobile page, with /video',
 716         'url': 'https://sp.nicovideo.jp/my/history/video',
 717         'only_matching': True,
 718     }, {
 719         'note': 'mobile page, without /video',
 720         'url': 'https://sp.nicovideo.jp/my/history',
 721         'only_matching': True,
 722     }, {
 723         'note': 'PC page',
 724         'url': 'https://www.nicovideo.jp/my/history/like',
 725         'only_matching': True,
 726     }, {
 727         'note': 'Mobile page',
 728         'url': 'https://sp.nicovideo.jp/my/history/like',
 729         'only_matching': True,
 730     }]
 731
 732     def _call_api(self, list_id, resource, query):
 733         path = 'likes' if list_id == 'history/like' else 'watch/history'
 734         return self._download_json(
 735             f'https://nvapi.nicovideo.jp/v1/users/me/{path}', list_id,
 736             f'Downloading {resource}', query=query, headers=self._API_HEADERS)['data']
 737
 738     def _real_extract(self, url):
 739         list_id = self._match_id(url)
 740         try:
 741             mylist = self._call_api(list_id, 'list', {'pageSize': 1})
 742         except ExtractorError as e:
 743             if isinstance(e.cause, HTTPError) and e.cause.status == 401:
 744                 self.raise_login_required('You have to be logged in to get your history')
 745             raise
 746         return self.playlist_result(self._entries(list_id), list_id, **self._parse_owner(mylist))
 747
 748
 749 class NicovideoSearchBaseIE(InfoExtractor):
 750     _SEARCH_TYPE = 'search'
 751
 752     def _entries(self, url, item_id, query=None, note='Downloading page %(page)s'):
 753         query = query or {}
 754         pages = [query['page']] if 'page' in query else itertools.count(1)
 755         for page_num in pages:
 756             query['page'] = str(page_num)
 757             webpage = self._download_webpage(url, item_id, query=query, note=note % {'page': page_num})
 758             results = re.findall(r'(?<=data-video-id=)["\']?(?P<videoid>.*?)(?=["\'])', webpage)
 759             for item in results:
 760                 yield self.url_result(f'https://www.nicovideo.jp/watch/{item}', 'Niconico', item)
 761             if not results:
 762                 break
 763
 764     def _search_results(self, query):
 765         return self._entries(
 766             self._proto_relative_url(f'//www.nicovideo.jp/{self._SEARCH_TYPE}/{query}'), query)
 767
 768
 769 class NicovideoSearchIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 770     IE_DESC = 'Nico video search'
 771     IE_NAME = 'nicovideo:search'
 772     _SEARCH_KEY = 'nicosearch'
 773
 774
 775 class NicovideoSearchURLIE(NicovideoSearchBaseIE):
 776     IE_NAME = f'{NicovideoSearchIE.IE_NAME}_url'
 777     IE_DESC = 'Nico video search URLs'
 778     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/search/(?P<id>[^?#&]+)?'
 779     _TESTS = [{
 780         'url': 'http://www.nicovideo.jp/search/sm9',
 781         'info_dict': {
 782             'id': 'sm9',
 783             'title': 'sm9'
 784         },
 785         'playlist_mincount': 40,
 786     }, {
 787         'url': 'https://www.nicovideo.jp/search/sm9?sort=h&order=d&end=2020-12-31&start=2020-01-01',
 788         'info_dict': {
 789             'id': 'sm9',
 790             'title': 'sm9'
 791         },
 792         'playlist_count': 31,
 793     }]
 794
 795     def _real_extract(self, url):
 796         query = self._match_id(url)
 797         return self.playlist_result(self._entries(url, query), query, query)
 798
 799
 800 class NicovideoSearchDateIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 801     IE_DESC = 'Nico video search, newest first'
 802     IE_NAME = f'{NicovideoSearchIE.IE_NAME}:date'
 803     _SEARCH_KEY = 'nicosearchdate'
 804     _TESTS = [{
 805         'url': 'nicosearchdateall:a',
 806         'info_dict': {
 807             'id': 'a',
 808             'title': 'a'
 809         },
 810         'playlist_mincount': 1610,
 811     }]
 812
 813     _START_DATE = datetime.date(2007, 1, 1)
 814     _RESULTS_PER_PAGE = 32
 815     _MAX_PAGES = 50
 816
 817     def _entries(self, url, item_id, start_date=None, end_date=None):
 818         start_date, end_date = start_date or self._START_DATE, end_date or datetime.datetime.now().date()
 819
 820         # If the last page has a full page of videos, we need to break down the query interval further
 821         last_page_len = len(list(self._get_entries_for_date(
 822             url, item_id, start_date, end_date, self._MAX_PAGES,
 823             note=f'Checking number of videos from {start_date} to {end_date}')))
 824         if (last_page_len == self._RESULTS_PER_PAGE and start_date != end_date):
 825             midpoint = start_date + ((end_date - start_date) // 2)
 826             yield from self._entries(url, item_id, midpoint, end_date)
 827             yield from self._entries(url, item_id, start_date, midpoint)
 828         else:
 829             self.to_screen(f'{item_id}: Downloading results from {start_date} to {end_date}')
 830             yield from self._get_entries_for_date(
 831                 url, item_id, start_date, end_date, note='    Downloading page %(page)s')
 832
 833     def _get_entries_for_date(self, url, item_id, start_date, end_date=None, page_num=None, note=None):
 834         query = {
 835             'start': str(start_date),
 836             'end': str(end_date or start_date),
 837             'sort': 'f',
 838             'order': 'd',
 839         }
 840         if page_num:
 841             query['page'] = str(page_num)
 842
 843         yield from super()._entries(url, item_id, query=query, note=note)
 844
 845
 846 class NicovideoTagURLIE(NicovideoSearchBaseIE):
 847     IE_NAME = 'niconico:tag'
 848     IE_DESC = 'NicoNico video tag URLs'
 849     _SEARCH_TYPE = 'tag'
 850     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/tag/(?P<id>[^?#&]+)?'
 851     _TESTS = [{
 852         'url': 'https://www.nicovideo.jp/tag/ドキュメンタリー淫夢',
 853         'info_dict': {
 854             'id': 'ドキュメンタリー淫夢',
 855             'title': 'ドキュメンタリー淫夢'
 856         },
 857         'playlist_mincount': 400,
 858     }]
 859
 860     def _real_extract(self, url):
 861         query = self._match_id(url)
 862         return self.playlist_result(self._entries(url, query), query, query)
 863
 864
 865 class NiconicoUserIE(InfoExtractor):
 866     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/user/(?P<id>\d+)/?(?:$|[#?])'
 867     _TEST = {
 868         'url': 'https://www.nicovideo.jp/user/419948',
 869         'info_dict': {
 870             'id': '419948',
 871         },
 872         'playlist_mincount': 101,
 873     }
 874     _API_URL = "https://nvapi.nicovideo.jp/v1/users/%s/videos?sortKey=registeredAt&sortOrder=desc&pageSize=%s&page=%s"
 875     _PAGE_SIZE = 100
 876
 877     _API_HEADERS = {
 878         'X-Frontend-ID': '6',
 879         'X-Frontend-Version': '0'
 880     }
 881
 882     def _entries(self, list_id):
 883         total_count = 1
 884         count = page_num = 0
 885         while count < total_count:
 886             json_parsed = self._download_json(
 887                 self._API_URL % (list_id, self._PAGE_SIZE, page_num + 1), list_id,
 888                 headers=self._API_HEADERS,
 889                 note='Downloading JSON metadata%s' % (' page %d' % page_num if page_num else ''))
 890             if not page_num:
 891                 total_count = int_or_none(json_parsed['data'].get('totalCount'))
 892             for entry in json_parsed["data"]["items"]:
 893                 count += 1
 894                 yield self.url_result('https://www.nicovideo.jp/watch/%s' % entry['id'])
 895             page_num += 1
 896
 897     def _real_extract(self, url):
 898         list_id = self._match_id(url)
 899         return self.playlist_result(self._entries(list_id), list_id, ie=NiconicoIE.ie_key())
 900
 901
 902 class NiconicoLiveIE(InfoExtractor):
 903     IE_NAME = 'niconico:live'
 904     IE_DESC = 'ニコニコ生放送'
 905     _VALID_URL = r'https?://(?:sp\.)?live2?\.nicovideo\.jp/(?:watch|gate)/(?P<id>lv\d+)'
 906     _TESTS = [{
 907         'note': 'this test case includes invisible characters for title, pasting them as-is',
 908         'url': 'https://live.nicovideo.jp/watch/lv339533123',
 909         'info_dict': {
 910             'id': 'lv339533123',
 911             'title': '激辛ペヤング食べます‪( ;ᯅ; )‬（歌枠オーディション参加中）',
 912             'view_count': 1526,
 913             'comment_count': 1772,
 914             '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',
 915             'uploader': 'もか',
 916             'channel': 'ゲストさんのコミュニティ',
 917             'channel_id': 'co5776900',
 918             'channel_url': 'https://com.nicovideo.jp/community/co5776900',
 919             'timestamp': 1670677328,
 920             'is_live': True,
 921         },
 922         'skip': 'livestream',
 923     }, {
 924         'url': 'https://live2.nicovideo.jp/watch/lv339533123',
 925         'only_matching': True,
 926     }, {
 927         'url': 'https://sp.live.nicovideo.jp/watch/lv339533123',
 928         'only_matching': True,
 929     }, {
 930         'url': 'https://sp.live2.nicovideo.jp/watch/lv339533123',
 931         'only_matching': True,
 932     }]
 933
 934     _KNOWN_LATENCY = ('high', 'low')
 935
 936     def _real_extract(self, url):
 937         if not websockets:
 938             raise ExtractorError('websockets library is not available. Please install it.', expected=True)
 939         video_id = self._match_id(url)
 940         webpage, urlh = self._download_webpage_handle(f'https://live.nicovideo.jp/watch/{video_id}', video_id)
 941
 942         embedded_data = self._parse_json(unescapeHTML(self._search_regex(
 943             r'<script\s+id="embedded-data"\s*data-props="(.+?)"', webpage, 'embedded data')), video_id)
 944
 945         ws_url = traverse_obj(embedded_data, ('site', 'relive', 'webSocketUrl'))
 946         if not ws_url:
 947             raise ExtractorError('The live hasn\'t started yet or already ended.', expected=True)
 948         ws_url = update_url_query(ws_url, {
 949             'frontend_id': traverse_obj(embedded_data, ('site', 'frontendId')) or '9',
 950         })
 951
 952         hostname = remove_start(urlparse(urlh.url).hostname, 'sp.')
 953         cookies = try_get(urlh.url, self._downloader._calc_cookies)
 954         latency = try_get(self._configuration_arg('latency'), lambda x: x[0])
 955         if latency not in self._KNOWN_LATENCY:
 956             latency = 'high'
 957
 958         ws = WebSocketsWrapper(ws_url, {
 959             'Cookies': str_or_none(cookies) or '',
 960             'Origin': f'https://{hostname}',
 961             'Accept': '*/*',
 962             'User-Agent': self.get_param('http_headers')['User-Agent'],
 963         })
 964
 965         self.write_debug('[debug] Sending HLS server request')
 966         ws.send(json.dumps({
 967             'type': 'startWatching',
 968             'data': {
 969                 'stream': {
 970                     'quality': 'abr',
 971                     'protocol': 'hls+fmp4',
 972                     'latency': latency,
 973                     'chasePlay': False
 974                 },
 975                 'room': {
 976                     'protocol': 'webSocket',
 977                     'commentable': True
 978                 },
 979                 'reconnect': False,
 980             }
 981         }))
 982
 983         while True:
 984             recv = ws.recv()
 985             if not recv:
 986                 continue
 987             data = json.loads(recv)
 988             if not isinstance(data, dict):
 989                 continue
 990             if data.get('type') == 'stream':
 991                 m3u8_url = data['data']['uri']
 992                 qualities = data['data']['availableQualities']
 993                 break
 994             elif data.get('type') == 'disconnect':
 995                 self.write_debug(recv)
 996                 raise ExtractorError('Disconnected at middle of extraction')
 997             elif data.get('type') == 'error':
 998                 self.write_debug(recv)
 999                 message = traverse_obj(data, ('body', 'code')) or recv
1000                 raise ExtractorError(message)
1001             elif self.get_param('verbose', False):
1002                 if len(recv) > 100:
1003                     recv = recv[:100] + '...'
1004                 self.write_debug('Server said: %s' % recv)
1005
1006         title = traverse_obj(embedded_data, ('program', 'title')) or self._html_search_meta(
1007             ('og:title', 'twitter:title'), webpage, 'live title', fatal=False)
1008
1009         raw_thumbs = traverse_obj(embedded_data, ('program', 'thumbnail')) or {}
1010         thumbnails = []
1011         for name, value in raw_thumbs.items():
1012             if not isinstance(value, dict):
1013                 thumbnails.append({
1014                     'id': name,
1015                     'url': value,
1016                     **parse_resolution(value, lenient=True),
1017                 })
1018                 continue
1019
1020             for k, img_url in value.items():
1021                 res = parse_resolution(k, lenient=True) or parse_resolution(img_url, lenient=True)
1022                 width, height = res.get('width'), res.get('height')
1023
1024                 thumbnails.append({
1025                     'id': f'{name}_{width}x{height}',
1026                     'url': img_url,
1027                     **res,
1028                 })
1029
1030         formats = self._extract_m3u8_formats(m3u8_url, video_id, ext='mp4', live=True)
1031         for fmt, q in zip(formats, reversed(qualities[1:])):
1032             fmt.update({
1033                 'format_id': q,
1034                 'protocol': 'niconico_live',
1035                 'ws': ws,
1036                 'video_id': video_id,
1037                 'cookies': cookies,
1038                 'live_latency': latency,
1039                 'origin': hostname,
1040             })
1041
1042         return {
1043             'id': video_id,
1044             'title': title,
1045             **traverse_obj(embedded_data, {
1046                 'view_count': ('program', 'statistics', 'watchCount'),
1047                 'comment_count': ('program', 'statistics', 'commentCount'),
1048                 'uploader': ('program', 'supplier', 'name'),
1049                 'channel': ('socialGroup', 'name'),
1050                 'channel_id': ('socialGroup', 'id'),
1051                 'channel_url': ('socialGroup', 'socialGroupPageUrl'),
1052             }),
1053             'description': clean_html(traverse_obj(embedded_data, ('program', 'description'))),
1054             'timestamp': int_or_none(traverse_obj(embedded_data, ('program', 'openTime'))),
1055             'is_live': True,
1056             'thumbnails': thumbnails,
1057             'formats': formats,
1058         }