yt_dlp/extractor/niconico.py

   1 import datetime
   2 import functools
   3 import itertools
   4 import json
   5 import re
   6 import time
   7
   8 from urllib.parse import urlparse
   9
  10 from .common import InfoExtractor, SearchInfoExtractor
  11 from ..networking import Request
  12 from ..networking.exceptions import HTTPError
  13 from ..utils import (
  14     ExtractorError,
  15     OnDemandPagedList,
  16     bug_reports_message,
  17     clean_html,
  18     float_or_none,
  19     int_or_none,
  20     join_nonempty,
  21     parse_duration,
  22     parse_filesize,
  23     parse_iso8601,
  24     parse_resolution,
  25     qualities,
  26     remove_start,
  27     str_or_none,
  28     traverse_obj,
  29     try_get,
  30     unescapeHTML,
  31     update_url_query,
  32     url_or_none,
  33     urlencode_postdata,
  34     urljoin,
  35 )
  36
  37
  38 class NiconicoIE(InfoExtractor):
  39     IE_NAME = 'niconico'
  40     IE_DESC = 'ニコニコ動画'
  41
  42     _TESTS = [{
  43         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  44         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  45         'info_dict': {
  46             'id': 'sm22312215',
  47             'ext': 'mp4',
  48             'title': 'Big Buck Bunny',
  49             'thumbnail': r're:https?://.*',
  50             'uploader': 'takuya0301',
  51             'uploader_id': '2698420',
  52             'upload_date': '20131123',
  53             'timestamp': int,  # timestamp is unstable
  54             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  55             'duration': 33,
  56             'view_count': int,
  57             'comment_count': int,
  58         },
  59         'skip': 'Requires an account',
  60     }, {
  61         # File downloaded with and without credentials are different, so omit
  62         # the md5 field
  63         'url': 'http://www.nicovideo.jp/watch/nm14296458',
  64         'info_dict': {
  65             'id': 'nm14296458',
  66             'ext': 'swf',
  67             'title': '【鏡音リン】Dance on media【オリジナル】take2!',
  68             'description': 'md5:689f066d74610b3b22e0f1739add0f58',
  69             'thumbnail': r're:https?://.*',
  70             'uploader': 'りょうた',
  71             'uploader_id': '18822557',
  72             'upload_date': '20110429',
  73             'timestamp': 1304065916,
  74             'duration': 209,
  75         },
  76         'skip': 'Requires an account',
  77     }, {
  78         # 'video exists but is marked as "deleted"
  79         # md5 is unstable
  80         'url': 'http://www.nicovideo.jp/watch/sm10000',
  81         'info_dict': {
  82             'id': 'sm10000',
  83             'ext': 'unknown_video',
  84             'description': 'deleted',
  85             'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞',
  86             'thumbnail': r're:https?://.*',
  87             'upload_date': '20071224',
  88             'timestamp': int,  # timestamp field has different value if logged in
  89             'duration': 304,
  90             'view_count': int,
  91         },
  92         'skip': 'Requires an account',
  93     }, {
  94         'url': 'http://www.nicovideo.jp/watch/so22543406',
  95         'info_dict': {
  96             'id': '1388129933',
  97             'ext': 'mp4',
  98             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
  99             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
 100             'thumbnail': r're:https?://.*',
 101             'timestamp': 1388851200,
 102             'upload_date': '20140104',
 103             'uploader': 'アニメロチャンネル',
 104             'uploader_id': '312',
 105         },
 106         'skip': 'The viewing period of the video you were searching for has expired.',
 107     }, {
 108         # video not available via `getflv`; "old" HTML5 video
 109         'url': 'http://www.nicovideo.jp/watch/sm1151009',
 110         'md5': '8fa81c364eb619d4085354eab075598a',
 111         'info_dict': {
 112             'id': 'sm1151009',
 113             'ext': 'mp4',
 114             '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）',
 115             'description': 'md5:6ee077e0581ff5019773e2e714cdd0b7',
 116             'thumbnail': r're:https?://.*',
 117             'duration': 184,
 118             'timestamp': 1190868283,
 119             'upload_date': '20070927',
 120             'uploader': 'denden2',
 121             'uploader_id': '1392194',
 122             'view_count': int,
 123             'comment_count': int,
 124         },
 125         'skip': 'Requires an account',
 126     }, {
 127         # "New" HTML5 video
 128         # md5 is unstable
 129         'url': 'http://www.nicovideo.jp/watch/sm31464864',
 130         'info_dict': {
 131             'id': 'sm31464864',
 132             'ext': 'mp4',
 133             'title': '新作TVアニメ「戦姫絶唱シンフォギアAXZ」PV 最高画質',
 134             'description': 'md5:e52974af9a96e739196b2c1ca72b5feb',
 135             'timestamp': 1498514060,
 136             'upload_date': '20170626',
 137             'uploader': 'ゲスト',
 138             'uploader_id': '40826363',
 139             'thumbnail': r're:https?://.*',
 140             'duration': 198,
 141             'view_count': int,
 142             'comment_count': int,
 143         },
 144         'skip': 'Requires an account',
 145     }, {
 146         # Video without owner
 147         'url': 'http://www.nicovideo.jp/watch/sm18238488',
 148         'md5': 'd265680a1f92bdcbbd2a507fc9e78a9e',
 149         'info_dict': {
 150             'id': 'sm18238488',
 151             'ext': 'mp4',
 152             'title': '【実写版】ミュータントタートルズ',
 153             'description': 'md5:15df8988e47a86f9e978af2064bf6d8e',
 154             'timestamp': 1341160408,
 155             'upload_date': '20120701',
 156             'uploader': None,
 157             'uploader_id': None,
 158             'thumbnail': r're:https?://.*',
 159             'duration': 5271,
 160             'view_count': int,
 161             'comment_count': int,
 162         },
 163         'skip': 'Requires an account',
 164     }, {
 165         'url': 'http://sp.nicovideo.jp/watch/sm28964488?ss_pos=1&cp_in=wt_tg',
 166         'only_matching': True,
 167     }, {
 168         'note': 'a video that is only served as an ENCRYPTED HLS.',
 169         'url': 'https://www.nicovideo.jp/watch/so38016254',
 170         'only_matching': True,
 171     }]
 172
 173     _VALID_URL = r'https?://(?:(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch|nico\.ms)/(?P<id>(?:[a-z]{2})?[0-9]+)'
 174     _NETRC_MACHINE = 'niconico'
 175     _API_HEADERS = {
 176         'X-Frontend-ID': '6',
 177         'X-Frontend-Version': '0',
 178         'X-Niconico-Language': 'en-us',
 179         'Referer': 'https://www.nicovideo.jp/',
 180         'Origin': 'https://www.nicovideo.jp',
 181     }
 182
 183     def _perform_login(self, username, password):
 184         login_ok = True
 185         login_form_strs = {
 186             'mail_tel': username,
 187             'password': password,
 188         }
 189         self._request_webpage(
 190             'https://account.nicovideo.jp/login', None,
 191             note='Acquiring Login session')
 192         page = self._download_webpage(
 193             'https://account.nicovideo.jp/login/redirector?show_button_twitter=1&site=niconico&show_button_facebook=1', None,
 194             note='Logging in', errnote='Unable to log in',
 195             data=urlencode_postdata(login_form_strs),
 196             headers={
 197                 'Referer': 'https://account.nicovideo.jp/login',
 198                 'Content-Type': 'application/x-www-form-urlencoded',
 199             })
 200         if 'oneTimePw' in page:
 201             post_url = self._search_regex(
 202                 r'<form[^>]+action=(["\'])(?P<url>.+?)\1', page, 'post url', group='url')
 203             page = self._download_webpage(
 204                 urljoin('https://account.nicovideo.jp', post_url), None,
 205                 note='Performing MFA', errnote='Unable to complete MFA',
 206                 data=urlencode_postdata({
 207                     'otp': self._get_tfa_info('6 digits code')
 208                 }), headers={
 209                     'Content-Type': 'application/x-www-form-urlencoded',
 210                 })
 211             if 'oneTimePw' in page or 'formError' in page:
 212                 err_msg = self._html_search_regex(
 213                     r'formError["\']+>(.*?)</div>', page, 'form_error',
 214                     default='There\'s an error but the message can\'t be parsed.',
 215                     flags=re.DOTALL)
 216                 self.report_warning(f'Unable to log in: MFA challenge failed, "{err_msg}"')
 217                 return False
 218         login_ok = 'class="notice error"' not in page
 219         if not login_ok:
 220             self.report_warning('Unable to log in: bad username or password')
 221         return login_ok
 222
 223     def _get_heartbeat_info(self, info_dict):
 224         video_id, video_src_id, audio_src_id = info_dict['url'].split(':')[1].split('/')
 225         dmc_protocol = info_dict['expected_protocol']
 226
 227         api_data = (
 228             info_dict.get('_api_data')
 229             or self._parse_json(
 230                 self._html_search_regex(
 231                     'data-api-data="([^"]+)"',
 232                     self._download_webpage('https://www.nicovideo.jp/watch/' + video_id, video_id),
 233                     'API data', default='{}'),
 234                 video_id))
 235
 236         session_api_data = try_get(api_data, lambda x: x['media']['delivery']['movie']['session'])
 237         session_api_endpoint = try_get(session_api_data, lambda x: x['urls'][0])
 238
 239         def ping():
 240             tracking_id = traverse_obj(api_data, ('media', 'delivery', 'trackingId'))
 241             if tracking_id:
 242                 tracking_url = update_url_query('https://nvapi.nicovideo.jp/v1/2ab0cbaa/watch', {'t': tracking_id})
 243                 watch_request_response = self._download_json(
 244                     tracking_url, video_id,
 245                     note='Acquiring permission for downloading video', fatal=False,
 246                     headers=self._API_HEADERS)
 247                 if traverse_obj(watch_request_response, ('meta', 'status')) != 200:
 248                     self.report_warning('Failed to acquire permission for playing video. Video download may fail.')
 249
 250         yesno = lambda x: 'yes' if x else 'no'
 251
 252         if dmc_protocol == 'http':
 253             protocol = 'http'
 254             protocol_parameters = {
 255                 'http_output_download_parameters': {
 256                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 257                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 258                 }
 259             }
 260         elif dmc_protocol == 'hls':
 261             protocol = 'm3u8'
 262             segment_duration = try_get(self._configuration_arg('segment_duration'), lambda x: int(x[0])) or 6000
 263             parsed_token = self._parse_json(session_api_data['token'], video_id)
 264             encryption = traverse_obj(api_data, ('media', 'delivery', 'encryption'))
 265             protocol_parameters = {
 266                 'hls_parameters': {
 267                     'segment_duration': segment_duration,
 268                     'transfer_preset': '',
 269                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 270                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 271                 }
 272             }
 273             if 'hls_encryption' in parsed_token and encryption:
 274                 protocol_parameters['hls_parameters']['encryption'] = {
 275                     parsed_token['hls_encryption']: {
 276                         'encrypted_key': encryption['encryptedKey'],
 277                         'key_uri': encryption['keyUri'],
 278                     }
 279                 }
 280             else:
 281                 protocol = 'm3u8_native'
 282         else:
 283             raise ExtractorError(f'Unsupported DMC protocol: {dmc_protocol}')
 284
 285         session_response = self._download_json(
 286             session_api_endpoint['url'], video_id,
 287             query={'_format': 'json'},
 288             headers={'Content-Type': 'application/json'},
 289             note='Downloading JSON metadata for %s' % info_dict['format_id'],
 290             data=json.dumps({
 291                 'session': {
 292                     'client_info': {
 293                         'player_id': session_api_data.get('playerId'),
 294                     },
 295                     'content_auth': {
 296                         'auth_type': try_get(session_api_data, lambda x: x['authTypes'][session_api_data['protocols'][0]]),
 297                         'content_key_timeout': session_api_data.get('contentKeyTimeout'),
 298                         'service_id': 'nicovideo',
 299                         'service_user_id': session_api_data.get('serviceUserId')
 300                     },
 301                     'content_id': session_api_data.get('contentId'),
 302                     'content_src_id_sets': [{
 303                         'content_src_ids': [{
 304                             'src_id_to_mux': {
 305                                 'audio_src_ids': [audio_src_id],
 306                                 'video_src_ids': [video_src_id],
 307                             }
 308                         }]
 309                     }],
 310                     'content_type': 'movie',
 311                     'content_uri': '',
 312                     'keep_method': {
 313                         'heartbeat': {
 314                             'lifetime': session_api_data.get('heartbeatLifetime')
 315                         }
 316                     },
 317                     'priority': session_api_data['priority'],
 318                     'protocol': {
 319                         'name': 'http',
 320                         'parameters': {
 321                             'http_parameters': {
 322                                 'parameters': protocol_parameters
 323                             }
 324                         }
 325                     },
 326                     'recipe_id': session_api_data.get('recipeId'),
 327                     'session_operation_auth': {
 328                         'session_operation_auth_by_signature': {
 329                             'signature': session_api_data.get('signature'),
 330                             'token': session_api_data.get('token'),
 331                         }
 332                     },
 333                     'timing_constraint': 'unlimited'
 334                 }
 335             }).encode())
 336
 337         info_dict['url'] = session_response['data']['session']['content_uri']
 338         info_dict['protocol'] = protocol
 339
 340         # get heartbeat info
 341         heartbeat_info_dict = {
 342             'url': session_api_endpoint['url'] + '/' + session_response['data']['session']['id'] + '?_format=json&_method=PUT',
 343             'data': json.dumps(session_response['data']),
 344             # interval, convert milliseconds to seconds, then halve to make a buffer.
 345             'interval': float_or_none(session_api_data.get('heartbeatLifetime'), scale=3000),
 346             'ping': ping
 347         }
 348
 349         return info_dict, heartbeat_info_dict
 350
 351     def _extract_format_for_quality(self, video_id, audio_quality, video_quality, dmc_protocol):
 352
 353         if not audio_quality.get('isAvailable') or not video_quality.get('isAvailable'):
 354             return None
 355
 356         def extract_video_quality(video_quality):
 357             return parse_filesize('%sB' % self._search_regex(
 358                 r'\| ([0-9]*\.?[0-9]*[MK])', video_quality, 'vbr', default=''))
 359
 360         format_id = '-'.join(
 361             [remove_start(s['id'], 'archive_') for s in (video_quality, audio_quality)] + [dmc_protocol])
 362
 363         vid_qual_label = traverse_obj(video_quality, ('metadata', 'label'))
 364         vid_quality = traverse_obj(video_quality, ('metadata', 'bitrate'))
 365
 366         return {
 367             'url': 'niconico_dmc:%s/%s/%s' % (video_id, video_quality['id'], audio_quality['id']),
 368             'format_id': format_id,
 369             'format_note': join_nonempty('DMC', vid_qual_label, dmc_protocol.upper(), delim=' '),
 370             'ext': 'mp4',  # Session API are used in HTML5, which always serves mp4
 371             'acodec': 'aac',
 372             'vcodec': 'h264',
 373             'abr': float_or_none(traverse_obj(audio_quality, ('metadata', 'bitrate')), 1000),
 374             'vbr': float_or_none(vid_quality if vid_quality > 0 else extract_video_quality(vid_qual_label), 1000),
 375             'height': traverse_obj(video_quality, ('metadata', 'resolution', 'height')),
 376             'width': traverse_obj(video_quality, ('metadata', 'resolution', 'width')),
 377             'quality': -2 if 'low' in video_quality['id'] else None,
 378             'protocol': 'niconico_dmc',
 379             'expected_protocol': dmc_protocol,  # XXX: This is not a documented field
 380             'http_headers': {
 381                 'Origin': 'https://www.nicovideo.jp',
 382                 'Referer': 'https://www.nicovideo.jp/watch/' + video_id,
 383             }
 384         }
 385
 386     def _real_extract(self, url):
 387         video_id = self._match_id(url)
 388
 389         try:
 390             webpage, handle = self._download_webpage_handle(
 391                 'https://www.nicovideo.jp/watch/' + video_id, video_id)
 392             if video_id.startswith('so'):
 393                 video_id = self._match_id(handle.url)
 394
 395             api_data = self._parse_json(self._html_search_regex(
 396                 'data-api-data="([^"]+)"', webpage,
 397                 'API data', default='{}'), video_id)
 398         except ExtractorError as e:
 399             try:
 400                 api_data = self._download_json(
 401                     'https://www.nicovideo.jp/api/watch/v3/%s?_frontendId=6&_frontendVersion=0&actionTrackId=AAAAAAAAAA_%d' % (video_id, round(time.time() * 1000)), video_id,
 402                     note='Downloading API JSON', errnote='Unable to fetch data')['data']
 403             except ExtractorError:
 404                 if not isinstance(e.cause, HTTPError):
 405                     raise
 406                 webpage = e.cause.response.read().decode('utf-8', 'replace')
 407                 error_msg = self._html_search_regex(
 408                     r'(?s)<section\s+class="(?:(?:ErrorMessage|WatchExceptionPage-message)\s*)+">(.+?)</section>',
 409                     webpage, 'error reason', default=None)
 410                 if not error_msg:
 411                     raise
 412                 raise ExtractorError(re.sub(r'\s+', ' ', error_msg), expected=True)
 413
 414         formats = []
 415
 416         def get_video_info(*items, get_first=True, **kwargs):
 417             return traverse_obj(api_data, ('video', *items), get_all=not get_first, **kwargs)
 418
 419         quality_info = api_data['media']['delivery']['movie']
 420         session_api_data = quality_info['session']
 421         for (audio_quality, video_quality, protocol) in itertools.product(quality_info['audios'], quality_info['videos'], session_api_data['protocols']):
 422             fmt = self._extract_format_for_quality(video_id, audio_quality, video_quality, protocol)
 423             if fmt:
 424                 formats.append(fmt)
 425
 426         # Start extracting information
 427         tags = None
 428         if webpage:
 429             # use og:video:tag (not logged in)
 430             og_video_tags = re.finditer(r'<meta\s+property="og:video:tag"\s*content="(.*?)">', webpage)
 431             tags = list(filter(None, (clean_html(x.group(1)) for x in og_video_tags)))
 432             if not tags:
 433                 # use keywords and split with comma (not logged in)
 434                 kwds = self._html_search_meta('keywords', webpage, default=None)
 435                 if kwds:
 436                     tags = [x for x in kwds.split(',') if x]
 437         if not tags:
 438             # find in json (logged in)
 439             tags = traverse_obj(api_data, ('tag', 'items', ..., 'name'))
 440
 441         thumb_prefs = qualities(['url', 'middleUrl', 'largeUrl', 'player', 'ogp'])
 442
 443         return {
 444             'id': video_id,
 445             '_api_data': api_data,
 446             'title': get_video_info(('originalTitle', 'title')) or self._og_search_title(webpage, default=None),
 447             'formats': formats,
 448             'thumbnails': [{
 449                 'id': key,
 450                 'url': url,
 451                 'ext': 'jpg',
 452                 'preference': thumb_prefs(key),
 453                 **parse_resolution(url, lenient=True),
 454             } for key, url in (get_video_info('thumbnail') or {}).items() if url],
 455             'description': clean_html(get_video_info('description')),
 456             'uploader': traverse_obj(api_data, ('owner', 'nickname'), ('channel', 'name'), ('community', 'name')),
 457             'uploader_id': str_or_none(traverse_obj(api_data, ('owner', 'id'), ('channel', 'id'), ('community', 'id'))),
 458             'timestamp': parse_iso8601(get_video_info('registeredAt')) or parse_iso8601(
 459                 self._html_search_meta('video:release_date', webpage, 'date published', default=None)),
 460             'channel': traverse_obj(api_data, ('channel', 'name'), ('community', 'name')),
 461             'channel_id': traverse_obj(api_data, ('channel', 'id'), ('community', 'id')),
 462             'view_count': int_or_none(get_video_info('count', 'view')),
 463             'tags': tags,
 464             'genre': traverse_obj(api_data, ('genre', 'label'), ('genre', 'key')),
 465             'comment_count': get_video_info('count', 'comment', expected_type=int),
 466             'duration': (
 467                 parse_duration(self._html_search_meta('video:duration', webpage, 'video duration', default=None))
 468                 or get_video_info('duration')),
 469             'webpage_url': url_or_none(url) or f'https://www.nicovideo.jp/watch/{video_id}',
 470             'subtitles': self.extract_subtitles(video_id, api_data),
 471         }
 472
 473     def _get_subtitles(self, video_id, api_data):
 474         comments_info = traverse_obj(api_data, ('comment', 'nvComment', {dict})) or {}
 475         danmaku = traverse_obj(self._download_json(
 476             f'{comments_info.get("server")}/v1/threads', video_id, data=json.dumps({
 477                 'additionals': {},
 478                 'params': comments_info.get('params'),
 479                 'threadKey': comments_info.get('threadKey'),
 480             }).encode(), fatal=False,
 481             headers={
 482                 'Referer': 'https://www.nicovideo.jp/',
 483                 'Origin': 'https://www.nicovideo.jp',
 484                 'Content-Type': 'text/plain;charset=UTF-8',
 485                 'x-client-os-type': 'others',
 486                 'x-frontend-id': '6',
 487                 'x-frontend-version': '0',
 488             },
 489             note='Downloading comments', errnote='Failed to download comments'),
 490             ('data', 'threads', ..., 'comments', ...))
 491
 492         if not danmaku:
 493             self.report_warning(f'Failed to get comments. {bug_reports_message()}')
 494             return
 495
 496         return {
 497             'comments': [{
 498                 'ext': 'json',
 499                 'data': json.dumps(danmaku),
 500             }],
 501         }
 502
 503
 504 class NiconicoPlaylistBaseIE(InfoExtractor):
 505     _PAGE_SIZE = 100
 506
 507     _API_HEADERS = {
 508         'X-Frontend-ID': '6',
 509         'X-Frontend-Version': '0',
 510         'X-Niconico-Language': 'en-us'
 511     }
 512
 513     def _call_api(self, list_id, resource, query):
 514         raise NotImplementedError('Must be implemented in subclasses')
 515
 516     @staticmethod
 517     def _parse_owner(item):
 518         return {
 519             'uploader': traverse_obj(item, ('owner', 'name')),
 520             'uploader_id': traverse_obj(item, ('owner', 'id')),
 521         }
 522
 523     def _fetch_page(self, list_id, page):
 524         page += 1
 525         resp = self._call_api(list_id, 'page %d' % page, {
 526             'page': page,
 527             'pageSize': self._PAGE_SIZE,
 528         })
 529         # this is needed to support both mylist and user
 530         for video in traverse_obj(resp, ('items', ..., ('video', None))) or []:
 531             video_id = video.get('id')
 532             if not video_id:
 533                 # skip {"video": {"id": "blablabla", ...}}
 534                 continue
 535             count = video.get('count') or {}
 536             get_count = lambda x: int_or_none(count.get(x))
 537             yield {
 538                 '_type': 'url',
 539                 'id': video_id,
 540                 'title': video.get('title'),
 541                 'url': f'https://www.nicovideo.jp/watch/{video_id}',
 542                 'description': video.get('shortDescription'),
 543                 'duration': int_or_none(video.get('duration')),
 544                 'view_count': get_count('view'),
 545                 'comment_count': get_count('comment'),
 546                 'thumbnail': traverse_obj(video, ('thumbnail', ('nHdUrl', 'largeUrl', 'listingUrl', 'url'))),
 547                 'ie_key': NiconicoIE.ie_key(),
 548                 **self._parse_owner(video),
 549             }
 550
 551     def _entries(self, list_id):
 552         return OnDemandPagedList(functools.partial(self._fetch_page, list_id), self._PAGE_SIZE)
 553
 554
 555 class NiconicoPlaylistIE(NiconicoPlaylistBaseIE):
 556     IE_NAME = 'niconico:playlist'
 557     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp|nico\.ms)/(?:user/\d+/)?(?:my/)?mylist/(?:#/)?(?P<id>\d+)'
 558
 559     _TESTS = [{
 560         'url': 'http://www.nicovideo.jp/mylist/27411728',
 561         'info_dict': {
 562             'id': '27411728',
 563             'title': 'AKB48のオールナイトニッポン',
 564             'description': 'md5:d89694c5ded4b6c693dea2db6e41aa08',
 565             'uploader': 'のっく',
 566             'uploader_id': '805442',
 567         },
 568         'playlist_mincount': 291,
 569     }, {
 570         'url': 'https://www.nicovideo.jp/user/805442/mylist/27411728',
 571         'only_matching': True,
 572     }, {
 573         'url': 'https://www.nicovideo.jp/my/mylist/#/68048635',
 574         'only_matching': True,
 575     }]
 576
 577     def _call_api(self, list_id, resource, query):
 578         return self._download_json(
 579             f'https://nvapi.nicovideo.jp/v2/mylists/{list_id}', list_id,
 580             f'Downloading {resource}', query=query,
 581             headers=self._API_HEADERS)['data']['mylist']
 582
 583     def _real_extract(self, url):
 584         list_id = self._match_id(url)
 585         mylist = self._call_api(list_id, 'list', {
 586             'pageSize': 1,
 587         })
 588         return self.playlist_result(
 589             self._entries(list_id), list_id,
 590             mylist.get('name'), mylist.get('description'), **self._parse_owner(mylist))
 591
 592
 593 class NiconicoSeriesIE(InfoExtractor):
 594     IE_NAME = 'niconico:series'
 595     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp(?:/user/\d+)?|nico\.ms)/series/(?P<id>\d+)'
 596
 597     _TESTS = [{
 598         'url': 'https://www.nicovideo.jp/user/44113208/series/110226',
 599         'info_dict': {
 600             'id': '110226',
 601             'title': 'ご立派ァ！のシリーズ',
 602         },
 603         'playlist_mincount': 10,
 604     }, {
 605         'url': 'https://www.nicovideo.jp/series/12312/',
 606         'info_dict': {
 607             'id': '12312',
 608             'title': 'バトルスピリッツ　お勧めカード紹介(調整中)',
 609         },
 610         'playlist_mincount': 103,
 611     }, {
 612         'url': 'https://nico.ms/series/203559',
 613         'only_matching': True,
 614     }]
 615
 616     def _real_extract(self, url):
 617         list_id = self._match_id(url)
 618         webpage = self._download_webpage(url, list_id)
 619
 620         title = self._search_regex(
 621             (r'<title>「(.+)（全',
 622              r'<div class="TwitterShareButton"\s+data-text="(.+)\s+https:'),
 623             webpage, 'title', fatal=False)
 624         if title:
 625             title = unescapeHTML(title)
 626         json_data = next(self._yield_json_ld(webpage, None, fatal=False))
 627         return self.playlist_from_matches(
 628             traverse_obj(json_data, ('itemListElement', ..., 'url')), list_id, title, ie=NiconicoIE)
 629
 630
 631 class NiconicoHistoryIE(NiconicoPlaylistBaseIE):
 632     IE_NAME = 'niconico:history'
 633     IE_DESC = 'NicoNico user history or likes. Requires cookies.'
 634     _VALID_URL = r'https?://(?:www\.|sp\.)?nicovideo\.jp/my/(?P<id>history(?:/like)?)'
 635
 636     _TESTS = [{
 637         'note': 'PC page, with /video',
 638         'url': 'https://www.nicovideo.jp/my/history/video',
 639         'only_matching': True,
 640     }, {
 641         'note': 'PC page, without /video',
 642         'url': 'https://www.nicovideo.jp/my/history',
 643         'only_matching': True,
 644     }, {
 645         'note': 'mobile page, with /video',
 646         'url': 'https://sp.nicovideo.jp/my/history/video',
 647         'only_matching': True,
 648     }, {
 649         'note': 'mobile page, without /video',
 650         'url': 'https://sp.nicovideo.jp/my/history',
 651         'only_matching': True,
 652     }, {
 653         'note': 'PC page',
 654         'url': 'https://www.nicovideo.jp/my/history/like',
 655         'only_matching': True,
 656     }, {
 657         'note': 'Mobile page',
 658         'url': 'https://sp.nicovideo.jp/my/history/like',
 659         'only_matching': True,
 660     }]
 661
 662     def _call_api(self, list_id, resource, query):
 663         path = 'likes' if list_id == 'history/like' else 'watch/history'
 664         return self._download_json(
 665             f'https://nvapi.nicovideo.jp/v1/users/me/{path}', list_id,
 666             f'Downloading {resource}', query=query, headers=self._API_HEADERS)['data']
 667
 668     def _real_extract(self, url):
 669         list_id = self._match_id(url)
 670         try:
 671             mylist = self._call_api(list_id, 'list', {'pageSize': 1})
 672         except ExtractorError as e:
 673             if isinstance(e.cause, HTTPError) and e.cause.status == 401:
 674                 self.raise_login_required('You have to be logged in to get your history')
 675             raise
 676         return self.playlist_result(self._entries(list_id), list_id, **self._parse_owner(mylist))
 677
 678
 679 class NicovideoSearchBaseIE(InfoExtractor):
 680     _SEARCH_TYPE = 'search'
 681
 682     def _entries(self, url, item_id, query=None, note='Downloading page %(page)s'):
 683         query = query or {}
 684         pages = [query['page']] if 'page' in query else itertools.count(1)
 685         for page_num in pages:
 686             query['page'] = str(page_num)
 687             webpage = self._download_webpage(url, item_id, query=query, note=note % {'page': page_num})
 688             results = re.findall(r'(?<=data-video-id=)["\']?(?P<videoid>.*?)(?=["\'])', webpage)
 689             for item in results:
 690                 yield self.url_result(f'https://www.nicovideo.jp/watch/{item}', 'Niconico', item)
 691             if not results:
 692                 break
 693
 694     def _search_results(self, query):
 695         return self._entries(
 696             self._proto_relative_url(f'//www.nicovideo.jp/{self._SEARCH_TYPE}/{query}'), query)
 697
 698
 699 class NicovideoSearchIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 700     IE_DESC = 'Nico video search'
 701     IE_NAME = 'nicovideo:search'
 702     _SEARCH_KEY = 'nicosearch'
 703
 704
 705 class NicovideoSearchURLIE(NicovideoSearchBaseIE):
 706     IE_NAME = f'{NicovideoSearchIE.IE_NAME}_url'
 707     IE_DESC = 'Nico video search URLs'
 708     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/search/(?P<id>[^?#&]+)?'
 709     _TESTS = [{
 710         'url': 'http://www.nicovideo.jp/search/sm9',
 711         'info_dict': {
 712             'id': 'sm9',
 713             'title': 'sm9'
 714         },
 715         'playlist_mincount': 40,
 716     }, {
 717         'url': 'https://www.nicovideo.jp/search/sm9?sort=h&order=d&end=2020-12-31&start=2020-01-01',
 718         'info_dict': {
 719             'id': 'sm9',
 720             'title': 'sm9'
 721         },
 722         'playlist_count': 31,
 723     }]
 724
 725     def _real_extract(self, url):
 726         query = self._match_id(url)
 727         return self.playlist_result(self._entries(url, query), query, query)
 728
 729
 730 class NicovideoSearchDateIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 731     IE_DESC = 'Nico video search, newest first'
 732     IE_NAME = f'{NicovideoSearchIE.IE_NAME}:date'
 733     _SEARCH_KEY = 'nicosearchdate'
 734     _TESTS = [{
 735         'url': 'nicosearchdateall:a',
 736         'info_dict': {
 737             'id': 'a',
 738             'title': 'a'
 739         },
 740         'playlist_mincount': 1610,
 741     }]
 742
 743     _START_DATE = datetime.date(2007, 1, 1)
 744     _RESULTS_PER_PAGE = 32
 745     _MAX_PAGES = 50
 746
 747     def _entries(self, url, item_id, start_date=None, end_date=None):
 748         start_date, end_date = start_date or self._START_DATE, end_date or datetime.datetime.now().date()
 749
 750         # If the last page has a full page of videos, we need to break down the query interval further
 751         last_page_len = len(list(self._get_entries_for_date(
 752             url, item_id, start_date, end_date, self._MAX_PAGES,
 753             note=f'Checking number of videos from {start_date} to {end_date}')))
 754         if (last_page_len == self._RESULTS_PER_PAGE and start_date != end_date):
 755             midpoint = start_date + ((end_date - start_date) // 2)
 756             yield from self._entries(url, item_id, midpoint, end_date)
 757             yield from self._entries(url, item_id, start_date, midpoint)
 758         else:
 759             self.to_screen(f'{item_id}: Downloading results from {start_date} to {end_date}')
 760             yield from self._get_entries_for_date(
 761                 url, item_id, start_date, end_date, note='    Downloading page %(page)s')
 762
 763     def _get_entries_for_date(self, url, item_id, start_date, end_date=None, page_num=None, note=None):
 764         query = {
 765             'start': str(start_date),
 766             'end': str(end_date or start_date),
 767             'sort': 'f',
 768             'order': 'd',
 769         }
 770         if page_num:
 771             query['page'] = str(page_num)
 772
 773         yield from super()._entries(url, item_id, query=query, note=note)
 774
 775
 776 class NicovideoTagURLIE(NicovideoSearchBaseIE):
 777     IE_NAME = 'niconico:tag'
 778     IE_DESC = 'NicoNico video tag URLs'
 779     _SEARCH_TYPE = 'tag'
 780     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/tag/(?P<id>[^?#&]+)?'
 781     _TESTS = [{
 782         'url': 'https://www.nicovideo.jp/tag/ドキュメンタリー淫夢',
 783         'info_dict': {
 784             'id': 'ドキュメンタリー淫夢',
 785             'title': 'ドキュメンタリー淫夢'
 786         },
 787         'playlist_mincount': 400,
 788     }]
 789
 790     def _real_extract(self, url):
 791         query = self._match_id(url)
 792         return self.playlist_result(self._entries(url, query), query, query)
 793
 794
 795 class NiconicoUserIE(InfoExtractor):
 796     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/user/(?P<id>\d+)/?(?:$|[#?])'
 797     _TEST = {
 798         'url': 'https://www.nicovideo.jp/user/419948',
 799         'info_dict': {
 800             'id': '419948',
 801         },
 802         'playlist_mincount': 101,
 803     }
 804     _API_URL = "https://nvapi.nicovideo.jp/v1/users/%s/videos?sortKey=registeredAt&sortOrder=desc&pageSize=%s&page=%s"
 805     _PAGE_SIZE = 100
 806
 807     _API_HEADERS = {
 808         'X-Frontend-ID': '6',
 809         'X-Frontend-Version': '0'
 810     }
 811
 812     def _entries(self, list_id):
 813         total_count = 1
 814         count = page_num = 0
 815         while count < total_count:
 816             json_parsed = self._download_json(
 817                 self._API_URL % (list_id, self._PAGE_SIZE, page_num + 1), list_id,
 818                 headers=self._API_HEADERS,
 819                 note='Downloading JSON metadata%s' % (' page %d' % page_num if page_num else ''))
 820             if not page_num:
 821                 total_count = int_or_none(json_parsed['data'].get('totalCount'))
 822             for entry in json_parsed["data"]["items"]:
 823                 count += 1
 824                 yield self.url_result('https://www.nicovideo.jp/watch/%s' % entry['id'])
 825             page_num += 1
 826
 827     def _real_extract(self, url):
 828         list_id = self._match_id(url)
 829         return self.playlist_result(self._entries(list_id), list_id, ie=NiconicoIE.ie_key())
 830
 831
 832 class NiconicoLiveIE(InfoExtractor):
 833     IE_NAME = 'niconico:live'
 834     IE_DESC = 'ニコニコ生放送'
 835     _VALID_URL = r'https?://(?:sp\.)?live2?\.nicovideo\.jp/(?:watch|gate)/(?P<id>lv\d+)'
 836     _TESTS = [{
 837         'note': 'this test case includes invisible characters for title, pasting them as-is',
 838         'url': 'https://live.nicovideo.jp/watch/lv339533123',
 839         'info_dict': {
 840             'id': 'lv339533123',
 841             'title': '激辛ペヤング食べます‪( ;ᯅ; )‬（歌枠オーディション参加中）',
 842             'view_count': 1526,
 843             'comment_count': 1772,
 844             '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',
 845             'uploader': 'もか',
 846             'channel': 'ゲストさんのコミュニティ',
 847             'channel_id': 'co5776900',
 848             'channel_url': 'https://com.nicovideo.jp/community/co5776900',
 849             'timestamp': 1670677328,
 850             'is_live': True,
 851         },
 852         'skip': 'livestream',
 853     }, {
 854         'url': 'https://live2.nicovideo.jp/watch/lv339533123',
 855         'only_matching': True,
 856     }, {
 857         'url': 'https://sp.live.nicovideo.jp/watch/lv339533123',
 858         'only_matching': True,
 859     }, {
 860         'url': 'https://sp.live2.nicovideo.jp/watch/lv339533123',
 861         'only_matching': True,
 862     }]
 863
 864     _KNOWN_LATENCY = ('high', 'low')
 865
 866     def _real_extract(self, url):
 867         video_id = self._match_id(url)
 868         webpage, urlh = self._download_webpage_handle(f'https://live.nicovideo.jp/watch/{video_id}', video_id)
 869
 870         embedded_data = self._parse_json(unescapeHTML(self._search_regex(
 871             r'<script\s+id="embedded-data"\s*data-props="(.+?)"', webpage, 'embedded data')), video_id)
 872
 873         ws_url = traverse_obj(embedded_data, ('site', 'relive', 'webSocketUrl'))
 874         if not ws_url:
 875             raise ExtractorError('The live hasn\'t started yet or already ended.', expected=True)
 876         ws_url = update_url_query(ws_url, {
 877             'frontend_id': traverse_obj(embedded_data, ('site', 'frontendId')) or '9',
 878         })
 879
 880         hostname = remove_start(urlparse(urlh.url).hostname, 'sp.')
 881         latency = try_get(self._configuration_arg('latency'), lambda x: x[0])
 882         if latency not in self._KNOWN_LATENCY:
 883             latency = 'high'
 884
 885         ws = self._request_webpage(
 886             Request(ws_url, headers={'Origin': f'https://{hostname}'}),
 887             video_id=video_id, note='Connecting to WebSocket server')
 888
 889         self.write_debug('[debug] Sending HLS server request')
 890         ws.send(json.dumps({
 891             'type': 'startWatching',
 892             'data': {
 893                 'stream': {
 894                     'quality': 'abr',
 895                     'protocol': 'hls+fmp4',
 896                     'latency': latency,
 897                     'chasePlay': False
 898                 },
 899                 'room': {
 900                     'protocol': 'webSocket',
 901                     'commentable': True
 902                 },
 903                 'reconnect': False,
 904             }
 905         }))
 906
 907         while True:
 908             recv = ws.recv()
 909             if not recv:
 910                 continue
 911             data = json.loads(recv)
 912             if not isinstance(data, dict):
 913                 continue
 914             if data.get('type') == 'stream':
 915                 m3u8_url = data['data']['uri']
 916                 qualities = data['data']['availableQualities']
 917                 break
 918             elif data.get('type') == 'disconnect':
 919                 self.write_debug(recv)
 920                 raise ExtractorError('Disconnected at middle of extraction')
 921             elif data.get('type') == 'error':
 922                 self.write_debug(recv)
 923                 message = traverse_obj(data, ('body', 'code')) or recv
 924                 raise ExtractorError(message)
 925             elif self.get_param('verbose', False):
 926                 if len(recv) > 100:
 927                     recv = recv[:100] + '...'
 928                 self.write_debug('Server said: %s' % recv)
 929
 930         title = traverse_obj(embedded_data, ('program', 'title')) or self._html_search_meta(
 931             ('og:title', 'twitter:title'), webpage, 'live title', fatal=False)
 932
 933         raw_thumbs = traverse_obj(embedded_data, ('program', 'thumbnail')) or {}
 934         thumbnails = []
 935         for name, value in raw_thumbs.items():
 936             if not isinstance(value, dict):
 937                 thumbnails.append({
 938                     'id': name,
 939                     'url': value,
 940                     **parse_resolution(value, lenient=True),
 941                 })
 942                 continue
 943
 944             for k, img_url in value.items():
 945                 res = parse_resolution(k, lenient=True) or parse_resolution(img_url, lenient=True)
 946                 width, height = res.get('width'), res.get('height')
 947
 948                 thumbnails.append({
 949                     'id': f'{name}_{width}x{height}',
 950                     'url': img_url,
 951                     **res,
 952                 })
 953
 954         formats = self._extract_m3u8_formats(m3u8_url, video_id, ext='mp4', live=True)
 955         for fmt, q in zip(formats, reversed(qualities[1:])):
 956             fmt.update({
 957                 'format_id': q,
 958                 'protocol': 'niconico_live',
 959                 'ws': ws,
 960                 'video_id': video_id,
 961                 'live_latency': latency,
 962                 'origin': hostname,
 963             })
 964
 965         return {
 966             'id': video_id,
 967             'title': title,
 968             **traverse_obj(embedded_data, {
 969                 'view_count': ('program', 'statistics', 'watchCount'),
 970                 'comment_count': ('program', 'statistics', 'commentCount'),
 971                 'uploader': ('program', 'supplier', 'name'),
 972                 'channel': ('socialGroup', 'name'),
 973                 'channel_id': ('socialGroup', 'id'),
 974                 'channel_url': ('socialGroup', 'socialGroupPageUrl'),
 975             }),
 976             'description': clean_html(traverse_obj(embedded_data, ('program', 'description'))),
 977             'timestamp': int_or_none(traverse_obj(embedded_data, ('program', 'openTime'))),
 978             'is_live': True,
 979             'thumbnails': thumbnails,
 980             'formats': formats,
 981         }