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 .common import InfoExtractor, SearchInfoExtractor
   9 from ..compat import (
  10     compat_HTTPError,
  11 )
  12 from ..utils import (
  13     ExtractorError,
  14     OnDemandPagedList,
  15     bug_reports_message,
  16     clean_html,
  17     float_or_none,
  18     int_or_none,
  19     join_nonempty,
  20     parse_duration,
  21     parse_filesize,
  22     parse_iso8601,
  23     parse_resolution,
  24     qualities,
  25     remove_start,
  26     str_or_none,
  27     traverse_obj,
  28     try_get,
  29     unescapeHTML,
  30     update_url_query,
  31     url_or_none,
  32     urlencode_postdata,
  33     urljoin,
  34 )
  35
  36
  37 class NiconicoIE(InfoExtractor):
  38     IE_NAME = 'niconico'
  39     IE_DESC = 'ニコニコ動画'
  40
  41     _TESTS = [{
  42         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  43         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  44         'info_dict': {
  45             'id': 'sm22312215',
  46             'ext': 'mp4',
  47             'title': 'Big Buck Bunny',
  48             'thumbnail': r're:https?://.*',
  49             'uploader': 'takuya0301',
  50             'uploader_id': '2698420',
  51             'upload_date': '20131123',
  52             'timestamp': int,  # timestamp is unstable
  53             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  54             'duration': 33,
  55             'view_count': int,
  56             'comment_count': int,
  57         },
  58         'skip': 'Requires an account',
  59     }, {
  60         # File downloaded with and without credentials are different, so omit
  61         # the md5 field
  62         'url': 'http://www.nicovideo.jp/watch/nm14296458',
  63         'info_dict': {
  64             'id': 'nm14296458',
  65             'ext': 'swf',
  66             'title': '【鏡音リン】Dance on media【オリジナル】take2!',
  67             'description': 'md5:689f066d74610b3b22e0f1739add0f58',
  68             'thumbnail': r're:https?://.*',
  69             'uploader': 'りょうた',
  70             'uploader_id': '18822557',
  71             'upload_date': '20110429',
  72             'timestamp': 1304065916,
  73             'duration': 209,
  74         },
  75         'skip': 'Requires an account',
  76     }, {
  77         # 'video exists but is marked as "deleted"
  78         # md5 is unstable
  79         'url': 'http://www.nicovideo.jp/watch/sm10000',
  80         'info_dict': {
  81             'id': 'sm10000',
  82             'ext': 'unknown_video',
  83             'description': 'deleted',
  84             'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞',
  85             'thumbnail': r're:https?://.*',
  86             'upload_date': '20071224',
  87             'timestamp': int,  # timestamp field has different value if logged in
  88             'duration': 304,
  89             'view_count': int,
  90         },
  91         'skip': 'Requires an account',
  92     }, {
  93         'url': 'http://www.nicovideo.jp/watch/so22543406',
  94         'info_dict': {
  95             'id': '1388129933',
  96             'ext': 'mp4',
  97             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
  98             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
  99             'thumbnail': r're:https?://.*',
 100             'timestamp': 1388851200,
 101             'upload_date': '20140104',
 102             'uploader': 'アニメロチャンネル',
 103             'uploader_id': '312',
 104         },
 105         'skip': 'The viewing period of the video you were searching for has expired.',
 106     }, {
 107         # video not available via `getflv`; "old" HTML5 video
 108         'url': 'http://www.nicovideo.jp/watch/sm1151009',
 109         'md5': '8fa81c364eb619d4085354eab075598a',
 110         'info_dict': {
 111             'id': 'sm1151009',
 112             'ext': 'mp4',
 113             '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）',
 114             'description': 'md5:6ee077e0581ff5019773e2e714cdd0b7',
 115             'thumbnail': r're:https?://.*',
 116             'duration': 184,
 117             'timestamp': 1190868283,
 118             'upload_date': '20070927',
 119             'uploader': 'denden2',
 120             'uploader_id': '1392194',
 121             'view_count': int,
 122             'comment_count': int,
 123         },
 124         'skip': 'Requires an account',
 125     }, {
 126         # "New" HTML5 video
 127         # md5 is unstable
 128         'url': 'http://www.nicovideo.jp/watch/sm31464864',
 129         'info_dict': {
 130             'id': 'sm31464864',
 131             'ext': 'mp4',
 132             'title': '新作TVアニメ「戦姫絶唱シンフォギアAXZ」PV 最高画質',
 133             'description': 'md5:e52974af9a96e739196b2c1ca72b5feb',
 134             'timestamp': 1498514060,
 135             'upload_date': '20170626',
 136             'uploader': 'ゲスト',
 137             'uploader_id': '40826363',
 138             'thumbnail': r're:https?://.*',
 139             'duration': 198,
 140             'view_count': int,
 141             'comment_count': int,
 142         },
 143         'skip': 'Requires an account',
 144     }, {
 145         # Video without owner
 146         'url': 'http://www.nicovideo.jp/watch/sm18238488',
 147         'md5': 'd265680a1f92bdcbbd2a507fc9e78a9e',
 148         'info_dict': {
 149             'id': 'sm18238488',
 150             'ext': 'mp4',
 151             'title': '【実写版】ミュータントタートルズ',
 152             'description': 'md5:15df8988e47a86f9e978af2064bf6d8e',
 153             'timestamp': 1341160408,
 154             'upload_date': '20120701',
 155             'uploader': None,
 156             'uploader_id': None,
 157             'thumbnail': r're:https?://.*',
 158             'duration': 5271,
 159             'view_count': int,
 160             'comment_count': int,
 161         },
 162         'skip': 'Requires an account',
 163     }, {
 164         'url': 'http://sp.nicovideo.jp/watch/sm28964488?ss_pos=1&cp_in=wt_tg',
 165         'only_matching': True,
 166     }, {
 167         'note': 'a video that is only served as an ENCRYPTED HLS.',
 168         'url': 'https://www.nicovideo.jp/watch/so38016254',
 169         'only_matching': True,
 170     }]
 171
 172     _VALID_URL = r'https?://(?:(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch|nico\.ms)/(?P<id>(?:[a-z]{2})?[0-9]+)'
 173     _NETRC_MACHINE = 'niconico'
 174     _COMMENT_API_ENDPOINTS = (
 175         'https://nvcomment.nicovideo.jp/legacy/api.json',
 176         'https://nmsg.nicovideo.jp/api.json',)
 177     _API_HEADERS = {
 178         'X-Frontend-ID': '6',
 179         'X-Frontend-Version': '0',
 180         'X-Niconico-Language': 'en-us',
 181         'Referer': 'https://www.nicovideo.jp/',
 182         'Origin': 'https://www.nicovideo.jp',
 183     }
 184
 185     def _perform_login(self, username, password):
 186         login_ok = True
 187         login_form_strs = {
 188             'mail_tel': username,
 189             'password': password,
 190         }
 191         self._request_webpage(
 192             'https://account.nicovideo.jp/login', None,
 193             note='Acquiring Login session')
 194         page = self._download_webpage(
 195             'https://account.nicovideo.jp/login/redirector?show_button_twitter=1&site=niconico&show_button_facebook=1', None,
 196             note='Logging in', errnote='Unable to log in',
 197             data=urlencode_postdata(login_form_strs),
 198             headers={
 199                 'Referer': 'https://account.nicovideo.jp/login',
 200                 'Content-Type': 'application/x-www-form-urlencoded',
 201             })
 202         if 'oneTimePw' in page:
 203             post_url = self._search_regex(
 204                 r'<form[^>]+action=(["\'])(?P<url>.+?)\1', page, 'post url', group='url')
 205             page = self._download_webpage(
 206                 urljoin('https://account.nicovideo.jp', post_url), None,
 207                 note='Performing MFA', errnote='Unable to complete MFA',
 208                 data=urlencode_postdata({
 209                     'otp': self._get_tfa_info('6 digits code')
 210                 }), headers={
 211                     'Content-Type': 'application/x-www-form-urlencoded',
 212                 })
 213             if 'oneTimePw' in page or 'formError' in page:
 214                 err_msg = self._html_search_regex(
 215                     r'formError["\']+>(.*?)</div>', page, 'form_error',
 216                     default='There\'s an error but the message can\'t be parsed.',
 217                     flags=re.DOTALL)
 218                 self.report_warning(f'Unable to log in: MFA challenge failed, "{err_msg}"')
 219                 return False
 220         login_ok = 'class="notice error"' not in page
 221         if not login_ok:
 222             self.report_warning('Unable to log in: bad username or password')
 223         return login_ok
 224
 225     def _get_heartbeat_info(self, info_dict):
 226         video_id, video_src_id, audio_src_id = info_dict['url'].split(':')[1].split('/')
 227         dmc_protocol = info_dict['expected_protocol']
 228
 229         api_data = (
 230             info_dict.get('_api_data')
 231             or self._parse_json(
 232                 self._html_search_regex(
 233                     'data-api-data="([^"]+)"',
 234                     self._download_webpage('http://www.nicovideo.jp/watch/' + video_id, video_id),
 235                     'API data', default='{}'),
 236                 video_id))
 237
 238         session_api_data = try_get(api_data, lambda x: x['media']['delivery']['movie']['session'])
 239         session_api_endpoint = try_get(session_api_data, lambda x: x['urls'][0])
 240
 241         def ping():
 242             tracking_id = traverse_obj(api_data, ('media', 'delivery', 'trackingId'))
 243             if tracking_id:
 244                 tracking_url = update_url_query('https://nvapi.nicovideo.jp/v1/2ab0cbaa/watch', {'t': tracking_id})
 245                 watch_request_response = self._download_json(
 246                     tracking_url, video_id,
 247                     note='Acquiring permission for downloading video', fatal=False,
 248                     headers=self._API_HEADERS)
 249                 if traverse_obj(watch_request_response, ('meta', 'status')) != 200:
 250                     self.report_warning('Failed to acquire permission for playing video. Video download may fail.')
 251
 252         yesno = lambda x: 'yes' if x else 'no'
 253
 254         if dmc_protocol == 'http':
 255             protocol = 'http'
 256             protocol_parameters = {
 257                 'http_output_download_parameters': {
 258                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 259                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 260                 }
 261             }
 262         elif dmc_protocol == 'hls':
 263             protocol = 'm3u8'
 264             segment_duration = try_get(self._configuration_arg('segment_duration'), lambda x: int(x[0])) or 6000
 265             parsed_token = self._parse_json(session_api_data['token'], video_id)
 266             encryption = traverse_obj(api_data, ('media', 'delivery', 'encryption'))
 267             protocol_parameters = {
 268                 'hls_parameters': {
 269                     'segment_duration': segment_duration,
 270                     'transfer_preset': '',
 271                     'use_ssl': yesno(session_api_data['urls'][0]['isSsl']),
 272                     'use_well_known_port': yesno(session_api_data['urls'][0]['isWellKnownPort']),
 273                 }
 274             }
 275             if 'hls_encryption' in parsed_token and encryption:
 276                 protocol_parameters['hls_parameters']['encryption'] = {
 277                     parsed_token['hls_encryption']: {
 278                         'encrypted_key': encryption['encryptedKey'],
 279                         'key_uri': encryption['keyUri'],
 280                     }
 281                 }
 282             else:
 283                 protocol = 'm3u8_native'
 284         else:
 285             raise ExtractorError(f'Unsupported DMC protocol: {dmc_protocol}')
 286
 287         session_response = self._download_json(
 288             session_api_endpoint['url'], video_id,
 289             query={'_format': 'json'},
 290             headers={'Content-Type': 'application/json'},
 291             note='Downloading JSON metadata for %s' % info_dict['format_id'],
 292             data=json.dumps({
 293                 'session': {
 294                     'client_info': {
 295                         'player_id': session_api_data.get('playerId'),
 296                     },
 297                     'content_auth': {
 298                         'auth_type': try_get(session_api_data, lambda x: x['authTypes'][session_api_data['protocols'][0]]),
 299                         'content_key_timeout': session_api_data.get('contentKeyTimeout'),
 300                         'service_id': 'nicovideo',
 301                         'service_user_id': session_api_data.get('serviceUserId')
 302                     },
 303                     'content_id': session_api_data.get('contentId'),
 304                     'content_src_id_sets': [{
 305                         'content_src_ids': [{
 306                             'src_id_to_mux': {
 307                                 'audio_src_ids': [audio_src_id],
 308                                 'video_src_ids': [video_src_id],
 309                             }
 310                         }]
 311                     }],
 312                     'content_type': 'movie',
 313                     'content_uri': '',
 314                     'keep_method': {
 315                         'heartbeat': {
 316                             'lifetime': session_api_data.get('heartbeatLifetime')
 317                         }
 318                     },
 319                     'priority': session_api_data['priority'],
 320                     'protocol': {
 321                         'name': 'http',
 322                         'parameters': {
 323                             'http_parameters': {
 324                                 'parameters': protocol_parameters
 325                             }
 326                         }
 327                     },
 328                     'recipe_id': session_api_data.get('recipeId'),
 329                     'session_operation_auth': {
 330                         'session_operation_auth_by_signature': {
 331                             'signature': session_api_data.get('signature'),
 332                             'token': session_api_data.get('token'),
 333                         }
 334                     },
 335                     'timing_constraint': 'unlimited'
 336                 }
 337             }).encode())
 338
 339         info_dict['url'] = session_response['data']['session']['content_uri']
 340         info_dict['protocol'] = protocol
 341
 342         # get heartbeat info
 343         heartbeat_info_dict = {
 344             'url': session_api_endpoint['url'] + '/' + session_response['data']['session']['id'] + '?_format=json&_method=PUT',
 345             'data': json.dumps(session_response['data']),
 346             # interval, convert milliseconds to seconds, then halve to make a buffer.
 347             'interval': float_or_none(session_api_data.get('heartbeatLifetime'), scale=3000),
 348             'ping': ping
 349         }
 350
 351         return info_dict, heartbeat_info_dict
 352
 353     def _extract_format_for_quality(self, video_id, audio_quality, video_quality, dmc_protocol):
 354
 355         if not audio_quality.get('isAvailable') or not video_quality.get('isAvailable'):
 356             return None
 357
 358         def extract_video_quality(video_quality):
 359             return parse_filesize('%sB' % self._search_regex(
 360                 r'\| ([0-9]*\.?[0-9]*[MK])', video_quality, 'vbr', default=''))
 361
 362         format_id = '-'.join(
 363             [remove_start(s['id'], 'archive_') for s in (video_quality, audio_quality)] + [dmc_protocol])
 364
 365         vid_qual_label = traverse_obj(video_quality, ('metadata', 'label'))
 366         vid_quality = traverse_obj(video_quality, ('metadata', 'bitrate'))
 367
 368         return {
 369             'url': 'niconico_dmc:%s/%s/%s' % (video_id, video_quality['id'], audio_quality['id']),
 370             'format_id': format_id,
 371             'format_note': join_nonempty('DMC', vid_qual_label, dmc_protocol.upper(), delim=' '),
 372             'ext': 'mp4',  # Session API are used in HTML5, which always serves mp4
 373             'acodec': 'aac',
 374             'vcodec': 'h264',
 375             'abr': float_or_none(traverse_obj(audio_quality, ('metadata', 'bitrate')), 1000),
 376             'vbr': float_or_none(vid_quality if vid_quality > 0 else extract_video_quality(vid_qual_label), 1000),
 377             'height': traverse_obj(video_quality, ('metadata', 'resolution', 'height')),
 378             'width': traverse_obj(video_quality, ('metadata', 'resolution', 'width')),
 379             'quality': -2 if 'low' in video_quality['id'] else None,
 380             'protocol': 'niconico_dmc',
 381             'expected_protocol': dmc_protocol,  # XXX: This is not a documented field
 382             'http_headers': {
 383                 'Origin': 'https://www.nicovideo.jp',
 384                 'Referer': 'https://www.nicovideo.jp/watch/' + video_id,
 385             }
 386         }
 387
 388     def _real_extract(self, url):
 389         video_id = self._match_id(url)
 390
 391         try:
 392             webpage, handle = self._download_webpage_handle(
 393                 'http://www.nicovideo.jp/watch/' + video_id, video_id)
 394             if video_id.startswith('so'):
 395                 video_id = self._match_id(handle.geturl())
 396
 397             api_data = self._parse_json(self._html_search_regex(
 398                 'data-api-data="([^"]+)"', webpage,
 399                 'API data', default='{}'), video_id)
 400         except ExtractorError as e:
 401             try:
 402                 api_data = self._download_json(
 403                     'https://www.nicovideo.jp/api/watch/v3/%s?_frontendId=6&_frontendVersion=0&actionTrackId=AAAAAAAAAA_%d' % (video_id, round(time.time() * 1000)), video_id,
 404                     note='Downloading API JSON', errnote='Unable to fetch data')['data']
 405             except ExtractorError:
 406                 if not isinstance(e.cause, compat_HTTPError):
 407                     raise
 408                 webpage = e.cause.read().decode('utf-8', 'replace')
 409                 error_msg = self._html_search_regex(
 410                     r'(?s)<section\s+class="(?:(?:ErrorMessage|WatchExceptionPage-message)\s*)+">(.+?)</section>',
 411                     webpage, 'error reason', default=None)
 412                 if not error_msg:
 413                     raise
 414                 raise ExtractorError(re.sub(r'\s+', ' ', error_msg), expected=True)
 415
 416         formats = []
 417
 418         def get_video_info(*items, get_first=True, **kwargs):
 419             return traverse_obj(api_data, ('video', *items), get_all=not get_first, **kwargs)
 420
 421         quality_info = api_data['media']['delivery']['movie']
 422         session_api_data = quality_info['session']
 423         for (audio_quality, video_quality, protocol) in itertools.product(quality_info['audios'], quality_info['videos'], session_api_data['protocols']):
 424             fmt = self._extract_format_for_quality(video_id, audio_quality, video_quality, protocol)
 425             if fmt:
 426                 formats.append(fmt)
 427
 428         self._sort_formats(formats)
 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         raw_danmaku = self._extract_all_comments(video_id, thread_ids, user_id_str, comment_user_key)
 483         if not raw_danmaku:
 484             self.report_warning(f'Failed to get comments. {bug_reports_message()}')
 485             return
 486         return {
 487             'comments': [{
 488                 'ext': 'json',
 489                 'data': json.dumps(raw_danmaku),
 490             }],
 491         }
 492
 493     def _extract_all_comments(self, video_id, threads, user_id, user_key):
 494         auth_data = {
 495             'user_id': user_id,
 496             'userkey': user_key,
 497         } if user_id and user_key else {'user_id': ''}
 498
 499         # Request Start
 500         post_data = [{'ping': {'content': 'rs:0'}}]
 501         for i, thread in enumerate(threads):
 502             thread_id = thread['id']
 503             thread_fork = thread['fork']
 504             # Post Start (2N)
 505             post_data.append({'ping': {'content': f'ps:{i * 2}'}})
 506             post_data.append({'thread': {
 507                 'fork': thread_fork,
 508                 'language': 0,
 509                 'nicoru': 3,
 510                 'scores': 1,
 511                 'thread': thread_id,
 512                 'version': '20090904',
 513                 'with_global': 1,
 514                 **auth_data,
 515             }})
 516             # Post Final (2N)
 517             post_data.append({'ping': {'content': f'pf:{i * 2}'}})
 518
 519             # Post Start (2N+1)
 520             post_data.append({'ping': {'content': f'ps:{i * 2 + 1}'}})
 521             post_data.append({'thread_leaves': {
 522                 # format is '<bottom of minute range>-<top of minute range>:<comments per minute>,<total last comments'
 523                 # unfortunately NND limits (deletes?) comment returns this way, so you're only able to grab the last 1000 per language
 524                 'content': '0-999999:999999,999999,nicoru:999999',
 525                 'fork': thread_fork,
 526                 'language': 0,
 527                 'nicoru': 3,
 528                 'scores': 1,
 529                 'thread': thread_id,
 530                 **auth_data,
 531             }})
 532             # Post Final (2N+1)
 533             post_data.append({'ping': {'content': f'pf:{i * 2 + 1}'}})
 534         # Request Final
 535         post_data.append({'ping': {'content': 'rf:0'}})
 536
 537         for api_url in self._COMMENT_API_ENDPOINTS:
 538             comments = self._download_json(
 539                 api_url, video_id, data=json.dumps(post_data).encode(), fatal=False,
 540                 headers={
 541                     'Referer': 'https://www.nicovideo.jp/watch/%s' % video_id,
 542                     'Origin': 'https://www.nicovideo.jp',
 543                     'Content-Type': 'text/plain;charset=UTF-8',
 544                 },
 545                 note='Downloading comments', errnote=f'Failed to access endpoint {api_url}')
 546             if comments:
 547                 return comments
 548
 549
 550 class NiconicoPlaylistBaseIE(InfoExtractor):
 551     _PAGE_SIZE = 100
 552
 553     _API_HEADERS = {
 554         'X-Frontend-ID': '6',
 555         'X-Frontend-Version': '0',
 556         'X-Niconico-Language': 'en-us'
 557     }
 558
 559     def _call_api(self, list_id, resource, query):
 560         "Implement this in child class"
 561         pass
 562
 563     @staticmethod
 564     def _parse_owner(item):
 565         return {
 566             'uploader': traverse_obj(item, ('owner', 'name')),
 567             'uploader_id': traverse_obj(item, ('owner', 'id')),
 568         }
 569
 570     def _fetch_page(self, list_id, page):
 571         page += 1
 572         resp = self._call_api(list_id, 'page %d' % page, {
 573             'page': page,
 574             'pageSize': self._PAGE_SIZE,
 575         })
 576         # this is needed to support both mylist and user
 577         for video in traverse_obj(resp, ('items', ..., ('video', None))) or []:
 578             video_id = video.get('id')
 579             if not video_id:
 580                 # skip {"video": {"id": "blablabla", ...}}
 581                 continue
 582             count = video.get('count') or {}
 583             get_count = lambda x: int_or_none(count.get(x))
 584             yield {
 585                 '_type': 'url',
 586                 'id': video_id,
 587                 'title': video.get('title'),
 588                 'url': f'https://www.nicovideo.jp/watch/{video_id}',
 589                 'description': video.get('shortDescription'),
 590                 'duration': int_or_none(video.get('duration')),
 591                 'view_count': get_count('view'),
 592                 'comment_count': get_count('comment'),
 593                 'thumbnail': traverse_obj(video, ('thumbnail', ('nHdUrl', 'largeUrl', 'listingUrl', 'url'))),
 594                 'ie_key': NiconicoIE.ie_key(),
 595                 **self._parse_owner(video),
 596             }
 597
 598     def _entries(self, list_id):
 599         return OnDemandPagedList(functools.partial(self._fetch_page, list_id), self._PAGE_SIZE)
 600
 601
 602 class NiconicoPlaylistIE(NiconicoPlaylistBaseIE):
 603     IE_NAME = 'niconico:playlist'
 604     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp|nico\.ms)/(?:user/\d+/)?(?:my/)?mylist/(?:#/)?(?P<id>\d+)'
 605
 606     _TESTS = [{
 607         'url': 'http://www.nicovideo.jp/mylist/27411728',
 608         'info_dict': {
 609             'id': '27411728',
 610             'title': 'AKB48のオールナイトニッポン',
 611             'description': 'md5:d89694c5ded4b6c693dea2db6e41aa08',
 612             'uploader': 'のっく',
 613             'uploader_id': '805442',
 614         },
 615         'playlist_mincount': 291,
 616     }, {
 617         'url': 'https://www.nicovideo.jp/user/805442/mylist/27411728',
 618         'only_matching': True,
 619     }, {
 620         'url': 'https://www.nicovideo.jp/my/mylist/#/68048635',
 621         'only_matching': True,
 622     }]
 623
 624     def _call_api(self, list_id, resource, query):
 625         return self._download_json(
 626             f'https://nvapi.nicovideo.jp/v2/mylists/{list_id}', list_id,
 627             f'Downloading {resource}', query=query,
 628             headers=self._API_HEADERS)['data']['mylist']
 629
 630     def _real_extract(self, url):
 631         list_id = self._match_id(url)
 632         mylist = self._call_api(list_id, 'list', {
 633             'pageSize': 1,
 634         })
 635         return self.playlist_result(
 636             self._entries(list_id), list_id,
 637             mylist.get('name'), mylist.get('description'), **self._parse_owner(mylist))
 638
 639
 640 class NiconicoSeriesIE(InfoExtractor):
 641     IE_NAME = 'niconico:series'
 642     _VALID_URL = r'https?://(?:(?:www\.|sp\.)?nicovideo\.jp|nico\.ms)/series/(?P<id>\d+)'
 643
 644     _TESTS = [{
 645         'url': 'https://www.nicovideo.jp/series/110226',
 646         'info_dict': {
 647             'id': '110226',
 648             'title': 'ご立派ァ！のシリーズ',
 649         },
 650         'playlist_mincount': 10,
 651     }, {
 652         'url': 'https://www.nicovideo.jp/series/12312/',
 653         'info_dict': {
 654             'id': '12312',
 655             'title': 'バトルスピリッツ　お勧めカード紹介(調整中)',
 656         },
 657         'playlist_mincount': 103,
 658     }, {
 659         'url': 'https://nico.ms/series/203559',
 660         'only_matching': True,
 661     }]
 662
 663     def _real_extract(self, url):
 664         list_id = self._match_id(url)
 665         webpage = self._download_webpage(f'https://www.nicovideo.jp/series/{list_id}', list_id)
 666
 667         title = self._search_regex(
 668             (r'<title>「(.+)（全',
 669              r'<div class="TwitterShareButton"\s+data-text="(.+)\s+https:'),
 670             webpage, 'title', fatal=False)
 671         if title:
 672             title = unescapeHTML(title)
 673         playlist = [
 674             self.url_result(f'https://www.nicovideo.jp/watch/{v_id}', video_id=v_id)
 675             for v_id in re.findall(r'data-href=[\'"](?:https://www\.nicovideo\.jp)?/watch/([a-z0-9]+)', webpage)]
 676         return self.playlist_result(playlist, list_id, title)
 677
 678
 679 class NiconicoHistoryIE(NiconicoPlaylistBaseIE):
 680     IE_NAME = 'niconico:history'
 681     IE_DESC = 'NicoNico user history. Requires cookies.'
 682     _VALID_URL = r'https?://(?:www\.|sp\.)?nicovideo\.jp/my/history'
 683
 684     _TESTS = [{
 685         'note': 'PC page, with /video',
 686         'url': 'https://www.nicovideo.jp/my/history/video',
 687         'only_matching': True,
 688     }, {
 689         'note': 'PC page, without /video',
 690         'url': 'https://www.nicovideo.jp/my/history',
 691         'only_matching': True,
 692     }, {
 693         'note': 'mobile page, with /video',
 694         'url': 'https://sp.nicovideo.jp/my/history/video',
 695         'only_matching': True,
 696     }, {
 697         'note': 'mobile page, without /video',
 698         'url': 'https://sp.nicovideo.jp/my/history',
 699         'only_matching': True,
 700     }]
 701
 702     def _call_api(self, list_id, resource, query):
 703         return self._download_json(
 704             'https://nvapi.nicovideo.jp/v1/users/me/watch/history', 'history',
 705             f'Downloading {resource}', query=query,
 706             headers=self._API_HEADERS)['data']
 707
 708     def _real_extract(self, url):
 709         list_id = 'history'
 710         try:
 711             mylist = self._call_api(list_id, 'list', {
 712                 'pageSize': 1,
 713             })
 714         except ExtractorError as e:
 715             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
 716                 self.raise_login_required('You have to be logged in to get your watch history')
 717             raise
 718         return self.playlist_result(self._entries(list_id), list_id, **self._parse_owner(mylist))
 719
 720
 721 class NicovideoSearchBaseIE(InfoExtractor):
 722     _SEARCH_TYPE = 'search'
 723
 724     def _entries(self, url, item_id, query=None, note='Downloading page %(page)s'):
 725         query = query or {}
 726         pages = [query['page']] if 'page' in query else itertools.count(1)
 727         for page_num in pages:
 728             query['page'] = str(page_num)
 729             webpage = self._download_webpage(url, item_id, query=query, note=note % {'page': page_num})
 730             results = re.findall(r'(?<=data-video-id=)["\']?(?P<videoid>.*?)(?=["\'])', webpage)
 731             for item in results:
 732                 yield self.url_result(f'http://www.nicovideo.jp/watch/{item}', 'Niconico', item)
 733             if not results:
 734                 break
 735
 736     def _search_results(self, query):
 737         return self._entries(
 738             self._proto_relative_url(f'//www.nicovideo.jp/{self._SEARCH_TYPE}/{query}'), query)
 739
 740
 741 class NicovideoSearchIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 742     IE_DESC = 'Nico video search'
 743     IE_NAME = 'nicovideo:search'
 744     _SEARCH_KEY = 'nicosearch'
 745
 746
 747 class NicovideoSearchURLIE(NicovideoSearchBaseIE):
 748     IE_NAME = f'{NicovideoSearchIE.IE_NAME}_url'
 749     IE_DESC = 'Nico video search URLs'
 750     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/search/(?P<id>[^?#&]+)?'
 751     _TESTS = [{
 752         'url': 'http://www.nicovideo.jp/search/sm9',
 753         'info_dict': {
 754             'id': 'sm9',
 755             'title': 'sm9'
 756         },
 757         'playlist_mincount': 40,
 758     }, {
 759         'url': 'https://www.nicovideo.jp/search/sm9?sort=h&order=d&end=2020-12-31&start=2020-01-01',
 760         'info_dict': {
 761             'id': 'sm9',
 762             'title': 'sm9'
 763         },
 764         'playlist_count': 31,
 765     }]
 766
 767     def _real_extract(self, url):
 768         query = self._match_id(url)
 769         return self.playlist_result(self._entries(url, query), query, query)
 770
 771
 772 class NicovideoSearchDateIE(NicovideoSearchBaseIE, SearchInfoExtractor):
 773     IE_DESC = 'Nico video search, newest first'
 774     IE_NAME = f'{NicovideoSearchIE.IE_NAME}:date'
 775     _SEARCH_KEY = 'nicosearchdate'
 776     _TESTS = [{
 777         'url': 'nicosearchdateall:a',
 778         'info_dict': {
 779             'id': 'a',
 780             'title': 'a'
 781         },
 782         'playlist_mincount': 1610,
 783     }]
 784
 785     _START_DATE = datetime.date(2007, 1, 1)
 786     _RESULTS_PER_PAGE = 32
 787     _MAX_PAGES = 50
 788
 789     def _entries(self, url, item_id, start_date=None, end_date=None):
 790         start_date, end_date = start_date or self._START_DATE, end_date or datetime.datetime.now().date()
 791
 792         # If the last page has a full page of videos, we need to break down the query interval further
 793         last_page_len = len(list(self._get_entries_for_date(
 794             url, item_id, start_date, end_date, self._MAX_PAGES,
 795             note=f'Checking number of videos from {start_date} to {end_date}')))
 796         if (last_page_len == self._RESULTS_PER_PAGE and start_date != end_date):
 797             midpoint = start_date + ((end_date - start_date) // 2)
 798             yield from self._entries(url, item_id, midpoint, end_date)
 799             yield from self._entries(url, item_id, start_date, midpoint)
 800         else:
 801             self.to_screen(f'{item_id}: Downloading results from {start_date} to {end_date}')
 802             yield from self._get_entries_for_date(
 803                 url, item_id, start_date, end_date, note='    Downloading page %(page)s')
 804
 805     def _get_entries_for_date(self, url, item_id, start_date, end_date=None, page_num=None, note=None):
 806         query = {
 807             'start': str(start_date),
 808             'end': str(end_date or start_date),
 809             'sort': 'f',
 810             'order': 'd',
 811         }
 812         if page_num:
 813             query['page'] = str(page_num)
 814
 815         yield from super()._entries(url, item_id, query=query, note=note)
 816
 817
 818 class NicovideoTagURLIE(NicovideoSearchBaseIE):
 819     IE_NAME = 'niconico:tag'
 820     IE_DESC = 'NicoNico video tag URLs'
 821     _SEARCH_TYPE = 'tag'
 822     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/tag/(?P<id>[^?#&]+)?'
 823     _TESTS = [{
 824         'url': 'https://www.nicovideo.jp/tag/ドキュメンタリー淫夢',
 825         'info_dict': {
 826             'id': 'ドキュメンタリー淫夢',
 827             'title': 'ドキュメンタリー淫夢'
 828         },
 829         'playlist_mincount': 400,
 830     }]
 831
 832     def _real_extract(self, url):
 833         query = self._match_id(url)
 834         return self.playlist_result(self._entries(url, query), query, query)
 835
 836
 837 class NiconicoUserIE(InfoExtractor):
 838     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/user/(?P<id>\d+)/?(?:$|[#?])'
 839     _TEST = {
 840         'url': 'https://www.nicovideo.jp/user/419948',
 841         'info_dict': {
 842             'id': '419948',
 843         },
 844         'playlist_mincount': 101,
 845     }
 846     _API_URL = "https://nvapi.nicovideo.jp/v1/users/%s/videos?sortKey=registeredAt&sortOrder=desc&pageSize=%s&page=%s"
 847     _PAGE_SIZE = 100
 848
 849     _API_HEADERS = {
 850         'X-Frontend-ID': '6',
 851         'X-Frontend-Version': '0'
 852     }
 853
 854     def _entries(self, list_id):
 855         total_count = 1
 856         count = page_num = 0
 857         while count < total_count:
 858             json_parsed = self._download_json(
 859                 self._API_URL % (list_id, self._PAGE_SIZE, page_num + 1), list_id,
 860                 headers=self._API_HEADERS,
 861                 note='Downloading JSON metadata%s' % (' page %d' % page_num if page_num else ''))
 862             if not page_num:
 863                 total_count = int_or_none(json_parsed['data'].get('totalCount'))
 864             for entry in json_parsed["data"]["items"]:
 865                 count += 1
 866                 yield self.url_result('https://www.nicovideo.jp/watch/%s' % entry['id'])
 867             page_num += 1
 868
 869     def _real_extract(self, url):
 870         list_id = self._match_id(url)
 871         return self.playlist_result(self._entries(list_id), list_id, ie=NiconicoIE.ie_key())