youtube_dlc/extractor/niconico.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import datetime
   5 import functools
   6 import json
   7 import math
   8
   9 from .common import InfoExtractor
  10 from ..compat import (
  11     compat_parse_qs,
  12     compat_urllib_parse_urlparse,
  13 )
  14 from ..utils import (
  15     determine_ext,
  16     dict_get,
  17     ExtractorError,
  18     float_or_none,
  19     InAdvancePagedList,
  20     int_or_none,
  21     parse_duration,
  22     parse_iso8601,
  23     remove_start,
  24     try_get,
  25     unified_timestamp,
  26     urlencode_postdata,
  27     xpath_text,
  28 )
  29
  30
  31 class NiconicoIE(InfoExtractor):
  32     IE_NAME = 'niconico'
  33     IE_DESC = 'ニコニコ動画'
  34
  35     _TESTS = [{
  36         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  37         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  38         'info_dict': {
  39             'id': 'sm22312215',
  40             'ext': 'mp4',
  41             'title': 'Big Buck Bunny',
  42             'thumbnail': r're:https?://.*',
  43             'uploader': 'takuya0301',
  44             'uploader_id': '2698420',
  45             'upload_date': '20131123',
  46             'timestamp': int,  # timestamp is unstable
  47             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  48             'duration': 33,
  49             'view_count': int,
  50             'comment_count': int,
  51         },
  52         'skip': 'Requires an account',
  53     }, {
  54         # File downloaded with and without credentials are different, so omit
  55         # the md5 field
  56         'url': 'http://www.nicovideo.jp/watch/nm14296458',
  57         'info_dict': {
  58             'id': 'nm14296458',
  59             'ext': 'swf',
  60             'title': '【鏡音リン】Dance on media【オリジナル】take2!',
  61             'description': 'md5:689f066d74610b3b22e0f1739add0f58',
  62             'thumbnail': r're:https?://.*',
  63             'uploader': 'りょうた',
  64             'uploader_id': '18822557',
  65             'upload_date': '20110429',
  66             'timestamp': 1304065916,
  67             'duration': 209,
  68         },
  69         'skip': 'Requires an account',
  70     }, {
  71         # 'video exists but is marked as "deleted"
  72         # md5 is unstable
  73         'url': 'http://www.nicovideo.jp/watch/sm10000',
  74         'info_dict': {
  75             'id': 'sm10000',
  76             'ext': 'unknown_video',
  77             'description': 'deleted',
  78             'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞',
  79             'thumbnail': r're:https?://.*',
  80             'upload_date': '20071224',
  81             'timestamp': int,  # timestamp field has different value if logged in
  82             'duration': 304,
  83             'view_count': int,
  84         },
  85         'skip': 'Requires an account',
  86     }, {
  87         'url': 'http://www.nicovideo.jp/watch/so22543406',
  88         'info_dict': {
  89             'id': '1388129933',
  90             'ext': 'mp4',
  91             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
  92             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
  93             'thumbnail': r're:https?://.*',
  94             'timestamp': 1388851200,
  95             'upload_date': '20140104',
  96             'uploader': 'アニメロチャンネル',
  97             'uploader_id': '312',
  98         },
  99         'skip': 'The viewing period of the video you were searching for has expired.',
 100     }, {
 101         # video not available via `getflv`; "old" HTML5 video
 102         'url': 'http://www.nicovideo.jp/watch/sm1151009',
 103         'md5': '8fa81c364eb619d4085354eab075598a',
 104         'info_dict': {
 105             'id': 'sm1151009',
 106             'ext': 'mp4',
 107             '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）',
 108             'description': 'md5:6ee077e0581ff5019773e2e714cdd0b7',
 109             'thumbnail': r're:https?://.*',
 110             'duration': 184,
 111             'timestamp': 1190868283,
 112             'upload_date': '20070927',
 113             'uploader': 'denden2',
 114             'uploader_id': '1392194',
 115             'view_count': int,
 116             'comment_count': int,
 117         },
 118         'skip': 'Requires an account',
 119     }, {
 120         # "New" HTML5 video
 121         # md5 is unstable
 122         'url': 'http://www.nicovideo.jp/watch/sm31464864',
 123         'info_dict': {
 124             'id': 'sm31464864',
 125             'ext': 'mp4',
 126             'title': '新作TVアニメ「戦姫絶唱シンフォギアAXZ」PV 最高画質',
 127             'description': 'md5:e52974af9a96e739196b2c1ca72b5feb',
 128             'timestamp': 1498514060,
 129             'upload_date': '20170626',
 130             'uploader': 'ゲスト',
 131             'uploader_id': '40826363',
 132             'thumbnail': r're:https?://.*',
 133             'duration': 198,
 134             'view_count': int,
 135             'comment_count': int,
 136         },
 137         'skip': 'Requires an account',
 138     }, {
 139         # Video without owner
 140         'url': 'http://www.nicovideo.jp/watch/sm18238488',
 141         'md5': 'd265680a1f92bdcbbd2a507fc9e78a9e',
 142         'info_dict': {
 143             'id': 'sm18238488',
 144             'ext': 'mp4',
 145             'title': '【実写版】ミュータントタートルズ',
 146             'description': 'md5:15df8988e47a86f9e978af2064bf6d8e',
 147             'timestamp': 1341160408,
 148             'upload_date': '20120701',
 149             'uploader': None,
 150             'uploader_id': None,
 151             'thumbnail': r're:https?://.*',
 152             'duration': 5271,
 153             'view_count': int,
 154             'comment_count': int,
 155         },
 156         'skip': 'Requires an account',
 157     }, {
 158         'url': 'http://sp.nicovideo.jp/watch/sm28964488?ss_pos=1&cp_in=wt_tg',
 159         'only_matching': True,
 160     }]
 161
 162     _VALID_URL = r'https?://(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch/(?P<id>(?:[a-z]{2})?[0-9]+)'
 163     _NETRC_MACHINE = 'niconico'
 164
 165     def _real_initialize(self):
 166         self._login()
 167
 168     def _login(self):
 169         username, password = self._get_login_info()
 170         # No authentication to be performed
 171         if not username:
 172             return True
 173
 174         # Log in
 175         login_ok = True
 176         login_form_strs = {
 177             'mail_tel': username,
 178             'password': password,
 179         }
 180         urlh = self._request_webpage(
 181             'https://account.nicovideo.jp/api/v1/login', None,
 182             note='Logging in', errnote='Unable to log in',
 183             data=urlencode_postdata(login_form_strs))
 184         if urlh is False:
 185             login_ok = False
 186         else:
 187             parts = compat_urllib_parse_urlparse(urlh.geturl())
 188             if compat_parse_qs(parts.query).get('message', [None])[0] == 'cant_login':
 189                 login_ok = False
 190         if not login_ok:
 191             self._downloader.report_warning('unable to log in: bad username or password')
 192         return login_ok
 193
 194     def _extract_format_for_quality(self, api_data, video_id, audio_quality, video_quality):
 195         def yesno(boolean):
 196             return 'yes' if boolean else 'no'
 197
 198         session_api_data = api_data['video']['dmcInfo']['session_api']
 199         session_api_endpoint = session_api_data['urls'][0]
 200
 201         format_id = '-'.join(map(lambda s: remove_start(s['id'], 'archive_'), [video_quality, audio_quality]))
 202
 203         session_response = self._download_json(
 204             session_api_endpoint['url'], video_id,
 205             query={'_format': 'json'},
 206             headers={'Content-Type': 'application/json'},
 207             note='Downloading JSON metadata for %s' % format_id,
 208             data=json.dumps({
 209                 'session': {
 210                     'client_info': {
 211                         'player_id': session_api_data['player_id'],
 212                     },
 213                     'content_auth': {
 214                         'auth_type': session_api_data['auth_types'][session_api_data['protocols'][0]],
 215                         'content_key_timeout': session_api_data['content_key_timeout'],
 216                         'service_id': 'nicovideo',
 217                         'service_user_id': session_api_data['service_user_id']
 218                     },
 219                     'content_id': session_api_data['content_id'],
 220                     'content_src_id_sets': [{
 221                         'content_src_ids': [{
 222                             'src_id_to_mux': {
 223                                 'audio_src_ids': [audio_quality['id']],
 224                                 'video_src_ids': [video_quality['id']],
 225                             }
 226                         }]
 227                     }],
 228                     'content_type': 'movie',
 229                     'content_uri': '',
 230                     'keep_method': {
 231                         'heartbeat': {
 232                             'lifetime': session_api_data['heartbeat_lifetime']
 233                         }
 234                     },
 235                     'priority': session_api_data['priority'],
 236                     'protocol': {
 237                         'name': 'http',
 238                         'parameters': {
 239                             'http_parameters': {
 240                                 'parameters': {
 241                                     'http_output_download_parameters': {
 242                                         'use_ssl': yesno(session_api_endpoint['is_ssl']),
 243                                         'use_well_known_port': yesno(session_api_endpoint['is_well_known_port']),
 244                                     }
 245                                 }
 246                             }
 247                         }
 248                     },
 249                     'recipe_id': session_api_data['recipe_id'],
 250                     'session_operation_auth': {
 251                         'session_operation_auth_by_signature': {
 252                             'signature': session_api_data['signature'],
 253                             'token': session_api_data['token'],
 254                         }
 255                     },
 256                     'timing_constraint': 'unlimited'
 257                 }
 258             }).encode())
 259
 260         resolution = video_quality.get('resolution', {})
 261
 262         return {
 263             'url': session_response['data']['session']['content_uri'],
 264             'format_id': format_id,
 265             'ext': 'mp4',  # Session API are used in HTML5, which always serves mp4
 266             'abr': float_or_none(audio_quality.get('bitrate'), 1000),
 267             'vbr': float_or_none(video_quality.get('bitrate'), 1000),
 268             'height': resolution.get('height'),
 269             'width': resolution.get('width'),
 270         }
 271
 272     def _real_extract(self, url):
 273         video_id = self._match_id(url)
 274
 275         # Get video webpage. We are not actually interested in it for normal
 276         # cases, but need the cookies in order to be able to download the
 277         # info webpage
 278         webpage, handle = self._download_webpage_handle(
 279             'http://www.nicovideo.jp/watch/' + video_id, video_id)
 280         if video_id.startswith('so'):
 281             video_id = self._match_id(handle.geturl())
 282
 283         api_data = self._parse_json(self._html_search_regex(
 284             'data-api-data="([^"]+)"', webpage,
 285             'API data', default='{}'), video_id)
 286
 287         def _format_id_from_url(video_url):
 288             return 'economy' if video_real_url.endswith('low') else 'normal'
 289
 290         try:
 291             video_real_url = api_data['video']['smileInfo']['url']
 292         except KeyError:  # Flash videos
 293             # Get flv info
 294             flv_info_webpage = self._download_webpage(
 295                 'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
 296                 video_id, 'Downloading flv info')
 297
 298             flv_info = compat_parse_qs(flv_info_webpage)
 299             if 'url' not in flv_info:
 300                 if 'deleted' in flv_info:
 301                     raise ExtractorError('The video has been deleted.',
 302                                          expected=True)
 303                 elif 'closed' in flv_info:
 304                     raise ExtractorError('Niconico videos now require logging in',
 305                                          expected=True)
 306                 elif 'error' in flv_info:
 307                     raise ExtractorError('%s reports error: %s' % (
 308                         self.IE_NAME, flv_info['error'][0]), expected=True)
 309                 else:
 310                     raise ExtractorError('Unable to find video URL')
 311
 312             video_info_xml = self._download_xml(
 313                 'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id,
 314                 video_id, note='Downloading video info page')
 315
 316             def get_video_info(items):
 317                 if not isinstance(items, list):
 318                     items = [items]
 319                 for item in items:
 320                     ret = xpath_text(video_info_xml, './/' + item)
 321                     if ret:
 322                         return ret
 323
 324             video_real_url = flv_info['url'][0]
 325
 326             extension = get_video_info('movie_type')
 327             if not extension:
 328                 extension = determine_ext(video_real_url)
 329
 330             formats = [{
 331                 'url': video_real_url,
 332                 'ext': extension,
 333                 'format_id': _format_id_from_url(video_real_url),
 334             }]
 335         else:
 336             formats = []
 337
 338             dmc_info = api_data['video'].get('dmcInfo')
 339             if dmc_info:  # "New" HTML5 videos
 340                 quality_info = dmc_info['quality']
 341                 for audio_quality in quality_info['audios']:
 342                     for video_quality in quality_info['videos']:
 343                         if not audio_quality['available'] or not video_quality['available']:
 344                             continue
 345                         formats.append(self._extract_format_for_quality(
 346                             api_data, video_id, audio_quality, video_quality))
 347
 348                 self._sort_formats(formats)
 349             else:  # "Old" HTML5 videos
 350                 formats = [{
 351                     'url': video_real_url,
 352                     'ext': 'mp4',
 353                     'format_id': _format_id_from_url(video_real_url),
 354                 }]
 355
 356             def get_video_info(items):
 357                 return dict_get(api_data['video'], items)
 358
 359         # Start extracting information
 360         title = get_video_info('title')
 361         if not title:
 362             title = self._og_search_title(webpage, default=None)
 363         if not title:
 364             title = self._html_search_regex(
 365                 r'<span[^>]+class="videoHeaderTitle"[^>]*>([^<]+)</span>',
 366                 webpage, 'video title')
 367
 368         watch_api_data_string = self._html_search_regex(
 369             r'<div[^>]+id="watchAPIDataContainer"[^>]+>([^<]+)</div>',
 370             webpage, 'watch api data', default=None)
 371         watch_api_data = self._parse_json(watch_api_data_string, video_id) if watch_api_data_string else {}
 372         video_detail = watch_api_data.get('videoDetail', {})
 373
 374         thumbnail = (
 375             get_video_info(['thumbnail_url', 'thumbnailURL'])
 376             or self._html_search_meta('image', webpage, 'thumbnail', default=None)
 377             or video_detail.get('thumbnail'))
 378
 379         description = get_video_info('description')
 380
 381         timestamp = (parse_iso8601(get_video_info('first_retrieve'))
 382                      or unified_timestamp(get_video_info('postedDateTime')))
 383         if not timestamp:
 384             match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
 385             if match:
 386                 timestamp = parse_iso8601(match.replace('+', ':00+'))
 387         if not timestamp and video_detail.get('postedAt'):
 388             timestamp = parse_iso8601(
 389                 video_detail['postedAt'].replace('/', '-'),
 390                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 391
 392         view_count = int_or_none(get_video_info(['view_counter', 'viewCount']))
 393         if not view_count:
 394             match = self._html_search_regex(
 395                 r'>Views: <strong[^>]*>([^<]+)</strong>',
 396                 webpage, 'view count', default=None)
 397             if match:
 398                 view_count = int_or_none(match.replace(',', ''))
 399         view_count = view_count or video_detail.get('viewCount')
 400
 401         comment_count = (int_or_none(get_video_info('comment_num'))
 402                          or video_detail.get('commentCount')
 403                          or try_get(api_data, lambda x: x['thread']['commentCount']))
 404         if not comment_count:
 405             match = self._html_search_regex(
 406                 r'>Comments: <strong[^>]*>([^<]+)</strong>',
 407                 webpage, 'comment count', default=None)
 408             if match:
 409                 comment_count = int_or_none(match.replace(',', ''))
 410
 411         duration = (parse_duration(
 412             get_video_info('length')
 413             or self._html_search_meta(
 414                 'video:duration', webpage, 'video duration', default=None))
 415             or video_detail.get('length')
 416             or get_video_info('duration'))
 417
 418         webpage_url = get_video_info('watch_url') or url
 419
 420         # Note: cannot use api_data.get('owner', {}) because owner may be set to "null"
 421         # in the JSON, which will cause None to be returned instead of {}.
 422         owner = try_get(api_data, lambda x: x.get('owner'), dict) or {}
 423         uploader_id = get_video_info(['ch_id', 'user_id']) or owner.get('id')
 424         uploader = get_video_info(['ch_name', 'user_nickname']) or owner.get('nickname')
 425
 426         return {
 427             'id': video_id,
 428             'title': title,
 429             'formats': formats,
 430             'thumbnail': thumbnail,
 431             'description': description,
 432             'uploader': uploader,
 433             'timestamp': timestamp,
 434             'uploader_id': uploader_id,
 435             'view_count': view_count,
 436             'comment_count': comment_count,
 437             'duration': duration,
 438             'webpage_url': webpage_url,
 439         }
 440
 441
 442 class NiconicoPlaylistIE(InfoExtractor):
 443     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/(?:user/\d+/)?mylist/(?P<id>\d+)'
 444
 445     _TESTS = [{
 446         'url': 'http://www.nicovideo.jp/mylist/27411728',
 447         'info_dict': {
 448             'id': '27411728',
 449             'title': 'AKB48のオールナイトニッポン',
 450             'description': 'md5:d89694c5ded4b6c693dea2db6e41aa08',
 451             'uploader': 'のっく',
 452             'uploader_id': '805442',
 453         },
 454         'playlist_mincount': 225,
 455     }, {
 456         'url': 'https://www.nicovideo.jp/user/805442/mylist/27411728',
 457         'only_matching': True,
 458     }]
 459     _PAGE_SIZE = 100
 460
 461     def _call_api(self, list_id, resource, query):
 462         return self._download_json(
 463             'https://nvapi.nicovideo.jp/v2/mylists/' + list_id, list_id,
 464             'Downloading %s JSON metatdata' % resource, query=query,
 465             headers={'X-Frontend-Id': 6})['data']['mylist']
 466
 467     def _parse_owner(self, item):
 468         owner = item.get('owner') or {}
 469         if owner:
 470             return {
 471                 'uploader': owner.get('name'),
 472                 'uploader_id': owner.get('id'),
 473             }
 474         return {}
 475
 476     def _fetch_page(self, list_id, page):
 477         page += 1
 478         items = self._call_api(list_id, 'page %d' % page, {
 479             'page': page,
 480             'pageSize': self._PAGE_SIZE,
 481         })['items']
 482         for item in items:
 483             video = item.get('video') or {}
 484             video_id = video.get('id')
 485             if not video_id:
 486                 continue
 487             count = video.get('count') or {}
 488             get_count = lambda x: int_or_none(count.get(x))
 489             info = {
 490                 '_type': 'url',
 491                 'id': video_id,
 492                 'title': video.get('title'),
 493                 'url': 'https://www.nicovideo.jp/watch/' + video_id,
 494                 'description': video.get('shortDescription'),
 495                 'duration': int_or_none(video.get('duration')),
 496                 'view_count': get_count('view'),
 497                 'comment_count': get_count('comment'),
 498                 'ie_key': NiconicoIE.ie_key(),
 499             }
 500             info.update(self._parse_owner(video))
 501             yield info
 502
 503     def _real_extract(self, url):
 504         list_id = self._match_id(url)
 505         mylist = self._call_api(list_id, 'list', {
 506             'pageSize': 1,
 507         })
 508         entries = InAdvancePagedList(
 509             functools.partial(self._fetch_page, list_id),
 510             math.ceil(mylist['totalItemCount'] / self._PAGE_SIZE),
 511             self._PAGE_SIZE)
 512         result = self.playlist_result(
 513             entries, list_id, mylist.get('name'), mylist.get('description'))
 514         result.update(self._parse_owner(mylist))
 515         return result