Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/niconico.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import datetime
   6
   7 from .common import InfoExtractor
   8 from ..compat import (
   9     compat_parse_qs,
  10     compat_urlparse,
  11 )
  12 from ..utils import (
  13     determine_ext,
  14     dict_get,
  15     ExtractorError,
  16     int_or_none,
  17     float_or_none,
  18     parse_duration,
  19     parse_iso8601,
  20     remove_start,
  21     try_get,
  22     unified_timestamp,
  23     urlencode_postdata,
  24     xpath_text,
  25 )
  26
  27
  28 class NiconicoIE(InfoExtractor):
  29     IE_NAME = 'niconico'
  30     IE_DESC = 'ニコニコ動画'
  31
  32     _TESTS = [{
  33         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  34         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  35         'info_dict': {
  36             'id': 'sm22312215',
  37             'ext': 'mp4',
  38             'title': 'Big Buck Bunny',
  39             'thumbnail': r're:https?://.*',
  40             'uploader': 'takuya0301',
  41             'uploader_id': '2698420',
  42             'upload_date': '20131123',
  43             'timestamp': 1385182762,
  44             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  45             'duration': 33,
  46             'view_count': int,
  47             'comment_count': int,
  48         },
  49         'skip': 'Requires an account',
  50     }, {
  51         # File downloaded with and without credentials are different, so omit
  52         # the md5 field
  53         'url': 'http://www.nicovideo.jp/watch/nm14296458',
  54         'info_dict': {
  55             'id': 'nm14296458',
  56             'ext': 'swf',
  57             'title': '【鏡音リン】Dance on media【オリジナル】take2!',
  58             'description': 'md5:689f066d74610b3b22e0f1739add0f58',
  59             'thumbnail': r're:https?://.*',
  60             'uploader': 'りょうた',
  61             'uploader_id': '18822557',
  62             'upload_date': '20110429',
  63             'timestamp': 1304065916,
  64             'duration': 209,
  65         },
  66         'skip': 'Requires an account',
  67     }, {
  68         # 'video exists but is marked as "deleted"
  69         # md5 is unstable
  70         'url': 'http://www.nicovideo.jp/watch/sm10000',
  71         'info_dict': {
  72             'id': 'sm10000',
  73             'ext': 'unknown_video',
  74             'description': 'deleted',
  75             'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞',
  76             'thumbnail': r're:https?://.*',
  77             'upload_date': '20071224',
  78             'timestamp': int,  # timestamp field has different value if logged in
  79             'duration': 304,
  80             'view_count': int,
  81         },
  82         'skip': 'Requires an account',
  83     }, {
  84         'url': 'http://www.nicovideo.jp/watch/so22543406',
  85         'info_dict': {
  86             'id': '1388129933',
  87             'ext': 'mp4',
  88             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
  89             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
  90             'thumbnail': r're:https?://.*',
  91             'timestamp': 1388851200,
  92             'upload_date': '20140104',
  93             'uploader': 'アニメロチャンネル',
  94             'uploader_id': '312',
  95         },
  96         'skip': 'The viewing period of the video you were searching for has expired.',
  97     }, {
  98         # video not available via `getflv`; "old" HTML5 video
  99         'url': 'http://www.nicovideo.jp/watch/sm1151009',
 100         'md5': '8fa81c364eb619d4085354eab075598a',
 101         'info_dict': {
 102             'id': 'sm1151009',
 103             'ext': 'mp4',
 104             '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）',
 105             'description': 'md5:6ee077e0581ff5019773e2e714cdd0b7',
 106             'thumbnail': r're:https?://.*',
 107             'duration': 184,
 108             'timestamp': 1190868283,
 109             'upload_date': '20070927',
 110             'uploader': 'denden2',
 111             'uploader_id': '1392194',
 112             'view_count': int,
 113             'comment_count': int,
 114         },
 115         'skip': 'Requires an account',
 116     }, {
 117         # "New" HTML5 video
 118         'url': 'http://www.nicovideo.jp/watch/sm31464864',
 119         'md5': '351647b4917660986dc0fa8864085135',
 120         'info_dict': {
 121             'id': 'sm31464864',
 122             'ext': 'mp4',
 123             'title': '新作TVアニメ「戦姫絶唱シンフォギアAXZ」PV 最高画質',
 124             'description': 'md5:e52974af9a96e739196b2c1ca72b5feb',
 125             'timestamp': 1498514060,
 126             'upload_date': '20170626',
 127             'uploader': 'ゲス',
 128             'uploader_id': '40826363',
 129             'thumbnail': r're:https?://.*',
 130             'duration': 198,
 131             'view_count': int,
 132             'comment_count': int,
 133         },
 134         'skip': 'Requires an account',
 135     }, {
 136         'url': 'http://sp.nicovideo.jp/watch/sm28964488?ss_pos=1&cp_in=wt_tg',
 137         'only_matching': True,
 138     }]
 139
 140     _VALID_URL = r'https?://(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch/(?P<id>(?:[a-z]{2})?[0-9]+)'
 141     _NETRC_MACHINE = 'niconico'
 142
 143     def _real_initialize(self):
 144         self._login()
 145
 146     def _login(self):
 147         (username, password) = self._get_login_info()
 148         # No authentication to be performed
 149         if not username:
 150             return True
 151
 152         # Log in
 153         login_ok = True
 154         login_form_strs = {
 155             'mail_tel': username,
 156             'password': password,
 157         }
 158         urlh = self._request_webpage(
 159             'https://account.nicovideo.jp/api/v1/login', None,
 160             note='Logging in', errnote='Unable to log in',
 161             data=urlencode_postdata(login_form_strs))
 162         if urlh is False:
 163             login_ok = False
 164         else:
 165             parts = compat_urlparse.urlparse(urlh.geturl())
 166             if compat_parse_qs(parts.query).get('message', [None])[0] == 'cant_login':
 167                 login_ok = False
 168         if not login_ok:
 169             self._downloader.report_warning('unable to log in: bad username or password')
 170         return login_ok
 171
 172     def _extract_format_for_quality(self, api_data, video_id, audio_quality, video_quality):
 173         def yesno(boolean):
 174             return 'yes' if boolean else 'no'
 175
 176         session_api_data = api_data['video']['dmcInfo']['session_api']
 177         session_api_endpoint = session_api_data['urls'][0]
 178
 179         format_id = '-'.join(map(lambda s: remove_start(s['id'], 'archive_'), [video_quality, audio_quality]))
 180
 181         session_response = self._download_json(
 182             session_api_endpoint['url'], video_id,
 183             query={'_format': 'json'},
 184             headers={'Content-Type': 'application/json'},
 185             note='Downloading JSON metadata for %s' % format_id,
 186             data=json.dumps({
 187                 'session': {
 188                     'client_info': {
 189                         'player_id': session_api_data['player_id'],
 190                     },
 191                     'content_auth': {
 192                         'auth_type': session_api_data['auth_types'][session_api_data['protocols'][0]],
 193                         'content_key_timeout': session_api_data['content_key_timeout'],
 194                         'service_id': 'nicovideo',
 195                         'service_user_id': session_api_data['service_user_id']
 196                     },
 197                     'content_id': session_api_data['content_id'],
 198                     'content_src_id_sets': [{
 199                         'content_src_ids': [{
 200                             'src_id_to_mux': {
 201                                 'audio_src_ids': [audio_quality['id']],
 202                                 'video_src_ids': [video_quality['id']],
 203                             }
 204                         }]
 205                     }],
 206                     'content_type': 'movie',
 207                     'content_uri': '',
 208                     'keep_method': {
 209                         'heartbeat': {
 210                             'lifetime': session_api_data['heartbeat_lifetime']
 211                         }
 212                     },
 213                     'priority': session_api_data['priority'],
 214                     'protocol': {
 215                         'name': 'http',
 216                         'parameters': {
 217                             'http_parameters': {
 218                                 'parameters': {
 219                                     'http_output_download_parameters': {
 220                                         'use_ssl': yesno(session_api_endpoint['is_ssl']),
 221                                         'use_well_known_port': yesno(session_api_endpoint['is_well_known_port']),
 222                                     }
 223                                 }
 224                             }
 225                         }
 226                     },
 227                     'recipe_id': session_api_data['recipe_id'],
 228                     'session_operation_auth': {
 229                         'session_operation_auth_by_signature': {
 230                             'signature': session_api_data['signature'],
 231                             'token': session_api_data['token'],
 232                         }
 233                     },
 234                     'timing_constraint': 'unlimited'
 235                 }
 236             }))
 237
 238         resolution = video_quality.get('resolution', {})
 239
 240         return {
 241             'url': session_response['data']['session']['content_uri'],
 242             'format_id': format_id,
 243             'ext': 'mp4',  # Session API are used in HTML5, which always serves mp4
 244             'abr': float_or_none(audio_quality.get('bitrate'), 1000),
 245             'vbr': float_or_none(video_quality.get('bitrate'), 1000),
 246             'height': resolution.get('height'),
 247             'width': resolution.get('width'),
 248         }
 249
 250     def _real_extract(self, url):
 251         video_id = self._match_id(url)
 252
 253         # Get video webpage. We are not actually interested in it for normal
 254         # cases, but need the cookies in order to be able to download the
 255         # info webpage
 256         webpage, handle = self._download_webpage_handle(
 257             'http://www.nicovideo.jp/watch/' + video_id, video_id)
 258         if video_id.startswith('so'):
 259             video_id = self._match_id(handle.geturl())
 260
 261         api_data = self._parse_json(self._html_search_regex(
 262             'data-api-data="([^"]+)"', webpage,
 263             'API data', default='{}'), video_id)
 264
 265         def _format_id_from_url(video_url):
 266             return 'economy' if video_real_url.endswith('low') else 'normal'
 267
 268         try:
 269             video_real_url = api_data['video']['smileInfo']['url']
 270         except KeyError:  # Flash videos
 271             # Get flv info
 272             flv_info_webpage = self._download_webpage(
 273                 'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
 274                 video_id, 'Downloading flv info')
 275
 276             flv_info = compat_urlparse.parse_qs(flv_info_webpage)
 277             if 'url' not in flv_info:
 278                 if 'deleted' in flv_info:
 279                     raise ExtractorError('The video has been deleted.',
 280                                          expected=True)
 281                 elif 'closed' in flv_info:
 282                     raise ExtractorError('Niconico videos now require logging in',
 283                                          expected=True)
 284                 elif 'error' in flv_info:
 285                     raise ExtractorError('%s reports error: %s' % (
 286                         self.IE_NAME, flv_info['error'][0]), expected=True)
 287                 else:
 288                     raise ExtractorError('Unable to find video URL')
 289
 290             video_info_xml = self._download_xml(
 291                 'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id,
 292                 video_id, note='Downloading video info page')
 293
 294             def get_video_info(items):
 295                 if not isinstance(items, list):
 296                     items = [items]
 297                 for item in items:
 298                     ret = xpath_text(video_info_xml, './/' + item)
 299                     if ret:
 300                         return ret
 301
 302             video_real_url = flv_info['url'][0]
 303
 304             extension = get_video_info('movie_type')
 305             if not extension:
 306                 extension = determine_ext(video_real_url)
 307
 308             formats = [{
 309                 'url': video_real_url,
 310                 'ext': extension,
 311                 'format_id': _format_id_from_url(video_real_url),
 312             }]
 313         else:
 314             formats = []
 315
 316             dmc_info = api_data['video'].get('dmcInfo')
 317             if dmc_info:  # "New" HTML5 videos
 318                 quality_info = dmc_info['quality']
 319                 for audio_quality in quality_info['audios']:
 320                     for video_quality in quality_info['videos']:
 321                         if not audio_quality['available'] or not video_quality['available']:
 322                             continue
 323                         formats.append(self._extract_format_for_quality(
 324                             api_data, video_id, audio_quality, video_quality))
 325
 326                 self._sort_formats(formats)
 327             else:  # "Old" HTML5 videos
 328                 formats = [{
 329                     'url': video_real_url,
 330                     'ext': 'mp4',
 331                     'format_id': _format_id_from_url(video_real_url),
 332                 }]
 333
 334             def get_video_info(items):
 335                 return dict_get(api_data['video'], items)
 336
 337         # Start extracting information
 338         title = get_video_info('title')
 339         if not title:
 340             title = self._og_search_title(webpage, default=None)
 341         if not title:
 342             title = self._html_search_regex(
 343                 r'<span[^>]+class="videoHeaderTitle"[^>]*>([^<]+)</span>',
 344                 webpage, 'video title')
 345
 346         watch_api_data_string = self._html_search_regex(
 347             r'<div[^>]+id="watchAPIDataContainer"[^>]+>([^<]+)</div>',
 348             webpage, 'watch api data', default=None)
 349         watch_api_data = self._parse_json(watch_api_data_string, video_id) if watch_api_data_string else {}
 350         video_detail = watch_api_data.get('videoDetail', {})
 351
 352         thumbnail = (
 353             get_video_info(['thumbnail_url', 'thumbnailURL']) or
 354             self._html_search_meta('image', webpage, 'thumbnail', default=None) or
 355             video_detail.get('thumbnail'))
 356
 357         description = get_video_info('description')
 358
 359         timestamp = (parse_iso8601(get_video_info('first_retrieve')) or
 360                      unified_timestamp(get_video_info('postedDateTime')))
 361         if not timestamp:
 362             match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
 363             if match:
 364                 timestamp = parse_iso8601(match.replace('+', ':00+'))
 365         if not timestamp and video_detail.get('postedAt'):
 366             timestamp = parse_iso8601(
 367                 video_detail['postedAt'].replace('/', '-'),
 368                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 369
 370         view_count = int_or_none(get_video_info(['view_counter', 'viewCount']))
 371         if not view_count:
 372             match = self._html_search_regex(
 373                 r'>Views: <strong[^>]*>([^<]+)</strong>',
 374                 webpage, 'view count', default=None)
 375             if match:
 376                 view_count = int_or_none(match.replace(',', ''))
 377         view_count = view_count or video_detail.get('viewCount')
 378
 379         comment_count = (int_or_none(get_video_info('comment_num')) or
 380                          video_detail.get('commentCount') or
 381                          try_get(api_data, lambda x: x['thread']['commentCount']))
 382         if not comment_count:
 383             match = self._html_search_regex(
 384                 r'>Comments: <strong[^>]*>([^<]+)</strong>',
 385                 webpage, 'comment count', default=None)
 386             if match:
 387                 comment_count = int_or_none(match.replace(',', ''))
 388
 389         duration = (parse_duration(
 390             get_video_info('length') or
 391             self._html_search_meta(
 392                 'video:duration', webpage, 'video duration', default=None)) or
 393             video_detail.get('length') or
 394             get_video_info('duration'))
 395
 396         webpage_url = get_video_info('watch_url') or url
 397
 398         owner = api_data.get('owner', {})
 399         uploader_id = get_video_info(['ch_id', 'user_id']) or owner.get('id')
 400         uploader = get_video_info(['ch_name', 'user_nickname']) or owner.get('nickname')
 401
 402         return {
 403             'id': video_id,
 404             'title': title,
 405             'formats': formats,
 406             'thumbnail': thumbnail,
 407             'description': description,
 408             'uploader': uploader,
 409             'timestamp': timestamp,
 410             'uploader_id': uploader_id,
 411             'view_count': view_count,
 412             'comment_count': comment_count,
 413             'duration': duration,
 414             'webpage_url': webpage_url,
 415         }
 416
 417
 418 class NiconicoPlaylistIE(InfoExtractor):
 419     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/mylist/(?P<id>\d+)'
 420
 421     _TEST = {
 422         'url': 'http://www.nicovideo.jp/mylist/27411728',
 423         'info_dict': {
 424             'id': '27411728',
 425             'title': 'AKB48のオールナイトニッポン',
 426         },
 427         'playlist_mincount': 225,
 428     }
 429
 430     def _real_extract(self, url):
 431         list_id = self._match_id(url)
 432         webpage = self._download_webpage(url, list_id)
 433
 434         entries_json = self._search_regex(r'Mylist\.preload\(\d+, (\[.*\])\);',
 435                                           webpage, 'entries')
 436         entries = json.loads(entries_json)
 437         entries = [{
 438             '_type': 'url',
 439             'ie_key': NiconicoIE.ie_key(),
 440             'url': ('http://www.nicovideo.jp/watch/%s' %
 441                     entry['item_data']['video_id']),
 442         } for entry in entries]
 443
 444         return {
 445             '_type': 'playlist',
 446             'title': self._search_regex(r'\s+name: "(.*?)"', webpage, 'title'),
 447             'id': list_id,
 448             'entries': entries,
 449         }