_ Git - youtube-dl/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     parse_duration,
  18     parse_iso8601,
  19     try_get,
  20     unified_timestamp,
  21     urlencode_postdata,
  22     xpath_text,
  23 )
  24
  25
  26 class NiconicoIE(InfoExtractor):
  27     IE_NAME = 'niconico'
  28     IE_DESC = 'ニコニコ動画'
  29
  30     _TESTS = [{
  31         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  32         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  33         'info_dict': {
  34             'id': 'sm22312215',
  35             'ext': 'mp4',
  36             'title': 'Big Buck Bunny',
  37             'thumbnail': r're:https?://.*',
  38             'uploader': 'takuya0301',
  39             'uploader_id': '2698420',
  40             'upload_date': '20131123',
  41             'timestamp': 1385182762,
  42             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  43             'duration': 33,
  44             'view_count': int,
  45             'comment_count': int,
  46         },
  47         'skip': 'Requires an account',
  48     }, {
  49         # File downloaded with and without credentials are different, so omit
  50         # the md5 field
  51         'url': 'http://www.nicovideo.jp/watch/nm14296458',
  52         'info_dict': {
  53             'id': 'nm14296458',
  54             'ext': 'swf',
  55             'title': '【鏡音リン】Dance on media【オリジナル】take2!',
  56             'description': 'md5:689f066d74610b3b22e0f1739add0f58',
  57             'thumbnail': r're:https?://.*',
  58             'uploader': 'りょうた',
  59             'uploader_id': '18822557',
  60             'upload_date': '20110429',
  61             'timestamp': 1304065916,
  62             'duration': 209,
  63         },
  64         'skip': 'Requires an account',
  65     }, {
  66         # 'video exists but is marked as "deleted"
  67         # md5 is unstable
  68         'url': 'http://www.nicovideo.jp/watch/sm10000',
  69         'info_dict': {
  70             'id': 'sm10000',
  71             'ext': 'unknown_video',
  72             'description': 'deleted',
  73             'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞',
  74             'thumbnail': r're:https?://.*',
  75             'upload_date': '20071224',
  76             'timestamp': int,  # timestamp field has different value if logged in
  77             'duration': 304,
  78             'view_count': int,
  79         },
  80         'skip': 'Requires an account',
  81     }, {
  82         'url': 'http://www.nicovideo.jp/watch/so22543406',
  83         'info_dict': {
  84             'id': '1388129933',
  85             'ext': 'mp4',
  86             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
  87             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
  88             'thumbnail': r're:https?://.*',
  89             'timestamp': 1388851200,
  90             'upload_date': '20140104',
  91             'uploader': 'アニメロチャンネル',
  92             'uploader_id': '312',
  93         },
  94         'skip': 'The viewing period of the video you were searching for has expired.',
  95     }, {
  96         # video not available via `getflv`
  97         'url': 'http://www.nicovideo.jp/watch/sm1151009',
  98         'info_dict': {
  99             'id': 'sm1151009',
 100             'ext': 'flv',
 101             '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）',
 102             'description': 'md5:6ee077e0581ff5019773e2e714cdd0b7',
 103             'thumbnail': r're:https?://.*',
 104             'duration': 184,
 105             'timestamp': 1190868283,
 106             'upload_date': '20070927',
 107             'uploader': 'denden2',
 108             'uploader_id': '1392194',
 109             'view_count': int,
 110             'comment_count': int,
 111         },
 112         'skip': 'Requires an account',
 113     }, {
 114         'url': 'http://sp.nicovideo.jp/watch/sm28964488?ss_pos=1&cp_in=wt_tg',
 115         'only_matching': True,
 116     }]
 117
 118     _VALID_URL = r'https?://(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch/(?P<id>(?:[a-z]{2})?[0-9]+)'
 119     _NETRC_MACHINE = 'niconico'
 120
 121     def _real_initialize(self):
 122         self._login()
 123
 124     def _login(self):
 125         (username, password) = self._get_login_info()
 126         # No authentication to be performed
 127         if not username:
 128             return True
 129
 130         # Log in
 131         login_ok = True
 132         login_form_strs = {
 133             'mail_tel': username,
 134             'password': password,
 135         }
 136         urlh = self._request_webpage(
 137             'https://account.nicovideo.jp/api/v1/login', None,
 138             note='Logging in', errnote='Unable to log in',
 139             data=urlencode_postdata(login_form_strs))
 140         if urlh is False:
 141             login_ok = False
 142         else:
 143             parts = compat_urlparse.urlparse(urlh.geturl())
 144             if compat_parse_qs(parts.query).get('message', [None])[0] == 'cant_login':
 145                 login_ok = False
 146         if not login_ok:
 147             self._downloader.report_warning('unable to log in: bad username or password')
 148         return login_ok
 149
 150     def _real_extract(self, url):
 151         video_id = self._match_id(url)
 152
 153         # Get video webpage. We are not actually interested in it for normal
 154         # cases, but need the cookies in order to be able to download the
 155         # info webpage
 156         webpage, handle = self._download_webpage_handle(
 157             'http://www.nicovideo.jp/watch/' + video_id, video_id)
 158         if video_id.startswith('so'):
 159             video_id = self._match_id(handle.geturl())
 160
 161         api_data = self._parse_json(self._html_search_regex(
 162             'data-api-data="([^"]+)"', webpage,
 163             'API data', default='{}'), video_id)
 164         video_real_url = try_get(
 165             api_data, lambda x: x['video']['smileInfo']['url'])
 166
 167         if video_real_url:
 168             def get_video_info(items):
 169                 return dict_get(api_data['video'], items)
 170         else:
 171             # Get flv info
 172             flv_info_webpage = self._download_webpage(
 173                 'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
 174                 video_id, 'Downloading flv info')
 175
 176             flv_info = compat_urlparse.parse_qs(flv_info_webpage)
 177             if 'url' not in flv_info:
 178                 if 'deleted' in flv_info:
 179                     raise ExtractorError('The video has been deleted.',
 180                                          expected=True)
 181                 elif 'closed' in flv_info:
 182                     raise ExtractorError('Niconico videos now require logging in',
 183                                          expected=True)
 184                 elif 'error' in flv_info:
 185                     raise ExtractorError('%s reports error: %s' % (
 186                         self.IE_NAME, flv_info['error'][0]), expected=True)
 187                 else:
 188                     raise ExtractorError('Unable to find video URL')
 189
 190             video_real_url = flv_info['url'][0]
 191
 192             video_info_xml = self._download_xml(
 193                 'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id,
 194                 video_id, note='Downloading video info page')
 195
 196             def get_video_info(items):
 197                 if not isinstance(items, list):
 198                     items = [items]
 199                 for item in items:
 200                     ret = xpath_text(video_info_xml, './/' + item)
 201                     if ret:
 202                         return ret
 203
 204         # Start extracting information
 205         title = get_video_info('title')
 206         if not title:
 207             title = self._og_search_title(webpage, default=None)
 208         if not title:
 209             title = self._html_search_regex(
 210                 r'<span[^>]+class="videoHeaderTitle"[^>]*>([^<]+)</span>',
 211                 webpage, 'video title')
 212
 213         watch_api_data_string = self._html_search_regex(
 214             r'<div[^>]+id="watchAPIDataContainer"[^>]+>([^<]+)</div>',
 215             webpage, 'watch api data', default=None)
 216         watch_api_data = self._parse_json(watch_api_data_string, video_id) if watch_api_data_string else {}
 217         video_detail = watch_api_data.get('videoDetail', {})
 218
 219         extension = get_video_info(['movie_type', 'movieType'])
 220         if not extension:
 221             extension = determine_ext(video_real_url)
 222
 223         thumbnail = (
 224             get_video_info(['thumbnail_url', 'thumbnailURL']) or
 225             self._html_search_meta('image', webpage, 'thumbnail', default=None) or
 226             video_detail.get('thumbnail'))
 227
 228         description = get_video_info('description')
 229
 230         timestamp = (parse_iso8601(get_video_info('first_retrieve')) or
 231                      unified_timestamp(get_video_info('postedDateTime')))
 232         if not timestamp:
 233             match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
 234             if match:
 235                 timestamp = parse_iso8601(match.replace('+', ':00+'))
 236         if not timestamp and video_detail.get('postedAt'):
 237             timestamp = parse_iso8601(
 238                 video_detail['postedAt'].replace('/', '-'),
 239                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 240
 241         view_count = int_or_none(get_video_info(['view_counter', 'viewCount']))
 242         if not view_count:
 243             match = self._html_search_regex(
 244                 r'>Views: <strong[^>]*>([^<]+)</strong>',
 245                 webpage, 'view count', default=None)
 246             if match:
 247                 view_count = int_or_none(match.replace(',', ''))
 248         view_count = view_count or video_detail.get('viewCount')
 249
 250         comment_count = (int_or_none(get_video_info('comment_num')) or
 251                          video_detail.get('commentCount') or
 252                          try_get(api_data, lambda x: x['thread']['commentCount']))
 253         if not comment_count:
 254             match = self._html_search_regex(
 255                 r'>Comments: <strong[^>]*>([^<]+)</strong>',
 256                 webpage, 'comment count', default=None)
 257             if match:
 258                 comment_count = int_or_none(match.replace(',', ''))
 259
 260         duration = (parse_duration(
 261             get_video_info('length') or
 262             self._html_search_meta(
 263                 'video:duration', webpage, 'video duration', default=None)) or
 264             video_detail.get('length') or
 265             get_video_info('duration'))
 266
 267         webpage_url = get_video_info('watch_url') or url
 268
 269         owner = api_data.get('owner', {})
 270         uploader_id = get_video_info(['ch_id', 'user_id']) or owner.get('id')
 271         uploader = get_video_info(['ch_name', 'user_nickname']) or owner.get('nickname')
 272
 273         return {
 274             'id': video_id,
 275             'url': video_real_url,
 276             'title': title,
 277             'ext': extension,
 278             'format_id': 'economy' if video_real_url.endswith('low') else 'normal',
 279             'thumbnail': thumbnail,
 280             'description': description,
 281             'uploader': uploader,
 282             'timestamp': timestamp,
 283             'uploader_id': uploader_id,
 284             'view_count': view_count,
 285             'comment_count': comment_count,
 286             'duration': duration,
 287             'webpage_url': webpage_url,
 288         }
 289
 290
 291 class NiconicoPlaylistIE(InfoExtractor):
 292     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/mylist/(?P<id>\d+)'
 293
 294     _TEST = {
 295         'url': 'http://www.nicovideo.jp/mylist/27411728',
 296         'info_dict': {
 297             'id': '27411728',
 298             'title': 'AKB48のオールナイトニッポン',
 299         },
 300         'playlist_mincount': 225,
 301     }
 302
 303     def _real_extract(self, url):
 304         list_id = self._match_id(url)
 305         webpage = self._download_webpage(url, list_id)
 306
 307         entries_json = self._search_regex(r'Mylist\.preload\(\d+, (\[.*\])\);',
 308                                           webpage, 'entries')
 309         entries = json.loads(entries_json)
 310         entries = [{
 311             '_type': 'url',
 312             'ie_key': NiconicoIE.ie_key(),
 313             'url': ('http://www.nicovideo.jp/watch/%s' %
 314                     entry['item_data']['video_id']),
 315         } for entry in entries]
 316
 317         return {
 318             '_type': 'playlist',
 319             'title': self._search_regex(r'\s+name: "(.*?)"', webpage, 'title'),
 320             'id': list_id,
 321             'entries': entries,
 322         }