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