youtube_dl/extractor/niconico.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import json
   6 import datetime
   7
   8 from .common import InfoExtractor
   9 from ..compat import (
  10     compat_urlparse,
  11 )
  12 from ..utils import (
  13     ExtractorError,
  14     int_or_none,
  15     parse_duration,
  16     parse_iso8601,
  17     sanitized_Request,
  18     xpath_text,
  19     determine_ext,
  20     urlencode_postdata,
  21 )
  22
  23
  24 class NiconicoIE(InfoExtractor):
  25     IE_NAME = 'niconico'
  26     IE_DESC = 'ニコニコ動画'
  27
  28     _TESTS = [{
  29         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  30         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  31         'info_dict': {
  32             'id': 'sm22312215',
  33             'ext': 'mp4',
  34             'title': 'Big Buck Bunny',
  35             'uploader': 'takuya0301',
  36             'uploader_id': '2698420',
  37             'upload_date': '20131123',
  38             'timestamp': 1385182762,
  39             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  40             'duration': 33,
  41         },
  42         'skip': 'Requires an account',
  43     }, {
  44         # File downloaded with and without credentials are different, so omit
  45         # the md5 field
  46         'url': 'http://www.nicovideo.jp/watch/nm14296458',
  47         'info_dict': {
  48             'id': 'nm14296458',
  49             'ext': 'swf',
  50             'title': '【鏡音リン】Dance on media【オリジナル】take2!',
  51             'description': 'md5:689f066d74610b3b22e0f1739add0f58',
  52             'uploader': 'りょうた',
  53             'uploader_id': '18822557',
  54             'upload_date': '20110429',
  55             'timestamp': 1304065916,
  56             'duration': 209,
  57         },
  58         'skip': 'Requires an account',
  59     }, {
  60         # 'video exists but is marked as "deleted"
  61         # md5 is unstable
  62         'url': 'http://www.nicovideo.jp/watch/sm10000',
  63         'info_dict': {
  64             'id': 'sm10000',
  65             'ext': 'unknown_video',
  66             'description': 'deleted',
  67             'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞',
  68             'upload_date': '20071224',
  69             'timestamp': int,  # timestamp field has different value if logged in
  70             'duration': 304,
  71         },
  72         'skip': 'Requires an account',
  73     }, {
  74         'url': 'http://www.nicovideo.jp/watch/so22543406',
  75         'info_dict': {
  76             'id': '1388129933',
  77             'ext': 'mp4',
  78             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
  79             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
  80             'timestamp': 1388851200,
  81             'upload_date': '20140104',
  82             'uploader': 'アニメロチャンネル',
  83             'uploader_id': '312',
  84         },
  85         'skip': 'The viewing period of the video you were searching for has expired.',
  86     }]
  87
  88     _VALID_URL = r'https?://(?:www\.|secure\.)?nicovideo\.jp/watch/(?P<id>(?:[a-z]{2})?[0-9]+)'
  89     _NETRC_MACHINE = 'niconico'
  90
  91     def _real_initialize(self):
  92         self._login()
  93
  94     def _login(self):
  95         (username, password) = self._get_login_info()
  96         # No authentication to be performed
  97         if not username:
  98             return True
  99
 100         # Log in
 101         login_form_strs = {
 102             'mail': username,
 103             'password': password,
 104         }
 105         login_data = urlencode_postdata(login_form_strs)
 106         request = sanitized_Request(
 107             'https://secure.nicovideo.jp/secure/login', login_data)
 108         login_results = self._download_webpage(
 109             request, None, note='Logging in', errnote='Unable to log in')
 110         if re.search(r'(?i)<h1 class="mb8p4">Log in error</h1>', login_results) is not None:
 111             self._downloader.report_warning('unable to log in: bad username or password')
 112             return False
 113         return True
 114
 115     def _real_extract(self, url):
 116         video_id = self._match_id(url)
 117
 118         # Get video webpage. We are not actually interested in it for normal
 119         # cases, but need the cookies in order to be able to download the
 120         # info webpage
 121         webpage, handle = self._download_webpage_handle(
 122             'http://www.nicovideo.jp/watch/' + video_id, video_id)
 123         if video_id.startswith('so'):
 124             video_id = self._match_id(handle.geturl())
 125
 126         video_info = self._download_xml(
 127             'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id, video_id,
 128             note='Downloading video info page')
 129
 130         # Get flv info
 131         flv_info_webpage = self._download_webpage(
 132             'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
 133             video_id, 'Downloading flv info')
 134
 135         flv_info = compat_urlparse.parse_qs(flv_info_webpage)
 136         if 'url' not in flv_info:
 137             if 'deleted' in flv_info:
 138                 raise ExtractorError('The video has been deleted.',
 139                                      expected=True)
 140             elif 'closed' in flv_info:
 141                 raise ExtractorError('Niconico videos now require logging in',
 142                                      expected=True)
 143             else:
 144                 raise ExtractorError('Unable to find video URL')
 145
 146         video_real_url = flv_info['url'][0]
 147
 148         # Start extracting information
 149         title = xpath_text(video_info, './/title')
 150         if not title:
 151             title = self._og_search_title(webpage, default=None)
 152         if not title:
 153             title = self._html_search_regex(
 154                 r'<span[^>]+class="videoHeaderTitle"[^>]*>([^<]+)</span>',
 155                 webpage, 'video title')
 156
 157         watch_api_data_string = self._html_search_regex(
 158             r'<div[^>]+id="watchAPIDataContainer"[^>]+>([^<]+)</div>',
 159             webpage, 'watch api data', default=None)
 160         watch_api_data = self._parse_json(watch_api_data_string, video_id) if watch_api_data_string else {}
 161         video_detail = watch_api_data.get('videoDetail', {})
 162
 163         extension = xpath_text(video_info, './/movie_type')
 164         if not extension:
 165             extension = determine_ext(video_real_url)
 166
 167         thumbnail = (
 168             xpath_text(video_info, './/thumbnail_url') or
 169             self._html_search_meta('image', webpage, 'thumbnail', default=None) or
 170             video_detail.get('thumbnail'))
 171
 172         description = xpath_text(video_info, './/description')
 173
 174         timestamp = parse_iso8601(xpath_text(video_info, './/first_retrieve'))
 175         if not timestamp:
 176             match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
 177             if match:
 178                 timestamp = parse_iso8601(match.replace('+', ':00+'))
 179         if not timestamp and video_detail.get('postedAt'):
 180             timestamp = parse_iso8601(
 181                 video_detail['postedAt'].replace('/', '-'),
 182                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 183
 184         view_count = int_or_none(xpath_text(video_info, './/view_counter'))
 185         if not view_count:
 186             match = self._html_search_regex(
 187                 r'>Views: <strong[^>]*>([^<]+)</strong>',
 188                 webpage, 'view count', default=None)
 189             if match:
 190                 view_count = int_or_none(match.replace(',', ''))
 191         view_count = view_count or video_detail.get('viewCount')
 192
 193         comment_count = int_or_none(xpath_text(video_info, './/comment_num'))
 194         if not comment_count:
 195             match = self._html_search_regex(
 196                 r'>Comments: <strong[^>]*>([^<]+)</strong>',
 197                 webpage, 'comment count', default=None)
 198             if match:
 199                 comment_count = int_or_none(match.replace(',', ''))
 200         comment_count = comment_count or video_detail.get('commentCount')
 201
 202         duration = (parse_duration(
 203             xpath_text(video_info, './/length') or
 204             self._html_search_meta(
 205                 'video:duration', webpage, 'video duration', default=None)) or
 206             video_detail.get('length'))
 207
 208         webpage_url = xpath_text(video_info, './/watch_url') or url
 209
 210         if video_info.find('.//ch_id') is not None:
 211             uploader_id = video_info.find('.//ch_id').text
 212             uploader = video_info.find('.//ch_name').text
 213         elif video_info.find('.//user_id') is not None:
 214             uploader_id = video_info.find('.//user_id').text
 215             uploader = video_info.find('.//user_nickname').text
 216         else:
 217             uploader_id = uploader = None
 218
 219         return {
 220             'id': video_id,
 221             'url': video_real_url,
 222             'title': title,
 223             'ext': extension,
 224             'format_id': 'economy' if video_real_url.endswith('low') else 'normal',
 225             'thumbnail': thumbnail,
 226             'description': description,
 227             'uploader': uploader,
 228             'timestamp': timestamp,
 229             'uploader_id': uploader_id,
 230             'view_count': view_count,
 231             'comment_count': comment_count,
 232             'duration': duration,
 233             'webpage_url': webpage_url,
 234         }
 235
 236
 237 class NiconicoPlaylistIE(InfoExtractor):
 238     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/mylist/(?P<id>\d+)'
 239
 240     _TEST = {
 241         'url': 'http://www.nicovideo.jp/mylist/27411728',
 242         'info_dict': {
 243             'id': '27411728',
 244             'title': 'AKB48のオールナイトニッポン',
 245         },
 246         'playlist_mincount': 225,
 247     }
 248
 249     def _real_extract(self, url):
 250         list_id = self._match_id(url)
 251         webpage = self._download_webpage(url, list_id)
 252
 253         entries_json = self._search_regex(r'Mylist\.preload\(\d+, (\[.*\])\);',
 254                                           webpage, 'entries')
 255         entries = json.loads(entries_json)
 256         entries = [{
 257             '_type': 'url',
 258             'ie_key': NiconicoIE.ie_key(),
 259             'url': ('http://www.nicovideo.jp/watch/%s' %
 260                     entry['item_data']['video_id']),
 261         } for entry in entries]
 262
 263         return {
 264             '_type': 'playlist',
 265             'title': self._search_regex(r'\s+name: "(.*?)"', webpage, 'title'),
 266             'id': list_id,
 267             'entries': entries,
 268         }