youtube_dl/extractor/niconico.py

   1 # encoding: 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_urllib_parse,
  11     compat_urlparse,
  12 )
  13 from ..utils import (
  14     encode_dict,
  15     ExtractorError,
  16     int_or_none,
  17     parse_duration,
  18     parse_iso8601,
  19     sanitized_Request,
  20     xpath_text,
  21     determine_ext,
  22 )
  23
  24
  25 class NiconicoIE(InfoExtractor):
  26     IE_NAME = 'niconico'
  27     IE_DESC = 'ニコニコ動画'
  28
  29     _TESTS = [{
  30         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  31         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  32         'info_dict': {
  33             'id': 'sm22312215',
  34             'ext': 'mp4',
  35             'title': 'Big Buck Bunny',
  36             'uploader': 'takuya0301',
  37             'uploader_id': '2698420',
  38             'upload_date': '20131123',
  39             'timestamp': 1385182762,
  40             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  41             'duration': 33,
  42         },
  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     }, {
  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': 1198527840,  # timestamp field has different value if logged in
  69             'duration': 304,
  70         },
  71     }, {
  72         'url': 'http://www.nicovideo.jp/watch/so22543406',
  73         'info_dict': {
  74             'id': '1388129933',
  75             'ext': 'mp4',
  76             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
  77             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
  78             'timestamp': 1388851200,
  79             'upload_date': '20140104',
  80             'uploader': 'アニメロチャンネル',
  81             'uploader_id': '312',
  82         }
  83     }]
  84
  85     _VALID_URL = r'https?://(?:www\.|secure\.)?nicovideo\.jp/watch/(?P<id>(?:[a-z]{2})?[0-9]+)'
  86     _NETRC_MACHINE = 'niconico'
  87     # Determine whether the downloader used authentication to download video
  88     _AUTHENTICATED = False
  89
  90     def _real_initialize(self):
  91         self._login()
  92
  93     def _login(self):
  94         (username, password) = self._get_login_info()
  95         # No authentication to be performed
  96         if not username:
  97             return True
  98
  99         # Log in
 100         login_form_strs = {
 101             'mail': username,
 102             'password': password,
 103         }
 104         login_data = compat_urllib_parse.urlencode(encode_dict(login_form_strs)).encode('utf-8')
 105         request = sanitized_Request(
 106             'https://secure.nicovideo.jp/secure/login', login_data)
 107         login_results = self._download_webpage(
 108             request, None, note='Logging in', errnote='Unable to log in')
 109         if re.search(r'(?i)<h1 class="mb8p4">Log in error</h1>', login_results) is not None:
 110             self._downloader.report_warning('unable to log in: bad username or password')
 111             return False
 112         # Successful login
 113         self._AUTHENTICATED = True
 114         return True
 115
 116     def _real_extract(self, url):
 117         video_id = self._match_id(url)
 118
 119         # Get video webpage. We are not actually interested in it for normal
 120         # cases, but need the cookies in order to be able to download the
 121         # info webpage
 122         webpage, handle = self._download_webpage_handle(
 123             'http://www.nicovideo.jp/watch/' + video_id, video_id)
 124         if video_id.startswith('so'):
 125             video_id = self._match_id(handle.geturl())
 126
 127         video_info = self._download_xml(
 128             'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id, video_id,
 129             note='Downloading video info page')
 130
 131         if self._AUTHENTICATED:
 132             # Get flv info
 133             flv_info_webpage = self._download_webpage(
 134                 'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
 135                 video_id, 'Downloading flv info')
 136         else:
 137             # Get external player info
 138             ext_player_info = self._download_webpage(
 139                 'http://ext.nicovideo.jp/thumb_watch/' + video_id, video_id)
 140             thumb_play_key = self._search_regex(
 141                 r'\'thumbPlayKey\'\s*:\s*\'(.*?)\'', ext_player_info, 'thumbPlayKey')
 142
 143             # Get flv info
 144             flv_info_data = compat_urllib_parse.urlencode({
 145                 'k': thumb_play_key,
 146                 'v': video_id
 147             })
 148             flv_info_request = sanitized_Request(
 149                 'http://ext.nicovideo.jp/thumb_watch', flv_info_data,
 150                 {'Content-Type': 'application/x-www-form-urlencoded'})
 151             flv_info_webpage = self._download_webpage(
 152                 flv_info_request, video_id,
 153                 note='Downloading flv info', errnote='Unable to download flv info')
 154
 155         flv_info = compat_urlparse.parse_qs(flv_info_webpage)
 156         if 'url' not in flv_info:
 157             if 'deleted' in flv_info:
 158                 raise ExtractorError('The video has been deleted.',
 159                                      expected=True)
 160             else:
 161                 raise ExtractorError('Unable to find video URL')
 162
 163         video_real_url = flv_info['url'][0]
 164
 165         # Start extracting information
 166         title = xpath_text(video_info, './/title')
 167         if not title:
 168             title = self._og_search_title(webpage, default=None)
 169         if not title:
 170             title = self._html_search_regex(
 171                 r'<span[^>]+class="videoHeaderTitle"[^>]*>([^<]+)</span>',
 172                 webpage, 'video title')
 173
 174         watch_api_data_string = self._html_search_regex(
 175             r'<div[^>]+id="watchAPIDataContainer"[^>]+>([^<]+)</div>',
 176             webpage, 'watch api data', default=None)
 177         watch_api_data = self._parse_json(watch_api_data_string, video_id) if watch_api_data_string else {}
 178         video_detail = watch_api_data.get('videoDetail', {})
 179
 180         extension = xpath_text(video_info, './/movie_type')
 181         if not extension:
 182             extension = determine_ext(video_real_url)
 183
 184         thumbnail = (
 185             xpath_text(video_info, './/thumbnail_url') or
 186             self._html_search_meta('image', webpage, 'thumbnail', default=None) or
 187             video_detail.get('thumbnail'))
 188
 189         description = xpath_text(video_info, './/description')
 190
 191         timestamp = parse_iso8601(xpath_text(video_info, './/first_retrieve'))
 192         if not timestamp:
 193             match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
 194             if match:
 195                 timestamp = parse_iso8601(match.replace('+', ':00+'))
 196         if not timestamp and video_detail.get('postedAt'):
 197             timestamp = parse_iso8601(
 198                 video_detail['postedAt'].replace('/', '-'),
 199                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 200
 201         view_count = int_or_none(xpath_text(video_info, './/view_counter'))
 202         if not view_count:
 203             match = self._html_search_regex(
 204                 r'>Views: <strong[^>]*>([^<]+)</strong>',
 205                 webpage, 'view count', default=None)
 206             if match:
 207                 view_count = int_or_none(match.replace(',', ''))
 208         view_count = view_count or video_detail.get('viewCount')
 209
 210         comment_count = int_or_none(xpath_text(video_info, './/comment_num'))
 211         if not comment_count:
 212             match = self._html_search_regex(
 213                 r'>Comments: <strong[^>]*>([^<]+)</strong>',
 214                 webpage, 'comment count', default=None)
 215             if match:
 216                 comment_count = int_or_none(match.replace(',', ''))
 217         comment_count = comment_count or video_detail.get('commentCount')
 218
 219         duration = (parse_duration(
 220             xpath_text(video_info, './/length') or
 221             self._html_search_meta(
 222                 'video:duration', webpage, 'video duration', default=None)) or
 223             video_detail.get('length'))
 224
 225         webpage_url = xpath_text(video_info, './/watch_url') or url
 226
 227         if video_info.find('.//ch_id') is not None:
 228             uploader_id = video_info.find('.//ch_id').text
 229             uploader = video_info.find('.//ch_name').text
 230         elif video_info.find('.//user_id') is not None:
 231             uploader_id = video_info.find('.//user_id').text
 232             uploader = video_info.find('.//user_nickname').text
 233         else:
 234             uploader_id = uploader = None
 235
 236         return {
 237             'id': video_id,
 238             'url': video_real_url,
 239             'title': title,
 240             'ext': extension,
 241             'format_id': 'economy' if video_real_url.endswith('low') else 'normal',
 242             'thumbnail': thumbnail,
 243             'description': description,
 244             'uploader': uploader,
 245             'timestamp': timestamp,
 246             'uploader_id': uploader_id,
 247             'view_count': view_count,
 248             'comment_count': comment_count,
 249             'duration': duration,
 250             'webpage_url': webpage_url,
 251         }
 252
 253
 254 class NiconicoPlaylistIE(InfoExtractor):
 255     _VALID_URL = r'https?://www\.nicovideo\.jp/mylist/(?P<id>\d+)'
 256
 257     _TEST = {
 258         'url': 'http://www.nicovideo.jp/mylist/27411728',
 259         'info_dict': {
 260             'id': '27411728',
 261             'title': 'AKB48のオールナイトニッポン',
 262         },
 263         'playlist_mincount': 225,
 264     }
 265
 266     def _real_extract(self, url):
 267         list_id = self._match_id(url)
 268         webpage = self._download_webpage(url, list_id)
 269
 270         entries_json = self._search_regex(r'Mylist\.preload\(\d+, (\[.*\])\);',
 271                                           webpage, 'entries')
 272         entries = json.loads(entries_json)
 273         entries = [{
 274             '_type': 'url',
 275             'ie_key': NiconicoIE.ie_key(),
 276             'url': ('http://www.nicovideo.jp/watch/%s' %
 277                     entry['item_data']['video_id']),
 278         } for entry in entries]
 279
 280         return {
 281             '_type': 'playlist',
 282             'title': self._search_regex(r'\s+name: "(.*?)"', webpage, 'title'),
 283             'id': list_id,
 284             'entries': entries,
 285         }