_ Git - youtube-dl/blob - 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     # Determine whether the downloader used authentication to download video
  91     _AUTHENTICATED = False
  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_form_strs = {
 104             'mail': username,
 105             'password': password,
 106         }
 107         login_data = urlencode_postdata(login_form_strs)
 108         request = sanitized_Request(
 109             'https://secure.nicovideo.jp/secure/login', login_data)
 110         login_results = self._download_webpage(
 111             request, None, note='Logging in', errnote='Unable to log in')
 112         if re.search(r'(?i)<h1 class="mb8p4">Log in error</h1>', login_results) is not None:
 113             self._downloader.report_warning('unable to log in: bad username or password')
 114             return False
 115         # Successful login
 116         self._AUTHENTICATED = True
 117         return True
 118
 119     def _real_extract(self, url):
 120         video_id = self._match_id(url)
 121
 122         # Get video webpage. We are not actually interested in it for normal
 123         # cases, but need the cookies in order to be able to download the
 124         # info webpage
 125         webpage, handle = self._download_webpage_handle(
 126             'http://www.nicovideo.jp/watch/' + video_id, video_id)
 127         if video_id.startswith('so'):
 128             video_id = self._match_id(handle.geturl())
 129
 130         video_info = self._download_xml(
 131             'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id, video_id,
 132             note='Downloading video info page')
 133
 134         if self._AUTHENTICATED:
 135             # Get flv info
 136             flv_info_webpage = self._download_webpage(
 137                 'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
 138                 video_id, 'Downloading flv info')
 139         else:
 140             raise ExtractorError('Niconico videos now require logging in', expected=True)
 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             else:
 148                 raise ExtractorError('Unable to find video URL')
 149
 150         video_real_url = flv_info['url'][0]
 151
 152         # Start extracting information
 153         title = xpath_text(video_info, './/title')
 154         if not title:
 155             title = self._og_search_title(webpage, default=None)
 156         if not title:
 157             title = self._html_search_regex(
 158                 r'<span[^>]+class="videoHeaderTitle"[^>]*>([^<]+)</span>',
 159                 webpage, 'video title')
 160
 161         watch_api_data_string = self._html_search_regex(
 162             r'<div[^>]+id="watchAPIDataContainer"[^>]+>([^<]+)</div>',
 163             webpage, 'watch api data', default=None)
 164         watch_api_data = self._parse_json(watch_api_data_string, video_id) if watch_api_data_string else {}
 165         video_detail = watch_api_data.get('videoDetail', {})
 166
 167         extension = xpath_text(video_info, './/movie_type')
 168         if not extension:
 169             extension = determine_ext(video_real_url)
 170
 171         thumbnail = (
 172             xpath_text(video_info, './/thumbnail_url') or
 173             self._html_search_meta('image', webpage, 'thumbnail', default=None) or
 174             video_detail.get('thumbnail'))
 175
 176         description = xpath_text(video_info, './/description')
 177
 178         timestamp = parse_iso8601(xpath_text(video_info, './/first_retrieve'))
 179         if not timestamp:
 180             match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
 181             if match:
 182                 timestamp = parse_iso8601(match.replace('+', ':00+'))
 183         if not timestamp and video_detail.get('postedAt'):
 184             timestamp = parse_iso8601(
 185                 video_detail['postedAt'].replace('/', '-'),
 186                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 187
 188         view_count = int_or_none(xpath_text(video_info, './/view_counter'))
 189         if not view_count:
 190             match = self._html_search_regex(
 191                 r'>Views: <strong[^>]*>([^<]+)</strong>',
 192                 webpage, 'view count', default=None)
 193             if match:
 194                 view_count = int_or_none(match.replace(',', ''))
 195         view_count = view_count or video_detail.get('viewCount')
 196
 197         comment_count = int_or_none(xpath_text(video_info, './/comment_num'))
 198         if not comment_count:
 199             match = self._html_search_regex(
 200                 r'>Comments: <strong[^>]*>([^<]+)</strong>',
 201                 webpage, 'comment count', default=None)
 202             if match:
 203                 comment_count = int_or_none(match.replace(',', ''))
 204         comment_count = comment_count or video_detail.get('commentCount')
 205
 206         duration = (parse_duration(
 207             xpath_text(video_info, './/length') or
 208             self._html_search_meta(
 209                 'video:duration', webpage, 'video duration', default=None)) or
 210             video_detail.get('length'))
 211
 212         webpage_url = xpath_text(video_info, './/watch_url') or url
 213
 214         if video_info.find('.//ch_id') is not None:
 215             uploader_id = video_info.find('.//ch_id').text
 216             uploader = video_info.find('.//ch_name').text
 217         elif video_info.find('.//user_id') is not None:
 218             uploader_id = video_info.find('.//user_id').text
 219             uploader = video_info.find('.//user_nickname').text
 220         else:
 221             uploader_id = uploader = None
 222
 223         return {
 224             'id': video_id,
 225             'url': video_real_url,
 226             'title': title,
 227             'ext': extension,
 228             'format_id': 'economy' if video_real_url.endswith('low') else 'normal',
 229             'thumbnail': thumbnail,
 230             'description': description,
 231             'uploader': uploader,
 232             'timestamp': timestamp,
 233             'uploader_id': uploader_id,
 234             'view_count': view_count,
 235             'comment_count': comment_count,
 236             'duration': duration,
 237             'webpage_url': webpage_url,
 238         }
 239
 240
 241 class NiconicoPlaylistIE(InfoExtractor):
 242     _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/mylist/(?P<id>\d+)'
 243
 244     _TEST = {
 245         'url': 'http://www.nicovideo.jp/mylist/27411728',
 246         'info_dict': {
 247             'id': '27411728',
 248             'title': 'AKB48のオールナイトニッポン',
 249         },
 250         'playlist_mincount': 225,
 251     }
 252
 253     def _real_extract(self, url):
 254         list_id = self._match_id(url)
 255         webpage = self._download_webpage(url, list_id)
 256
 257         entries_json = self._search_regex(r'Mylist\.preload\(\d+, (\[.*\])\);',
 258                                           webpage, 'entries')
 259         entries = json.loads(entries_json)
 260         entries = [{
 261             '_type': 'url',
 262             'ie_key': NiconicoIE.ie_key(),
 263             'url': ('http://www.nicovideo.jp/watch/%s' %
 264                     entry['item_data']['video_id']),
 265         } for entry in entries]
 266
 267         return {
 268             '_type': 'playlist',
 269             'title': self._search_regex(r'\s+name: "(.*?)"', webpage, 'title'),
 270             'id': list_id,
 271             'entries': entries,
 272         }