_ Git - youtube-dl/blob - 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_urlencode,
  11     compat_urlparse,
  12 )
  13 from ..utils import (
  14     ExtractorError,
  15     int_or_none,
  16     parse_duration,
  17     parse_iso8601,
  18     sanitized_Request,
  19     xpath_text,
  20     determine_ext,
  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     }, {
  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     }, {
  58         # 'video exists but is marked as "deleted"
  59         # md5 is unstable
  60         'url': 'http://www.nicovideo.jp/watch/sm10000',
  61         'info_dict': {
  62             'id': 'sm10000',
  63             'ext': 'unknown_video',
  64             'description': 'deleted',
  65             'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞',
  66             'upload_date': '20071224',
  67             'timestamp': 1198527840,  # timestamp field has different value if logged in
  68             'duration': 304,
  69         },
  70     }, {
  71         'url': 'http://www.nicovideo.jp/watch/so22543406',
  72         'info_dict': {
  73             'id': '1388129933',
  74             'ext': 'mp4',
  75             'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～',
  76             'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1',
  77             'timestamp': 1388851200,
  78             'upload_date': '20140104',
  79             'uploader': 'アニメロチャンネル',
  80             'uploader_id': '312',
  81         }
  82     }]
  83
  84     _VALID_URL = r'https?://(?:www\.|secure\.)?nicovideo\.jp/watch/(?P<id>(?:[a-z]{2})?[0-9]+)'
  85     _NETRC_MACHINE = 'niconico'
  86     # Determine whether the downloader used authentication to download video
  87     _AUTHENTICATED = False
  88
  89     def _real_initialize(self):
  90         self._login()
  91
  92     def _login(self):
  93         (username, password) = self._get_login_info()
  94         # No authentication to be performed
  95         if not username:
  96             return True
  97
  98         # Log in
  99         login_form_strs = {
 100             'mail': username,
 101             'password': password,
 102         }
 103         login_data = compat_urllib_parse_urlencode(login_form_strs).encode('utf-8')
 104         request = sanitized_Request(
 105             'https://secure.nicovideo.jp/secure/login', login_data)
 106         login_results = self._download_webpage(
 107             request, None, note='Logging in', errnote='Unable to log in')
 108         if re.search(r'(?i)<h1 class="mb8p4">Log in error</h1>', login_results) is not None:
 109             self._downloader.report_warning('unable to log in: bad username or password')
 110             return False
 111         # Successful login
 112         self._AUTHENTICATED = True
 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         if self._AUTHENTICATED:
 131             # Get flv info
 132             flv_info_webpage = self._download_webpage(
 133                 'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
 134                 video_id, 'Downloading flv info')
 135         else:
 136             # Get external player info
 137             ext_player_info = self._download_webpage(
 138                 'http://ext.nicovideo.jp/thumb_watch/' + video_id, video_id)
 139             thumb_play_key = self._search_regex(
 140                 r'\'thumbPlayKey\'\s*:\s*\'(.*?)\'', ext_player_info, 'thumbPlayKey')
 141
 142             # Get flv info
 143             flv_info_data = compat_urllib_parse_urlencode({
 144                 'k': thumb_play_key,
 145                 'v': video_id
 146             })
 147             flv_info_request = sanitized_Request(
 148                 'http://ext.nicovideo.jp/thumb_watch', flv_info_data,
 149                 {'Content-Type': 'application/x-www-form-urlencoded'})
 150             flv_info_webpage = self._download_webpage(
 151                 flv_info_request, video_id,
 152                 note='Downloading flv info', errnote='Unable to download flv info')
 153
 154         flv_info = compat_urlparse.parse_qs(flv_info_webpage)
 155         if 'url' not in flv_info:
 156             if 'deleted' in flv_info:
 157                 raise ExtractorError('The video has been deleted.',
 158                                      expected=True)
 159             else:
 160                 raise ExtractorError('Unable to find video URL')
 161
 162         video_real_url = flv_info['url'][0]
 163
 164         # Start extracting information
 165         title = xpath_text(video_info, './/title')
 166         if not title:
 167             title = self._og_search_title(webpage, default=None)
 168         if not title:
 169             title = self._html_search_regex(
 170                 r'<span[^>]+class="videoHeaderTitle"[^>]*>([^<]+)</span>',
 171                 webpage, 'video title')
 172
 173         watch_api_data_string = self._html_search_regex(
 174             r'<div[^>]+id="watchAPIDataContainer"[^>]+>([^<]+)</div>',
 175             webpage, 'watch api data', default=None)
 176         watch_api_data = self._parse_json(watch_api_data_string, video_id) if watch_api_data_string else {}
 177         video_detail = watch_api_data.get('videoDetail', {})
 178
 179         extension = xpath_text(video_info, './/movie_type')
 180         if not extension:
 181             extension = determine_ext(video_real_url)
 182
 183         thumbnail = (
 184             xpath_text(video_info, './/thumbnail_url') or
 185             self._html_search_meta('image', webpage, 'thumbnail', default=None) or
 186             video_detail.get('thumbnail'))
 187
 188         description = xpath_text(video_info, './/description')
 189
 190         timestamp = parse_iso8601(xpath_text(video_info, './/first_retrieve'))
 191         if not timestamp:
 192             match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
 193             if match:
 194                 timestamp = parse_iso8601(match.replace('+', ':00+'))
 195         if not timestamp and video_detail.get('postedAt'):
 196             timestamp = parse_iso8601(
 197                 video_detail['postedAt'].replace('/', '-'),
 198                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 199
 200         view_count = int_or_none(xpath_text(video_info, './/view_counter'))
 201         if not view_count:
 202             match = self._html_search_regex(
 203                 r'>Views: <strong[^>]*>([^<]+)</strong>',
 204                 webpage, 'view count', default=None)
 205             if match:
 206                 view_count = int_or_none(match.replace(',', ''))
 207         view_count = view_count or video_detail.get('viewCount')
 208
 209         comment_count = int_or_none(xpath_text(video_info, './/comment_num'))
 210         if not comment_count:
 211             match = self._html_search_regex(
 212                 r'>Comments: <strong[^>]*>([^<]+)</strong>',
 213                 webpage, 'comment count', default=None)
 214             if match:
 215                 comment_count = int_or_none(match.replace(',', ''))
 216         comment_count = comment_count or video_detail.get('commentCount')
 217
 218         duration = (parse_duration(
 219             xpath_text(video_info, './/length') or
 220             self._html_search_meta(
 221                 'video:duration', webpage, 'video duration', default=None)) or
 222             video_detail.get('length'))
 223
 224         webpage_url = xpath_text(video_info, './/watch_url') or url
 225
 226         if video_info.find('.//ch_id') is not None:
 227             uploader_id = video_info.find('.//ch_id').text
 228             uploader = video_info.find('.//ch_name').text
 229         elif video_info.find('.//user_id') is not None:
 230             uploader_id = video_info.find('.//user_id').text
 231             uploader = video_info.find('.//user_nickname').text
 232         else:
 233             uploader_id = uploader = None
 234
 235         return {
 236             'id': video_id,
 237             'url': video_real_url,
 238             'title': title,
 239             'ext': extension,
 240             'format_id': 'economy' if video_real_url.endswith('low') else 'normal',
 241             'thumbnail': thumbnail,
 242             'description': description,
 243             'uploader': uploader,
 244             'timestamp': timestamp,
 245             'uploader_id': uploader_id,
 246             'view_count': view_count,
 247             'comment_count': comment_count,
 248             'duration': duration,
 249             'webpage_url': webpage_url,
 250         }
 251
 252
 253 class NiconicoPlaylistIE(InfoExtractor):
 254     _VALID_URL = r'https?://www\.nicovideo\.jp/mylist/(?P<id>\d+)'
 255
 256     _TEST = {
 257         'url': 'http://www.nicovideo.jp/mylist/27411728',
 258         'info_dict': {
 259             'id': '27411728',
 260             'title': 'AKB48のオールナイトニッポン',
 261         },
 262         'playlist_mincount': 225,
 263     }
 264
 265     def _real_extract(self, url):
 266         list_id = self._match_id(url)
 267         webpage = self._download_webpage(url, list_id)
 268
 269         entries_json = self._search_regex(r'Mylist\.preload\(\d+, (\[.*\])\);',
 270                                           webpage, 'entries')
 271         entries = json.loads(entries_json)
 272         entries = [{
 273             '_type': 'url',
 274             'ie_key': NiconicoIE.ie_key(),
 275             'url': ('http://www.nicovideo.jp/watch/%s' %
 276                     entry['item_data']['video_id']),
 277         } for entry in entries]
 278
 279         return {
 280             '_type': 'playlist',
 281             'title': self._search_regex(r'\s+name: "(.*?)"', webpage, 'title'),
 282             'id': list_id,
 283             'entries': entries,
 284         }