_ 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,
  11     compat_urllib_request,
  12     compat_urlparse,
  13 )
  14 from ..utils import (
  15     ExtractorError,
  16     int_or_none,
  17     parse_duration,
  18     parse_iso8601,
  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
  72     _VALID_URL = r'https?://(?:www\.|secure\.)?nicovideo\.jp/watch/(?P<id>(?:[a-z]{2})?[0-9]+)'
  73     _NETRC_MACHINE = 'niconico'
  74     # Determine whether the downloader used authentication to download video
  75     _AUTHENTICATED = False
  76
  77     def _real_initialize(self):
  78         self._login()
  79
  80     def _login(self):
  81         (username, password) = self._get_login_info()
  82         # No authentication to be performed
  83         if not username:
  84             return True
  85
  86         # Log in
  87         login_form_strs = {
  88             'mail': username,
  89             'password': password,
  90         }
  91         # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
  92         # chokes on unicode
  93         login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in login_form_strs.items())
  94         login_data = compat_urllib_parse.urlencode(login_form).encode('utf-8')
  95         request = compat_urllib_request.Request(
  96             'https://secure.nicovideo.jp/secure/login', login_data)
  97         login_results = self._download_webpage(
  98             request, None, note='Logging in', errnote='Unable to log in')
  99         if re.search(r'(?i)<h1 class="mb8p4">Log in error</h1>', login_results) is not None:
 100             self._downloader.report_warning('unable to log in: bad username or password')
 101             return False
 102         # Successful login
 103         self._AUTHENTICATED = True
 104         return True
 105
 106     def _real_extract(self, url):
 107         video_id = self._match_id(url)
 108
 109         # Get video webpage. We are not actually interested in it for normal
 110         # cases, but need the cookies in order to be able to download the
 111         # info webpage
 112         webpage = self._download_webpage('http://www.nicovideo.jp/watch/' + video_id, video_id)
 113
 114         video_info = self._download_xml(
 115             'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id, video_id,
 116             note='Downloading video info page')
 117
 118         if self._AUTHENTICATED:
 119             # Get flv info
 120             flv_info_webpage = self._download_webpage(
 121                 'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
 122                 video_id, 'Downloading flv info')
 123         else:
 124             # Get external player info
 125             ext_player_info = self._download_webpage(
 126                 'http://ext.nicovideo.jp/thumb_watch/' + video_id, video_id)
 127             thumb_play_key = self._search_regex(
 128                 r'\'thumbPlayKey\'\s*:\s*\'(.*?)\'', ext_player_info, 'thumbPlayKey')
 129
 130             # Get flv info
 131             flv_info_data = compat_urllib_parse.urlencode({
 132                 'k': thumb_play_key,
 133                 'v': video_id
 134             })
 135             flv_info_request = compat_urllib_request.Request(
 136                 'http://ext.nicovideo.jp/thumb_watch', flv_info_data,
 137                 {'Content-Type': 'application/x-www-form-urlencoded'})
 138             flv_info_webpage = self._download_webpage(
 139                 flv_info_request, video_id,
 140                 note='Downloading flv info', errnote='Unable to download 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             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         video_format = extension.upper()
 171
 172         thumbnail = (
 173             xpath_text(video_info, './/thumbnail_url') or
 174             self._html_search_meta('image', webpage, 'thumbnail', default=None) or
 175             video_detail.get('thumbnail'))
 176
 177         description = xpath_text(video_info, './/description')
 178
 179         timestamp = parse_iso8601(xpath_text(video_info, './/first_retrieve'))
 180         if not timestamp:
 181             match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
 182             if match:
 183                 timestamp = parse_iso8601(match.replace('+', ':00+'))
 184         if not timestamp and video_detail.get('postedAt'):
 185             timestamp = parse_iso8601(
 186                 video_detail['postedAt'].replace('/', '-'),
 187                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 188
 189         view_count = int_or_none(xpath_text(video_info, './/view_counter'))
 190         if not view_count:
 191             match = self._html_search_regex(
 192                 r'>Views: <strong[^>]*>([^<]+)</strong>',
 193                 webpage, 'view count', default=None)
 194             if match:
 195                 view_count = int_or_none(match.replace(',', ''))
 196         view_count = view_count or video_detail.get('viewCount')
 197
 198         comment_count = int_or_none(xpath_text(video_info, './/comment_num'))
 199         if not comment_count:
 200             match = self._html_search_regex(
 201                 r'>Comments: <strong[^>]*>([^<]+)</strong>',
 202                 webpage, 'comment count', default=None)
 203             if match:
 204                 comment_count = int_or_none(match.replace(',', ''))
 205         comment_count = comment_count or video_detail.get('commentCount')
 206
 207         duration = (parse_duration(
 208             xpath_text(video_info, './/length') or
 209             self._html_search_meta(
 210                 'video:duration', webpage, 'video duration', default=None)) or
 211             video_detail.get('length'))
 212
 213         webpage_url = xpath_text(video_info, './/watch_url') or url
 214
 215         if video_info.find('.//ch_id') is not None:
 216             uploader_id = video_info.find('.//ch_id').text
 217             uploader = video_info.find('.//ch_name').text
 218         elif video_info.find('.//user_id') is not None:
 219             uploader_id = video_info.find('.//user_id').text
 220             uploader = video_info.find('.//user_nickname').text
 221         else:
 222             uploader_id = uploader = None
 223
 224         return {
 225             'id': video_id,
 226             'url': video_real_url,
 227             'title': title,
 228             'ext': extension,
 229             'format': video_format,
 230             'thumbnail': thumbnail,
 231             'description': description,
 232             'uploader': uploader,
 233             'timestamp': timestamp,
 234             'uploader_id': uploader_id,
 235             'view_count': view_count,
 236             'comment_count': comment_count,
 237             'duration': duration,
 238             'webpage_url': webpage_url,
 239         }
 240
 241
 242 class NiconicoPlaylistIE(InfoExtractor):
 243     _VALID_URL = r'https?://www\.nicovideo\.jp/mylist/(?P<id>\d+)'
 244
 245     _TEST = {
 246         'url': 'http://www.nicovideo.jp/mylist/27411728',
 247         'info_dict': {
 248             'id': '27411728',
 249             'title': 'AKB48のオールナイトニッポン',
 250         },
 251         'playlist_mincount': 225,
 252     }
 253
 254     def _real_extract(self, url):
 255         list_id = self._match_id(url)
 256         webpage = self._download_webpage(url, list_id)
 257
 258         entries_json = self._search_regex(r'Mylist\.preload\(\d+, (\[.*\])\);',
 259                                           webpage, 'entries')
 260         entries = json.loads(entries_json)
 261         entries = [{
 262             '_type': 'url',
 263             'ie_key': NiconicoIE.ie_key(),
 264             'url': ('http://www.nicovideo.jp/watch/%s' %
 265                     entry['item_data']['video_id']),
 266         } for entry in entries]
 267
 268         return {
 269             '_type': 'playlist',
 270             'title': self._search_regex(r'\s+name: "(.*?)"', webpage, 'title'),
 271             'id': list_id,
 272             'entries': entries,
 273         }