Raphaël G. Git Repositories - youtubedl/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         '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         # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
 104         # chokes on unicode
 105         login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in login_form_strs.items())
 106         login_data = compat_urllib_parse.urlencode(login_form).encode('utf-8')
 107         request = compat_urllib_request.Request(
 108             'https://secure.nicovideo.jp/secure/login', login_data)
 109         login_results = self._download_webpage(
 110             request, None, note='Logging in', errnote='Unable to log in')
 111         if re.search(r'(?i)<h1 class="mb8p4">Log in error</h1>', login_results) is not None:
 112             self._downloader.report_warning('unable to log in: bad username or password')
 113             return False
 114         # Successful login
 115         self._AUTHENTICATED = True
 116         return True
 117
 118     def _real_extract(self, url):
 119         video_id = self._match_id(url)
 120
 121         # Get video webpage. We are not actually interested in it for normal
 122         # cases, but need the cookies in order to be able to download the
 123         # info webpage
 124         webpage, handle = self._download_webpage_handle(
 125             'http://www.nicovideo.jp/watch/' + video_id, video_id)
 126         if video_id.startswith('so'):
 127             video_id = self._match_id(handle.geturl())
 128
 129         video_info = self._download_xml(
 130             'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id, video_id,
 131             note='Downloading video info page')
 132
 133         if self._AUTHENTICATED:
 134             # Get flv info
 135             flv_info_webpage = self._download_webpage(
 136                 'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
 137                 video_id, 'Downloading flv info')
 138         else:
 139             # Get external player info
 140             ext_player_info = self._download_webpage(
 141                 'http://ext.nicovideo.jp/thumb_watch/' + video_id, video_id)
 142             thumb_play_key = self._search_regex(
 143                 r'\'thumbPlayKey\'\s*:\s*\'(.*?)\'', ext_player_info, 'thumbPlayKey')
 144
 145             # Get flv info
 146             flv_info_data = compat_urllib_parse.urlencode({
 147                 'k': thumb_play_key,
 148                 'v': video_id
 149             })
 150             flv_info_request = compat_urllib_request.Request(
 151                 'http://ext.nicovideo.jp/thumb_watch', flv_info_data,
 152                 {'Content-Type': 'application/x-www-form-urlencoded'})
 153             flv_info_webpage = self._download_webpage(
 154                 flv_info_request, video_id,
 155                 note='Downloading flv info', errnote='Unable to download flv info')
 156
 157         flv_info = compat_urlparse.parse_qs(flv_info_webpage)
 158         if 'url' not in flv_info:
 159             if 'deleted' in flv_info:
 160                 raise ExtractorError('The video has been deleted.',
 161                                      expected=True)
 162             else:
 163                 raise ExtractorError('Unable to find video URL')
 164
 165         video_real_url = flv_info['url'][0]
 166
 167         # Start extracting information
 168         title = xpath_text(video_info, './/title')
 169         if not title:
 170             title = self._og_search_title(webpage, default=None)
 171         if not title:
 172             title = self._html_search_regex(
 173                 r'<span[^>]+class="videoHeaderTitle"[^>]*>([^<]+)</span>',
 174                 webpage, 'video title')
 175
 176         watch_api_data_string = self._html_search_regex(
 177             r'<div[^>]+id="watchAPIDataContainer"[^>]+>([^<]+)</div>',
 178             webpage, 'watch api data', default=None)
 179         watch_api_data = self._parse_json(watch_api_data_string, video_id) if watch_api_data_string else {}
 180         video_detail = watch_api_data.get('videoDetail', {})
 181
 182         extension = xpath_text(video_info, './/movie_type')
 183         if not extension:
 184             extension = determine_ext(video_real_url)
 185         video_format = extension.upper()
 186
 187         thumbnail = (
 188             xpath_text(video_info, './/thumbnail_url') or
 189             self._html_search_meta('image', webpage, 'thumbnail', default=None) or
 190             video_detail.get('thumbnail'))
 191
 192         description = xpath_text(video_info, './/description')
 193
 194         timestamp = parse_iso8601(xpath_text(video_info, './/first_retrieve'))
 195         if not timestamp:
 196             match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
 197             if match:
 198                 timestamp = parse_iso8601(match.replace('+', ':00+'))
 199         if not timestamp and video_detail.get('postedAt'):
 200             timestamp = parse_iso8601(
 201                 video_detail['postedAt'].replace('/', '-'),
 202                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 203
 204         view_count = int_or_none(xpath_text(video_info, './/view_counter'))
 205         if not view_count:
 206             match = self._html_search_regex(
 207                 r'>Views: <strong[^>]*>([^<]+)</strong>',
 208                 webpage, 'view count', default=None)
 209             if match:
 210                 view_count = int_or_none(match.replace(',', ''))
 211         view_count = view_count or video_detail.get('viewCount')
 212
 213         comment_count = int_or_none(xpath_text(video_info, './/comment_num'))
 214         if not comment_count:
 215             match = self._html_search_regex(
 216                 r'>Comments: <strong[^>]*>([^<]+)</strong>',
 217                 webpage, 'comment count', default=None)
 218             if match:
 219                 comment_count = int_or_none(match.replace(',', ''))
 220         comment_count = comment_count or video_detail.get('commentCount')
 221
 222         duration = (parse_duration(
 223             xpath_text(video_info, './/length') or
 224             self._html_search_meta(
 225                 'video:duration', webpage, 'video duration', default=None)) or
 226             video_detail.get('length'))
 227
 228         webpage_url = xpath_text(video_info, './/watch_url') or url
 229
 230         if video_info.find('.//ch_id') is not None:
 231             uploader_id = video_info.find('.//ch_id').text
 232             uploader = video_info.find('.//ch_name').text
 233         elif video_info.find('.//user_id') is not None:
 234             uploader_id = video_info.find('.//user_id').text
 235             uploader = video_info.find('.//user_nickname').text
 236         else:
 237             uploader_id = uploader = None
 238
 239         return {
 240             'id': video_id,
 241             'url': video_real_url,
 242             'title': title,
 243             'ext': extension,
 244             'format': video_format,
 245             'thumbnail': thumbnail,
 246             'description': description,
 247             'uploader': uploader,
 248             'timestamp': timestamp,
 249             'uploader_id': uploader_id,
 250             'view_count': view_count,
 251             'comment_count': comment_count,
 252             'duration': duration,
 253             'webpage_url': webpage_url,
 254         }
 255
 256
 257 class NiconicoPlaylistIE(InfoExtractor):
 258     _VALID_URL = r'https?://www\.nicovideo\.jp/mylist/(?P<id>\d+)'
 259
 260     _TEST = {
 261         'url': 'http://www.nicovideo.jp/mylist/27411728',
 262         'info_dict': {
 263             'id': '27411728',
 264             'title': 'AKB48のオールナイトニッポン',
 265         },
 266         'playlist_mincount': 225,
 267     }
 268
 269     def _real_extract(self, url):
 270         list_id = self._match_id(url)
 271         webpage = self._download_webpage(url, list_id)
 272
 273         entries_json = self._search_regex(r'Mylist\.preload\(\d+, (\[.*\])\);',
 274                                           webpage, 'entries')
 275         entries = json.loads(entries_json)
 276         entries = [{
 277             '_type': 'url',
 278             'ie_key': NiconicoIE.ie_key(),
 279             'url': ('http://www.nicovideo.jp/watch/%s' %
 280                     entry['item_data']['video_id']),
 281         } for entry in entries]
 282
 283         return {
 284             '_type': 'playlist',
 285             'title': self._search_regex(r'\s+name: "(.*?)"', webpage, 'title'),
 286             'id': list_id,
 287             'entries': entries,
 288         }