_ Git - youtube-dl/blob - youtube_dl/extractor/funimation.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     clean_html,
   9     determine_ext,
  10     encode_dict,
  11     sanitized_Request,
  12     ExtractorError,
  13     urlencode_postdata
  14 )
  15
  16
  17 class FunimationIE(InfoExtractor):
  18     _VALID_URL = r'https?://(?:www\.)?funimation\.com/shows/[^/]+/videos/official/(?P<id>[^/?#&]+)'
  19
  20     _TESTS = [{
  21         'url': 'http://www.funimation.com/shows/air/videos/official/breeze',
  22         'info_dict': {
  23             'id': '658',
  24             'display_id': 'breeze',
  25             'ext': 'mp4',
  26             'title': 'Air - 1 - Breeze',
  27             'description': 'md5:1769f43cd5fc130ace8fd87232207892',
  28             'thumbnail': 're:https?://.*\.jpg',
  29         },
  30     }, {
  31         'url': 'http://www.funimation.com/shows/hacksign/videos/official/role-play',
  32         'info_dict': {
  33             'id': '31128',
  34             'display_id': 'role-play',
  35             'ext': 'mp4',
  36             'title': '.hack//SIGN - 1 - Role Play',
  37             'description': 'md5:b602bdc15eef4c9bbb201bb6e6a4a2dd',
  38             'thumbnail': 're:https?://.*\.jpg',
  39         },
  40     }]
  41
  42     def _login(self):
  43         (username, password) = self._get_login_info()
  44         if username is None:
  45             return
  46         data = urlencode_postdata(encode_dict({
  47             'email_field': username,
  48             'password_field': password,
  49         }))
  50         login_request = sanitized_Request('http://www.funimation.com/login', data, headers={
  51             'User-Agent': 'Mozilla/5.0 (Windows NT 5.2; WOW64; rv:42.0) Gecko/20100101 Firefox/42.0',
  52             'Content-Type': 'application/x-www-form-urlencoded'
  53         })
  54         login = self._download_webpage(
  55             login_request, None, 'Logging in as %s' % username)
  56         if re.search(r'<meta property="og:url" content="http://www.funimation.com/login"/>', login) is not None:
  57             raise ExtractorError('Unable to login, wrong username or password.', expected=True)
  58
  59     def _real_initialize(self):
  60         self._login()
  61
  62     def _real_extract(self, url):
  63         display_id = self._match_id(url)
  64
  65         errors = []
  66         formats = []
  67
  68         ERRORS_MAP = {
  69             'ERROR_MATURE_CONTENT_LOGGED_IN': 'matureContentLoggedIn',
  70             'ERROR_MATURE_CONTENT_LOGGED_OUT': 'matureContentLoggedOut',
  71             'ERROR_SUBSCRIPTION_LOGGED_OUT': 'subscriptionLoggedOut',
  72             'ERROR_VIDEO_EXPIRED': 'videoExpired',
  73             'ERROR_TERRITORY_UNAVAILABLE': 'territoryUnavailable',
  74             'SVODBASIC_SUBSCRIPTION_IN_PLAYER': 'basicSubscription',
  75             'SVODNON_SUBSCRIPTION_IN_PLAYER': 'nonSubscription',
  76             'ERROR_PLAYER_NOT_RESPONDING': 'playerNotResponding',
  77             'ERROR_UNABLE_TO_CONNECT_TO_CDN': 'unableToConnectToCDN',
  78             'ERROR_STREAM_NOT_FOUND': 'streamNotFound',
  79         }
  80
  81         USER_AGENTS = (
  82             # PC UA is served with m3u8 that provides some bonus lower quality formats
  83             ('pc', 'Mozilla/5.0 (Windows NT 5.2; WOW64; rv:42.0) Gecko/20100101 Firefox/42.0'),
  84             # Mobile UA allows to extract direct links and also does not fail when
  85             # PC UA fails with hulu error (e.g.
  86             # http://www.funimation.com/shows/hacksign/videos/official/role-play)
  87             ('mobile', 'Mozilla/5.0 (Linux; Android 4.4.2; Nexus 4 Build/KOT49H) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/34.0.1847.114 Mobile Safari/537.36'),
  88         )
  89
  90         for kind, user_agent in USER_AGENTS:
  91             request = sanitized_Request(url)
  92             request.add_header('User-Agent', user_agent)
  93             webpage = self._download_webpage(
  94                 request, display_id, 'Downloading %s webpage' % kind)
  95
  96             items = self._parse_json(
  97                 self._search_regex(
  98                     r'var\s+playersData\s*=\s*(\[.+?\]);\n',
  99                     webpage, 'players data'),
 100                 display_id)[0]['playlist'][0]['items']
 101
 102             item = next(item for item in items if item.get('itemAK') == display_id)
 103
 104             error_messages = {}
 105             video_error_messages = self._search_regex(
 106                 r'var\s+videoErrorMessages\s*=\s*({.+?});\n',
 107                 webpage, 'error messages', default=None)
 108             if video_error_messages:
 109                 error_messages_json = self._parse_json(video_error_messages, display_id, fatal=False)
 110                 if error_messages_json:
 111                     for _, error in error_messages_json.items():
 112                         type_ = error.get('type')
 113                         description = error.get('description')
 114                         content = error.get('content')
 115                         if type_ == 'text' and description and content:
 116                             error_message = ERRORS_MAP.get(description)
 117                             if error_message:
 118                                 error_messages[error_message] = content
 119
 120             for video in item.get('videoSet', []):
 121                 auth_token = video.get('authToken')
 122                 if not auth_token:
 123                     continue
 124                 funimation_id = video.get('FUNImationID') or video.get('videoId')
 125                 preference = 1 if video.get('languageMode') == 'dub' else 0
 126                 if not auth_token.startswith('?'):
 127                     auth_token = '?%s' % auth_token
 128                 for quality in ('sd', 'hd', 'hd1080'):
 129                     format_url = video.get('%sUrl' % quality)
 130                     if not format_url:
 131                         continue
 132                     if not format_url.startswith(('http', '//')):
 133                         errors.append(format_url)
 134                         continue
 135                     if determine_ext(format_url) == 'm3u8':
 136                         m3u8_formats = self._extract_m3u8_formats(
 137                             format_url + auth_token, display_id, 'mp4', entry_protocol='m3u8_native',
 138                             preference=preference, m3u8_id=funimation_id or 'hls', fatal=False)
 139                         if m3u8_formats:
 140                             formats.extend(m3u8_formats)
 141                     else:
 142                         f = {
 143                             'url': format_url + auth_token,
 144                             'format_id': funimation_id,
 145                             'preference': preference,
 146                         }
 147                         mobj = re.search(r'(?P<height>\d+)-(?P<tbr>\d+)[Kk]', format_url)
 148                         if mobj:
 149                             f.update({
 150                                 'height': int(mobj.group('height')),
 151                                 'tbr': int(mobj.group('tbr')),
 152                             })
 153                         formats.append(f)
 154
 155         if not formats and errors:
 156             raise ExtractorError(
 157                 '%s returned error: %s'
 158                 % (self.IE_NAME, clean_html(error_messages.get(errors[0], errors[0]))),
 159                 expected=True)
 160
 161         self._sort_formats(formats)
 162
 163         title = item['title']
 164         artist = item.get('artist')
 165         if artist:
 166             title = '%s - %s' % (artist, title)
 167         description = self._og_search_description(webpage) or item.get('description')
 168         thumbnail = self._og_search_thumbnail(webpage) or item.get('posterUrl')
 169         video_id = item.get('itemId') or display_id
 170
 171         return {
 172             'id': video_id,
 173             'display_id': display_id,
 174             'title': title,
 175             'description': description,
 176             'thumbnail': thumbnail,
 177             'formats': formats,
 178         }