Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/nexx.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import hashlib
   5 import random
   6 import re
   7 import time
   8
   9 from .common import InfoExtractor
  10 from ..compat import compat_str
  11 from ..utils import (
  12     ExtractorError,
  13     int_or_none,
  14     parse_duration,
  15     try_get,
  16     urlencode_postdata,
  17 )
  18
  19
  20 class NexxIE(InfoExtractor):
  21     _VALID_URL = r'''(?x)
  22                         (?:
  23                             https?://api\.nexx(?:\.cloud|cdn\.com)/v3/(?P<domain_id>\d+)/videos/byid/|
  24                             nexx:(?P<domain_id_s>\d+):
  25                         )
  26                         (?P<id>\d+)
  27                     '''
  28     _TESTS = [{
  29         # movie
  30         'url': 'https://api.nexx.cloud/v3/748/videos/byid/128907',
  31         'md5': '828cea195be04e66057b846288295ba1',
  32         'info_dict': {
  33             'id': '128907',
  34             'ext': 'mp4',
  35             'title': 'Stiftung Warentest',
  36             'alt_title': 'Wie ein Test abläuft',
  37             'description': 'md5:d1ddb1ef63de721132abd38639cc2fd2',
  38             'release_year': 2013,
  39             'creator': 'SPIEGEL TV',
  40             'thumbnail': r're:^https?://.*\.jpg$',
  41             'duration': 2509,
  42             'timestamp': 1384264416,
  43             'upload_date': '20131112',
  44         },
  45     }, {
  46         # episode
  47         'url': 'https://api.nexx.cloud/v3/741/videos/byid/247858',
  48         'info_dict': {
  49             'id': '247858',
  50             'ext': 'mp4',
  51             'title': 'Return of the Golden Child (OV)',
  52             'description': 'md5:5d969537509a92b733de21bae249dc63',
  53             'release_year': 2017,
  54             'thumbnail': r're:^https?://.*\.jpg$',
  55             'duration': 1397,
  56             'timestamp': 1495033267,
  57             'upload_date': '20170517',
  58             'episode_number': 2,
  59             'season_number': 2,
  60         },
  61         'params': {
  62             'skip_download': True,
  63         },
  64     }, {
  65         'url': 'https://api.nexxcdn.com/v3/748/videos/byid/128907',
  66         'only_matching': True,
  67     }, {
  68         'url': 'nexx:748:128907',
  69         'only_matching': True,
  70     }]
  71
  72     @staticmethod
  73     def _extract_domain_id(webpage):
  74         mobj = re.search(
  75             r'<script\b[^>]+\bsrc=["\'](?:https?:)?//require\.nexx(?:\.cloud|cdn\.com)/(?P<id>\d+)',
  76             webpage)
  77         return mobj.group('id') if mobj else None
  78
  79     @staticmethod
  80     def _extract_urls(webpage):
  81         # Reference:
  82         # 1. https://nx-s.akamaized.net/files/201510/44.pdf
  83
  84         entries = []
  85
  86         # JavaScript Integration
  87         domain_id = NexxIE._extract_domain_id(webpage)
  88         if domain_id:
  89             for video_id in re.findall(
  90                     r'(?is)onPLAYReady.+?_play\.init\s*\(.+?\s*,\s*["\']?(\d+)',
  91                     webpage):
  92                 entries.append(
  93                     'https://api.nexx.cloud/v3/%s/videos/byid/%s'
  94                     % (domain_id, video_id))
  95
  96         # TODO: support more embed formats
  97
  98         return entries
  99
 100     @staticmethod
 101     def _extract_url(webpage):
 102         return NexxIE._extract_urls(webpage)[0]
 103
 104     def _handle_error(self, response):
 105         status = int_or_none(try_get(
 106             response, lambda x: x['metadata']['status']) or 200)
 107         if 200 <= status < 300:
 108             return
 109         raise ExtractorError(
 110             '%s said: %s' % (self.IE_NAME, response['metadata']['errorhint']),
 111             expected=True)
 112
 113     def _call_api(self, domain_id, path, video_id, data=None, headers={}):
 114         headers['Content-Type'] = 'application/x-www-form-urlencoded; charset=UTF-8'
 115         result = self._download_json(
 116             'https://api.nexx.cloud/v3/%s/%s' % (domain_id, path), video_id,
 117             'Downloading %s JSON' % path, data=urlencode_postdata(data),
 118             headers=headers)
 119         self._handle_error(result)
 120         return result['result']
 121
 122     def _real_extract(self, url):
 123         mobj = re.match(self._VALID_URL, url)
 124         domain_id = mobj.group('domain_id') or mobj.group('domain_id_s')
 125         video_id = mobj.group('id')
 126
 127         # Reverse engineered from JS code (see getDeviceID function)
 128         device_id = '%d:%d:%d%d' % (
 129             random.randint(1, 4), int(time.time()),
 130             random.randint(1e4, 99999), random.randint(1, 9))
 131
 132         result = self._call_api(domain_id, 'session/init', video_id, data={
 133             'nxp_devh': device_id,
 134             'nxp_userh': '',
 135             'precid': '0',
 136             'playlicense': '0',
 137             'screenx': '1920',
 138             'screeny': '1080',
 139             'playerversion': '6.0.00',
 140             'gateway': 'html5',
 141             'adGateway': '',
 142             'explicitlanguage': 'en-US',
 143             'addTextTemplates': '1',
 144             'addDomainData': '1',
 145             'addAdModel': '1',
 146         }, headers={
 147             'X-Request-Enable-Auth-Fallback': '1',
 148         })
 149
 150         cid = result['general']['cid']
 151
 152         # As described in [1] X-Request-Token generation algorithm is
 153         # as follows:
 154         #   md5( operation + domain_id + domain_secret )
 155         # where domain_secret is a static value that will be given by nexx.tv
 156         # as per [1]. Here is how this "secret" is generated (reversed
 157         # from _play.api.init function, search for clienttoken). So it's
 158         # actually not static and not that much of a secret.
 159         # 1. https://nexxtvstorage.blob.core.windows.net/files/201610/27.pdf
 160         secret = result['device']['clienttoken'][int(device_id[0]):]
 161         secret = secret[0:len(secret) - int(device_id[-1])]
 162
 163         op = 'byid'
 164
 165         # Reversed from JS code for _play.api.call function (search for
 166         # X-Request-Token)
 167         request_token = hashlib.md5(
 168             ''.join((op, domain_id, secret)).encode('utf-8')).hexdigest()
 169
 170         video = self._call_api(
 171             domain_id, 'videos/%s/%s' % (op, video_id), video_id, data={
 172                 'additionalfields': 'language,channel,actors,studio,licenseby,slug,subtitle,teaser,description',
 173                 'addInteractionOptions': '1',
 174                 'addStatusDetails': '1',
 175                 'addStreamDetails': '1',
 176                 'addCaptions': '1',
 177                 'addScenes': '1',
 178                 'addHotSpots': '1',
 179                 'addBumpers': '1',
 180                 'captionFormat': 'data',
 181             }, headers={
 182                 'X-Request-CID': cid,
 183                 'X-Request-Token': request_token,
 184             })
 185
 186         general = video['general']
 187         title = general['title']
 188
 189         stream_data = video['streamdata']
 190         language = general.get('language_raw') or ''
 191
 192         # TODO: reverse more cdns
 193
 194         cdn = stream_data['cdnType']
 195         assert cdn == 'azure'
 196
 197         azure_locator = stream_data['azureLocator']
 198
 199         AZURE_URL = 'http://nx%s%02d.akamaized.net/'
 200
 201         def get_cdn_shield_base(shield_type='', prefix='-p'):
 202             for secure in ('', 's'):
 203                 cdn_shield = stream_data.get('cdnShield%sHTTP%s' % (shield_type, secure.upper()))
 204                 if cdn_shield:
 205                     return 'http%s://%s' % (secure, cdn_shield)
 206             else:
 207                 return AZURE_URL % (prefix, int(stream_data['azureAccount'].replace('nexxplayplus', '')))
 208
 209         azure_stream_base = get_cdn_shield_base()
 210         is_ml = ',' in language
 211         azure_manifest_url = '%s%s/%s_src%s.ism/Manifest' % (
 212             azure_stream_base, azure_locator, video_id, ('_manifest' if is_ml else '')) + '%s'
 213
 214         protection_token = try_get(
 215             video, lambda x: x['protectiondata']['token'], compat_str)
 216         if protection_token:
 217             azure_manifest_url += '?hdnts=%s' % protection_token
 218
 219         formats = self._extract_m3u8_formats(
 220             azure_manifest_url % '(format=m3u8-aapl)',
 221             video_id, 'mp4', 'm3u8_native',
 222             m3u8_id='%s-hls' % cdn, fatal=False)
 223         formats.extend(self._extract_mpd_formats(
 224             azure_manifest_url % '(format=mpd-time-csf)',
 225             video_id, mpd_id='%s-dash' % cdn, fatal=False))
 226         formats.extend(self._extract_ism_formats(
 227             azure_manifest_url % '', video_id, ism_id='%s-mss' % cdn, fatal=False))
 228
 229         azure_progressive_base = get_cdn_shield_base('Prog', '-d')
 230         azure_file_distribution = stream_data.get('azureFileDistribution')
 231         if azure_file_distribution:
 232             fds = azure_file_distribution.split(',')
 233             if fds:
 234                 for fd in fds:
 235                     ss = fd.split(':')
 236                     if len(ss) == 2:
 237                         tbr = int_or_none(ss[0])
 238                         if tbr:
 239                             f = {
 240                                 'url': '%s%s/%s_src_%s_%d.mp4' % (
 241                                     azure_progressive_base, azure_locator, video_id, ss[1], tbr),
 242                                 'format_id': '%s-http-%d' % (cdn, tbr),
 243                                 'tbr': tbr,
 244                             }
 245                             width_height = ss[1].split('x')
 246                             if len(width_height) == 2:
 247                                 f.update({
 248                                     'width': int_or_none(width_height[0]),
 249                                     'height': int_or_none(width_height[1]),
 250                                 })
 251                             formats.append(f)
 252
 253         self._sort_formats(formats)
 254
 255         return {
 256             'id': video_id,
 257             'title': title,
 258             'alt_title': general.get('subtitle'),
 259             'description': general.get('description'),
 260             'release_year': int_or_none(general.get('year')),
 261             'creator': general.get('studio') or general.get('studio_adref'),
 262             'thumbnail': try_get(
 263                 video, lambda x: x['imagedata']['thumb'], compat_str),
 264             'duration': parse_duration(general.get('runtime')),
 265             'timestamp': int_or_none(general.get('uploaded')),
 266             'episode_number': int_or_none(try_get(
 267                 video, lambda x: x['episodedata']['episode'])),
 268             'season_number': int_or_none(try_get(
 269                 video, lambda x: x['episodedata']['season'])),
 270             'formats': formats,
 271         }
 272
 273
 274 class NexxEmbedIE(InfoExtractor):
 275     _VALID_URL = r'https?://embed\.nexx(?:\.cloud|cdn\.com)/\d+/(?P<id>[^/?#&]+)'
 276     _TEST = {
 277         'url': 'http://embed.nexx.cloud/748/KC1614647Z27Y7T?autoplay=1',
 278         'md5': '16746bfc28c42049492385c989b26c4a',
 279         'info_dict': {
 280             'id': '161464',
 281             'ext': 'mp4',
 282             'title': 'Nervenkitzel Achterbahn',
 283             'alt_title': 'Karussellbauer in Deutschland',
 284             'description': 'md5:ffe7b1cc59a01f585e0569949aef73cc',
 285             'release_year': 2005,
 286             'creator': 'SPIEGEL TV',
 287             'thumbnail': r're:^https?://.*\.jpg$',
 288             'duration': 2761,
 289             'timestamp': 1394021479,
 290             'upload_date': '20140305',
 291         },
 292         'params': {
 293             'format': 'bestvideo',
 294             'skip_download': True,
 295         },
 296     }
 297
 298     @staticmethod
 299     def _extract_urls(webpage):
 300         # Reference:
 301         # 1. https://nx-s.akamaized.net/files/201510/44.pdf
 302
 303         # iFrame Embed Integration
 304         return [mobj.group('url') for mobj in re.finditer(
 305             r'<iframe[^>]+\bsrc=(["\'])(?P<url>(?:https?:)?//embed\.nexx(?:\.cloud|cdn\.com)/\d+/(?:(?!\1).)+)\1',
 306             webpage)]
 307
 308     def _real_extract(self, url):
 309         embed_id = self._match_id(url)
 310
 311         webpage = self._download_webpage(url, embed_id)
 312
 313         return self.url_result(NexxIE._extract_url(webpage), ie=NexxIE.ie_key())