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'https?://api\.nexx(?:\.cloud|cdn\.com)/v3/(?P<domain_id>\d+)/videos/byid/(?P<id>\d+)'
  22     _TESTS = [{
  23         # movie
  24         'url': 'https://api.nexx.cloud/v3/748/videos/byid/128907',
  25         'md5': '16746bfc28c42049492385c989b26c4a',
  26         'info_dict': {
  27             'id': '128907',
  28             'ext': 'mp4',
  29             'title': 'Stiftung Warentest',
  30             'alt_title': 'Wie ein Test abläuft',
  31             'description': 'md5:d1ddb1ef63de721132abd38639cc2fd2',
  32             'release_year': 2013,
  33             'creator': 'SPIEGEL TV',
  34             'thumbnail': r're:^https?://.*\.jpg$',
  35             'duration': 2509,
  36             'timestamp': 1384264416,
  37             'upload_date': '20131112',
  38         },
  39         'params': {
  40             'format': 'bestvideo',
  41         },
  42     }, {
  43         # episode
  44         'url': 'https://api.nexx.cloud/v3/741/videos/byid/247858',
  45         'info_dict': {
  46             'id': '247858',
  47             'ext': 'mp4',
  48             'title': 'Return of the Golden Child (OV)',
  49             'description': 'md5:5d969537509a92b733de21bae249dc63',
  50             'release_year': 2017,
  51             'thumbnail': r're:^https?://.*\.jpg$',
  52             'duration': 1397,
  53             'timestamp': 1495033267,
  54             'upload_date': '20170517',
  55             'episode_number': 2,
  56             'season_number': 2,
  57         },
  58         'params': {
  59             'format': 'bestvideo',
  60             'skip_download': True,
  61         },
  62     }, {
  63         'url': 'https://api.nexxcdn.com/v3/748/videos/byid/128907',
  64         'only_matching': True,
  65     }]
  66
  67     @staticmethod
  68     def _extract_urls(webpage):
  69         # Reference:
  70         # 1. https://nx-s.akamaized.net/files/201510/44.pdf
  71
  72         entries = []
  73
  74         # JavaScript Integration
  75         mobj = re.search(
  76             r'<script\b[^>]+\bsrc=["\']https?://require\.nexx(?:\.cloud|cdn\.com)/(?P<id>\d+)',
  77             webpage)
  78         if mobj:
  79             domain_id = mobj.group('id')
  80             for video_id in re.findall(
  81                     r'(?is)onPLAYReady.+?_play\.init\s*\(.+?\s*,\s*["\']?(\d+)',
  82                     webpage):
  83                 entries.append(
  84                     'https://api.nexx.cloud/v3/%s/videos/byid/%s'
  85                     % (domain_id, video_id))
  86
  87         # TODO: support more embed formats
  88
  89         return entries
  90
  91     @staticmethod
  92     def _extract_url(webpage):
  93         return NexxIE._extract_urls(webpage)[0]
  94
  95     def _handle_error(self, response):
  96         status = int_or_none(try_get(
  97             response, lambda x: x['metadata']['status']) or 200)
  98         if 200 <= status < 300:
  99             return
 100         raise ExtractorError(
 101             '%s said: %s' % (self.IE_NAME, response['metadata']['errorhint']),
 102             expected=True)
 103
 104     def _call_api(self, domain_id, path, video_id, data=None, headers={}):
 105         headers['Content-Type'] = 'application/x-www-form-urlencoded; charset=UTF-8'
 106         result = self._download_json(
 107             'https://api.nexx.cloud/v3/%s/%s' % (domain_id, path), video_id,
 108             'Downloading %s JSON' % path, data=urlencode_postdata(data),
 109             headers=headers)
 110         self._handle_error(result)
 111         return result['result']
 112
 113     def _real_extract(self, url):
 114         mobj = re.match(self._VALID_URL, url)
 115         domain_id, video_id = mobj.group('domain_id', 'id')
 116
 117         # Reverse engineered from JS code (see getDeviceID function)
 118         device_id = '%d:%d:%d%d' % (
 119             random.randint(1, 4), int(time.time()),
 120             random.randint(1e4, 99999), random.randint(1, 9))
 121
 122         result = self._call_api(domain_id, 'session/init', video_id, data={
 123             'nxp_devh': device_id,
 124             'nxp_userh': '',
 125             'precid': '0',
 126             'playlicense': '0',
 127             'screenx': '1920',
 128             'screeny': '1080',
 129             'playerversion': '6.0.00',
 130             'gateway': 'html5',
 131             'adGateway': '',
 132             'explicitlanguage': 'en-US',
 133             'addTextTemplates': '1',
 134             'addDomainData': '1',
 135             'addAdModel': '1',
 136         }, headers={
 137             'X-Request-Enable-Auth-Fallback': '1',
 138         })
 139
 140         cid = result['general']['cid']
 141
 142         # As described in [1] X-Request-Token generation algorithm is
 143         # as follows:
 144         #   md5( operation + domain_id + domain_secret )
 145         # where domain_secret is a static value that will be given by nexx.tv
 146         # as per [1]. Here is how this "secret" is generated (reversed
 147         # from _play.api.init function, search for clienttoken). So it's
 148         # actually not static and not that much of a secret.
 149         # 1. https://nexxtvstorage.blob.core.windows.net/files/201610/27.pdf
 150         secret = result['device']['clienttoken'][int(device_id[0]):]
 151         secret = secret[0:len(secret) - int(device_id[-1])]
 152
 153         op = 'byid'
 154
 155         # Reversed from JS code for _play.api.call function (search for
 156         # X-Request-Token)
 157         request_token = hashlib.md5(
 158             ''.join((op, domain_id, secret)).encode('utf-8')).hexdigest()
 159
 160         video = self._call_api(
 161             domain_id, 'videos/%s/%s' % (op, video_id), video_id, data={
 162                 'additionalfields': 'language,channel,actors,studio,licenseby,slug,subtitle,teaser,description',
 163                 'addInteractionOptions': '1',
 164                 'addStatusDetails': '1',
 165                 'addStreamDetails': '1',
 166                 'addCaptions': '1',
 167                 'addScenes': '1',
 168                 'addHotSpots': '1',
 169                 'addBumpers': '1',
 170                 'captionFormat': 'data',
 171             }, headers={
 172                 'X-Request-CID': cid,
 173                 'X-Request-Token': request_token,
 174             })
 175
 176         general = video['general']
 177         title = general['title']
 178
 179         stream_data = video['streamdata']
 180         language = general.get('language_raw') or ''
 181
 182         # TODO: reverse more cdns and formats
 183
 184         cdn = stream_data['cdnType']
 185         assert cdn == 'azure'
 186
 187         azure_locator = stream_data['azureLocator']
 188
 189         AZURE_URL = 'http://nx-p%02d.akamaized.net/'
 190
 191         for secure in ('s', ''):
 192             cdn_shield = stream_data.get('cdnShieldHTTP%s' % secure.upper())
 193             if cdn_shield:
 194                 azure_base = 'http%s://%s' % (secure, cdn_shield)
 195                 break
 196         else:
 197             azure_base = AZURE_URL % int(stream_data['azureAccount'].replace('nexxplayplus', ''))
 198
 199         is_ml = ',' in language
 200         azure_m3u8_url = '%s%s/%s_src%s.ism/Manifest(format=m3u8-aapl)' % (
 201             azure_base, azure_locator, video_id, ('_manifest' if is_ml else ''))
 202
 203         protection_token = try_get(
 204             video, lambda x: x['protectiondata']['token'], compat_str)
 205         if protection_token:
 206             azure_m3u8_url += '?hdnts=%s' % protection_token
 207
 208         formats = self._extract_m3u8_formats(
 209             azure_m3u8_url, video_id, 'mp4', entry_protocol='m3u8_native',
 210             m3u8_id='%s-hls' % cdn)
 211         self._sort_formats(formats)
 212
 213         return {
 214             'id': video_id,
 215             'title': title,
 216             'alt_title': general.get('subtitle'),
 217             'description': general.get('description'),
 218             'release_year': int_or_none(general.get('year')),
 219             'creator': general.get('studio') or general.get('studio_adref'),
 220             'thumbnail': try_get(
 221                 video, lambda x: x['imagedata']['thumb'], compat_str),
 222             'duration': parse_duration(general.get('runtime')),
 223             'timestamp': int_or_none(general.get('uploaded')),
 224             'episode_number': int_or_none(try_get(
 225                 video, lambda x: x['episodedata']['episode'])),
 226             'season_number': int_or_none(try_get(
 227                 video, lambda x: x['episodedata']['season'])),
 228             'formats': formats,
 229         }
 230
 231
 232 class NexxEmbedIE(InfoExtractor):
 233     _VALID_URL = r'https?://embed\.nexx(?:\.cloud|cdn\.com)/\d+/(?P<id>[^/?#&]+)'
 234     _TEST = {
 235         'url': 'http://embed.nexx.cloud/748/KC1614647Z27Y7T?autoplay=1',
 236         'md5': '16746bfc28c42049492385c989b26c4a',
 237         'info_dict': {
 238             'id': '161464',
 239             'ext': 'mp4',
 240             'title': 'Nervenkitzel Achterbahn',
 241             'alt_title': 'Karussellbauer in Deutschland',
 242             'description': 'md5:ffe7b1cc59a01f585e0569949aef73cc',
 243             'release_year': 2005,
 244             'creator': 'SPIEGEL TV',
 245             'thumbnail': r're:^https?://.*\.jpg$',
 246             'duration': 2761,
 247             'timestamp': 1394021479,
 248             'upload_date': '20140305',
 249         },
 250         'params': {
 251             'format': 'bestvideo',
 252             'skip_download': True,
 253         },
 254     }
 255
 256     @staticmethod
 257     def _extract_urls(webpage):
 258         # Reference:
 259         # 1. https://nx-s.akamaized.net/files/201510/44.pdf
 260
 261         # iFrame Embed Integration
 262         return [mobj.group('url') for mobj in re.finditer(
 263             r'<iframe[^>]+\bsrc=(["\'])(?P<url>(?:https?:)?//embed\.nexx(?:\.cloud|cdn\.com)/\d+/(?:(?!\1).)+)\1',
 264             webpage)]
 265
 266     def _real_extract(self, url):
 267         embed_id = self._match_id(url)
 268
 269         webpage = self._download_webpage(url, embed_id)
 270
 271         return self.url_result(NexxIE._extract_url(webpage), ie=NexxIE.ie_key())