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                             https?://arc\.nexx\.cloud/api/video/
  26                         )
  27                         (?P<id>\d+)
  28                     '''
  29     _TESTS = [{
  30         # movie
  31         'url': 'https://api.nexx.cloud/v3/748/videos/byid/128907',
  32         'md5': '828cea195be04e66057b846288295ba1',
  33         'info_dict': {
  34             'id': '128907',
  35             'ext': 'mp4',
  36             'title': 'Stiftung Warentest',
  37             'alt_title': 'Wie ein Test abläuft',
  38             'description': 'md5:d1ddb1ef63de721132abd38639cc2fd2',
  39             'release_year': 2013,
  40             'creator': 'SPIEGEL TV',
  41             'thumbnail': r're:^https?://.*\.jpg$',
  42             'duration': 2509,
  43             'timestamp': 1384264416,
  44             'upload_date': '20131112',
  45         },
  46     }, {
  47         # episode
  48         'url': 'https://api.nexx.cloud/v3/741/videos/byid/247858',
  49         'info_dict': {
  50             'id': '247858',
  51             'ext': 'mp4',
  52             'title': 'Return of the Golden Child (OV)',
  53             'description': 'md5:5d969537509a92b733de21bae249dc63',
  54             'release_year': 2017,
  55             'thumbnail': r're:^https?://.*\.jpg$',
  56             'duration': 1397,
  57             'timestamp': 1495033267,
  58             'upload_date': '20170517',
  59             'episode_number': 2,
  60             'season_number': 2,
  61         },
  62         'params': {
  63             'skip_download': True,
  64         },
  65     }, {
  66         # does not work via arc
  67         'url': 'nexx:741:1269984',
  68         'md5': 'c714b5b238b2958dc8d5642addba6886',
  69         'info_dict': {
  70             'id': '1269984',
  71             'ext': 'mp4',
  72             'title': '1 TAG ohne KLO... wortwörtlich! 😑',
  73             'alt_title': '1 TAG ohne KLO... wortwörtlich! 😑',
  74             'description': 'md5:4604539793c49eda9443ab5c5b1d612f',
  75             'thumbnail': r're:^https?://.*\.jpg$',
  76             'duration': 607,
  77             'timestamp': 1518614955,
  78             'upload_date': '20180214',
  79         },
  80     }, {
  81         'url': 'https://api.nexxcdn.com/v3/748/videos/byid/128907',
  82         'only_matching': True,
  83     }, {
  84         'url': 'nexx:748:128907',
  85         'only_matching': True,
  86     }, {
  87         'url': 'nexx:128907',
  88         'only_matching': True,
  89     }, {
  90         'url': 'https://arc.nexx.cloud/api/video/128907.json',
  91         'only_matching': True,
  92     }]
  93
  94     @staticmethod
  95     def _extract_domain_id(webpage):
  96         mobj = re.search(
  97             r'<script\b[^>]+\bsrc=["\'](?:https?:)?//require\.nexx(?:\.cloud|cdn\.com)/(?P<id>\d+)',
  98             webpage)
  99         return mobj.group('id') if mobj else None
 100
 101     @staticmethod
 102     def _extract_urls(webpage):
 103         # Reference:
 104         # 1. https://nx-s.akamaized.net/files/201510/44.pdf
 105
 106         entries = []
 107
 108         # JavaScript Integration
 109         domain_id = NexxIE._extract_domain_id(webpage)
 110         if domain_id:
 111             for video_id in re.findall(
 112                     r'(?is)onPLAYReady.+?_play\.init\s*\(.+?\s*,\s*["\']?(\d+)',
 113                     webpage):
 114                 entries.append(
 115                     'https://api.nexx.cloud/v3/%s/videos/byid/%s'
 116                     % (domain_id, video_id))
 117
 118         # TODO: support more embed formats
 119
 120         return entries
 121
 122     @staticmethod
 123     def _extract_url(webpage):
 124         return NexxIE._extract_urls(webpage)[0]
 125
 126     def _handle_error(self, response):
 127         status = int_or_none(try_get(
 128             response, lambda x: x['metadata']['status']) or 200)
 129         if 200 <= status < 300:
 130             return
 131         raise ExtractorError(
 132             '%s said: %s' % (self.IE_NAME, response['metadata']['errorhint']),
 133             expected=True)
 134
 135     def _call_api(self, domain_id, path, video_id, data=None, headers={}):
 136         headers['Content-Type'] = 'application/x-www-form-urlencoded; charset=UTF-8'
 137         result = self._download_json(
 138             'https://api.nexx.cloud/v3/%s/%s' % (domain_id, path), video_id,
 139             'Downloading %s JSON' % path, data=urlencode_postdata(data),
 140             headers=headers)
 141         self._handle_error(result)
 142         return result['result']
 143
 144     def _real_extract(self, url):
 145         mobj = re.match(self._VALID_URL, url)
 146         domain_id = mobj.group('domain_id') or mobj.group('domain_id_s')
 147         video_id = mobj.group('id')
 148
 149         video = None
 150
 151         response = self._download_json(
 152             'https://arc.nexx.cloud/api/video/%s.json' % video_id,
 153             video_id, fatal=False)
 154         if response and isinstance(response, dict):
 155             result = response.get('result')
 156             if result and isinstance(result, dict):
 157                 video = result
 158
 159         # not all videos work via arc, e.g. nexx:741:1269984
 160         if not video:
 161             # Reverse engineered from JS code (see getDeviceID function)
 162             device_id = '%d:%d:%d%d' % (
 163                 random.randint(1, 4), int(time.time()),
 164                 random.randint(1e4, 99999), random.randint(1, 9))
 165
 166             result = self._call_api(domain_id, 'session/init', video_id, data={
 167                 'nxp_devh': device_id,
 168                 'nxp_userh': '',
 169                 'precid': '0',
 170                 'playlicense': '0',
 171                 'screenx': '1920',
 172                 'screeny': '1080',
 173                 'playerversion': '6.0.00',
 174                 'gateway': 'html5',
 175                 'adGateway': '',
 176                 'explicitlanguage': 'en-US',
 177                 'addTextTemplates': '1',
 178                 'addDomainData': '1',
 179                 'addAdModel': '1',
 180             }, headers={
 181                 'X-Request-Enable-Auth-Fallback': '1',
 182             })
 183
 184             cid = result['general']['cid']
 185
 186             # As described in [1] X-Request-Token generation algorithm is
 187             # as follows:
 188             #   md5( operation + domain_id + domain_secret )
 189             # where domain_secret is a static value that will be given by nexx.tv
 190             # as per [1]. Here is how this "secret" is generated (reversed
 191             # from _play.api.init function, search for clienttoken). So it's
 192             # actually not static and not that much of a secret.
 193             # 1. https://nexxtvstorage.blob.core.windows.net/files/201610/27.pdf
 194             secret = result['device']['clienttoken'][int(device_id[0]):]
 195             secret = secret[0:len(secret) - int(device_id[-1])]
 196
 197             op = 'byid'
 198
 199             # Reversed from JS code for _play.api.call function (search for
 200             # X-Request-Token)
 201             request_token = hashlib.md5(
 202                 ''.join((op, domain_id, secret)).encode('utf-8')).hexdigest()
 203
 204             video = self._call_api(
 205                 domain_id, 'videos/%s/%s' % (op, video_id), video_id, data={
 206                     'additionalfields': 'language,channel,actors,studio,licenseby,slug,subtitle,teaser,description',
 207                     'addInteractionOptions': '1',
 208                     'addStatusDetails': '1',
 209                     'addStreamDetails': '1',
 210                     'addCaptions': '1',
 211                     'addScenes': '1',
 212                     'addHotSpots': '1',
 213                     'addBumpers': '1',
 214                     'captionFormat': 'data',
 215                 }, headers={
 216                     'X-Request-CID': cid,
 217                     'X-Request-Token': request_token,
 218                 })
 219
 220         general = video['general']
 221         title = general['title']
 222
 223         stream_data = video['streamdata']
 224         language = general.get('language_raw') or ''
 225
 226         # TODO: reverse more cdns
 227
 228         cdn = stream_data['cdnType']
 229         assert cdn == 'azure'
 230
 231         azure_locator = stream_data['azureLocator']
 232
 233         AZURE_URL = 'http://nx%s%02d.akamaized.net/'
 234
 235         def get_cdn_shield_base(shield_type='', prefix='-p'):
 236             for secure in ('', 's'):
 237                 cdn_shield = stream_data.get('cdnShield%sHTTP%s' % (shield_type, secure.upper()))
 238                 if cdn_shield:
 239                     return 'http%s://%s' % (secure, cdn_shield)
 240             else:
 241                 return AZURE_URL % (prefix, int(stream_data['azureAccount'].replace('nexxplayplus', '')))
 242
 243         azure_stream_base = get_cdn_shield_base()
 244         is_ml = ',' in language
 245         azure_manifest_url = '%s%s/%s_src%s.ism/Manifest' % (
 246             azure_stream_base, azure_locator, video_id, ('_manifest' if is_ml else '')) + '%s'
 247
 248         protection_token = try_get(
 249             video, lambda x: x['protectiondata']['token'], compat_str)
 250         if protection_token:
 251             azure_manifest_url += '?hdnts=%s' % protection_token
 252
 253         formats = self._extract_m3u8_formats(
 254             azure_manifest_url % '(format=m3u8-aapl)',
 255             video_id, 'mp4', 'm3u8_native',
 256             m3u8_id='%s-hls' % cdn, fatal=False)
 257         formats.extend(self._extract_mpd_formats(
 258             azure_manifest_url % '(format=mpd-time-csf)',
 259             video_id, mpd_id='%s-dash' % cdn, fatal=False))
 260         formats.extend(self._extract_ism_formats(
 261             azure_manifest_url % '', video_id, ism_id='%s-mss' % cdn, fatal=False))
 262
 263         azure_progressive_base = get_cdn_shield_base('Prog', '-d')
 264         azure_file_distribution = stream_data.get('azureFileDistribution')
 265         if azure_file_distribution:
 266             fds = azure_file_distribution.split(',')
 267             if fds:
 268                 for fd in fds:
 269                     ss = fd.split(':')
 270                     if len(ss) == 2:
 271                         tbr = int_or_none(ss[0])
 272                         if tbr:
 273                             f = {
 274                                 'url': '%s%s/%s_src_%s_%d.mp4' % (
 275                                     azure_progressive_base, azure_locator, video_id, ss[1], tbr),
 276                                 'format_id': '%s-http-%d' % (cdn, tbr),
 277                                 'tbr': tbr,
 278                             }
 279                             width_height = ss[1].split('x')
 280                             if len(width_height) == 2:
 281                                 f.update({
 282                                     'width': int_or_none(width_height[0]),
 283                                     'height': int_or_none(width_height[1]),
 284                                 })
 285                             formats.append(f)
 286
 287         self._sort_formats(formats)
 288
 289         return {
 290             'id': video_id,
 291             'title': title,
 292             'alt_title': general.get('subtitle'),
 293             'description': general.get('description'),
 294             'release_year': int_or_none(general.get('year')),
 295             'creator': general.get('studio') or general.get('studio_adref'),
 296             'thumbnail': try_get(
 297                 video, lambda x: x['imagedata']['thumb'], compat_str),
 298             'duration': parse_duration(general.get('runtime')),
 299             'timestamp': int_or_none(general.get('uploaded')),
 300             'episode_number': int_or_none(try_get(
 301                 video, lambda x: x['episodedata']['episode'])),
 302             'season_number': int_or_none(try_get(
 303                 video, lambda x: x['episodedata']['season'])),
 304             'formats': formats,
 305         }
 306
 307
 308 class NexxEmbedIE(InfoExtractor):
 309     _VALID_URL = r'https?://embed\.nexx(?:\.cloud|cdn\.com)/\d+/(?P<id>[^/?#&]+)'
 310     _TEST = {
 311         'url': 'http://embed.nexx.cloud/748/KC1614647Z27Y7T?autoplay=1',
 312         'md5': '16746bfc28c42049492385c989b26c4a',
 313         'info_dict': {
 314             'id': '161464',
 315             'ext': 'mp4',
 316             'title': 'Nervenkitzel Achterbahn',
 317             'alt_title': 'Karussellbauer in Deutschland',
 318             'description': 'md5:ffe7b1cc59a01f585e0569949aef73cc',
 319             'release_year': 2005,
 320             'creator': 'SPIEGEL TV',
 321             'thumbnail': r're:^https?://.*\.jpg$',
 322             'duration': 2761,
 323             'timestamp': 1394021479,
 324             'upload_date': '20140305',
 325         },
 326         'params': {
 327             'format': 'bestvideo',
 328             'skip_download': True,
 329         },
 330     }
 331
 332     @staticmethod
 333     def _extract_urls(webpage):
 334         # Reference:
 335         # 1. https://nx-s.akamaized.net/files/201510/44.pdf
 336
 337         # iFrame Embed Integration
 338         return [mobj.group('url') for mobj in re.finditer(
 339             r'<iframe[^>]+\bsrc=(["\'])(?P<url>(?:https?:)?//embed\.nexx(?:\.cloud|cdn\.com)/\d+/(?:(?!\1).)+)\1',
 340             webpage)]
 341
 342     def _real_extract(self, url):
 343         embed_id = self._match_id(url)
 344
 345         webpage = self._download_webpage(url, embed_id)
 346
 347         return self.url_result(NexxIE._extract_url(webpage), ie=NexxIE.ie_key())