ytdl/youtube_dl/extractor/nexx.py

348 lines
13 KiB
Python
Raw Normal View History

# coding: utf-8
from __future__ import unicode_literals
import hashlib
import random
import re
import time
from .common import InfoExtractor
from ..compat import compat_str
from ..utils import (
ExtractorError,
int_or_none,
parse_duration,
try_get,
urlencode_postdata,
)
class NexxIE(InfoExtractor):
_VALID_URL = r'''(?x)
(?:
https?://api\.nexx(?:\.cloud|cdn\.com)/v3/(?P<domain_id>\d+)/videos/byid/|
2018-03-07 21:25:04 +01:00
nexx:(?:(?P<domain_id_s>\d+):)?|
https?://arc\.nexx\.cloud/api/video/
)
(?P<id>\d+)
'''
_TESTS = [{
# movie
'url': 'https://api.nexx.cloud/v3/748/videos/byid/128907',
2017-11-25 18:13:17 +01:00
'md5': '828cea195be04e66057b846288295ba1',
'info_dict': {
'id': '128907',
'ext': 'mp4',
'title': 'Stiftung Warentest',
'alt_title': 'Wie ein Test abläuft',
'description': 'md5:d1ddb1ef63de721132abd38639cc2fd2',
'release_year': 2013,
'creator': 'SPIEGEL TV',
'thumbnail': r're:^https?://.*\.jpg$',
'duration': 2509,
'timestamp': 1384264416,
'upload_date': '20131112',
},
}, {
# episode
'url': 'https://api.nexx.cloud/v3/741/videos/byid/247858',
'info_dict': {
'id': '247858',
'ext': 'mp4',
'title': 'Return of the Golden Child (OV)',
'description': 'md5:5d969537509a92b733de21bae249dc63',
'release_year': 2017,
'thumbnail': r're:^https?://.*\.jpg$',
'duration': 1397,
'timestamp': 1495033267,
'upload_date': '20170517',
'episode_number': 2,
'season_number': 2,
},
'params': {
'skip_download': True,
},
}, {
# does not work via arc
'url': 'nexx:741:1269984',
'md5': 'c714b5b238b2958dc8d5642addba6886',
'info_dict': {
'id': '1269984',
'ext': 'mp4',
'title': '1 TAG ohne KLO... wortwörtlich! 😑',
'alt_title': '1 TAG ohne KLO... wortwörtlich! 😑',
'description': 'md5:4604539793c49eda9443ab5c5b1d612f',
'thumbnail': r're:^https?://.*\.jpg$',
'duration': 607,
'timestamp': 1518614955,
'upload_date': '20180214',
},
}, {
'url': 'https://api.nexxcdn.com/v3/748/videos/byid/128907',
'only_matching': True,
}, {
'url': 'nexx:748:128907',
'only_matching': True,
}, {
'url': 'nexx:128907',
'only_matching': True,
}, {
'url': 'https://arc.nexx.cloud/api/video/128907.json',
'only_matching': True,
}]
@staticmethod
def _extract_domain_id(webpage):
mobj = re.search(
r'<script\b[^>]+\bsrc=["\'](?:https?:)?//require\.nexx(?:\.cloud|cdn\.com)/(?P<id>\d+)',
webpage)
return mobj.group('id') if mobj else None
@staticmethod
def _extract_urls(webpage):
# Reference:
# 1. https://nx-s.akamaized.net/files/201510/44.pdf
entries = []
# JavaScript Integration
domain_id = NexxIE._extract_domain_id(webpage)
if domain_id:
2017-07-15 23:30:48 +02:00
for video_id in re.findall(
r'(?is)onPLAYReady.+?_play\.init\s*\(.+?\s*,\s*["\']?(\d+)',
webpage):
entries.append(
'https://api.nexx.cloud/v3/%s/videos/byid/%s'
% (domain_id, video_id))
# TODO: support more embed formats
return entries
@staticmethod
def _extract_url(webpage):
return NexxIE._extract_urls(webpage)[0]
def _handle_error(self, response):
status = int_or_none(try_get(
response, lambda x: x['metadata']['status']) or 200)
if 200 <= status < 300:
return
raise ExtractorError(
'%s said: %s' % (self.IE_NAME, response['metadata']['errorhint']),
expected=True)
def _call_api(self, domain_id, path, video_id, data=None, headers={}):
headers['Content-Type'] = 'application/x-www-form-urlencoded; charset=UTF-8'
result = self._download_json(
'https://api.nexx.cloud/v3/%s/%s' % (domain_id, path), video_id,
'Downloading %s JSON' % path, data=urlencode_postdata(data),
headers=headers)
self._handle_error(result)
return result['result']
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
domain_id = mobj.group('domain_id') or mobj.group('domain_id_s')
video_id = mobj.group('id')
video = None
response = self._download_json(
'https://arc.nexx.cloud/api/video/%s.json' % video_id,
video_id, fatal=False)
if response and isinstance(response, dict):
result = response.get('result')
if result and isinstance(result, dict):
video = result
# not all videos work via arc, e.g. nexx:741:1269984
if not video:
# Reverse engineered from JS code (see getDeviceID function)
device_id = '%d:%d:%d%d' % (
random.randint(1, 4), int(time.time()),
random.randint(1e4, 99999), random.randint(1, 9))
result = self._call_api(domain_id, 'session/init', video_id, data={
'nxp_devh': device_id,
'nxp_userh': '',
'precid': '0',
'playlicense': '0',
'screenx': '1920',
'screeny': '1080',
'playerversion': '6.0.00',
'gateway': 'html5',
'adGateway': '',
'explicitlanguage': 'en-US',
'addTextTemplates': '1',
'addDomainData': '1',
'addAdModel': '1',
}, headers={
'X-Request-Enable-Auth-Fallback': '1',
})
cid = result['general']['cid']
# As described in [1] X-Request-Token generation algorithm is
# as follows:
# md5( operation + domain_id + domain_secret )
# where domain_secret is a static value that will be given by nexx.tv
# as per [1]. Here is how this "secret" is generated (reversed
# from _play.api.init function, search for clienttoken). So it's
# actually not static and not that much of a secret.
# 1. https://nexxtvstorage.blob.core.windows.net/files/201610/27.pdf
secret = result['device']['clienttoken'][int(device_id[0]):]
secret = secret[0:len(secret) - int(device_id[-1])]
op = 'byid'
# Reversed from JS code for _play.api.call function (search for
# X-Request-Token)
request_token = hashlib.md5(
''.join((op, domain_id, secret)).encode('utf-8')).hexdigest()
video = self._call_api(
domain_id, 'videos/%s/%s' % (op, video_id), video_id, data={
'additionalfields': 'language,channel,actors,studio,licenseby,slug,subtitle,teaser,description',
'addInteractionOptions': '1',
'addStatusDetails': '1',
'addStreamDetails': '1',
'addCaptions': '1',
'addScenes': '1',
'addHotSpots': '1',
'addBumpers': '1',
'captionFormat': 'data',
}, headers={
'X-Request-CID': cid,
'X-Request-Token': request_token,
})
general = video['general']
title = general['title']
stream_data = video['streamdata']
language = general.get('language_raw') or ''
2017-11-25 18:13:17 +01:00
# TODO: reverse more cdns
cdn = stream_data['cdnType']
assert cdn == 'azure'
azure_locator = stream_data['azureLocator']
2017-11-25 18:13:17 +01:00
AZURE_URL = 'http://nx%s%02d.akamaized.net/'
def get_cdn_shield_base(shield_type='', prefix='-p'):
2017-11-25 18:13:17 +01:00
for secure in ('', 's'):
cdn_shield = stream_data.get('cdnShield%sHTTP%s' % (shield_type, secure.upper()))
2017-11-25 18:13:17 +01:00
if cdn_shield:
return 'http%s://%s' % (secure, cdn_shield)
2017-11-25 18:13:17 +01:00
else:
return AZURE_URL % (prefix, int(stream_data['azureAccount'].replace('nexxplayplus', '')))
azure_stream_base = get_cdn_shield_base()
is_ml = ',' in language
2017-11-25 18:13:17 +01:00
azure_manifest_url = '%s%s/%s_src%s.ism/Manifest' % (
azure_stream_base, azure_locator, video_id, ('_manifest' if is_ml else '')) + '%s'
protection_token = try_get(
video, lambda x: x['protectiondata']['token'], compat_str)
if protection_token:
2017-11-25 18:13:17 +01:00
azure_manifest_url += '?hdnts=%s' % protection_token
formats = self._extract_m3u8_formats(
2017-11-25 18:13:17 +01:00
azure_manifest_url % '(format=m3u8-aapl)',
video_id, 'mp4', 'm3u8_native',
m3u8_id='%s-hls' % cdn, fatal=False)
formats.extend(self._extract_mpd_formats(
azure_manifest_url % '(format=mpd-time-csf)',
video_id, mpd_id='%s-dash' % cdn, fatal=False))
formats.extend(self._extract_ism_formats(
azure_manifest_url % '', video_id, ism_id='%s-mss' % cdn, fatal=False))
azure_progressive_base = get_cdn_shield_base('Prog', '-d')
2017-11-25 18:13:17 +01:00
azure_file_distribution = stream_data.get('azureFileDistribution')
if azure_file_distribution:
fds = azure_file_distribution.split(',')
if fds:
for fd in fds:
ss = fd.split(':')
if len(ss) == 2:
tbr = int_or_none(ss[0])
if tbr:
f = {
'url': '%s%s/%s_src_%s_%d.mp4' % (
azure_progressive_base, azure_locator, video_id, ss[1], tbr),
'format_id': '%s-http-%d' % (cdn, tbr),
2017-11-25 18:13:17 +01:00
'tbr': tbr,
}
width_height = ss[1].split('x')
if len(width_height) == 2:
f.update({
'width': int_or_none(width_height[0]),
'height': int_or_none(width_height[1]),
})
formats.append(f)
self._sort_formats(formats)
return {
'id': video_id,
'title': title,
'alt_title': general.get('subtitle'),
'description': general.get('description'),
'release_year': int_or_none(general.get('year')),
'creator': general.get('studio') or general.get('studio_adref'),
'thumbnail': try_get(
video, lambda x: x['imagedata']['thumb'], compat_str),
'duration': parse_duration(general.get('runtime')),
'timestamp': int_or_none(general.get('uploaded')),
'episode_number': int_or_none(try_get(
video, lambda x: x['episodedata']['episode'])),
'season_number': int_or_none(try_get(
video, lambda x: x['episodedata']['season'])),
'formats': formats,
}
class NexxEmbedIE(InfoExtractor):
_VALID_URL = r'https?://embed\.nexx(?:\.cloud|cdn\.com)/\d+/(?P<id>[^/?#&]+)'
_TEST = {
'url': 'http://embed.nexx.cloud/748/KC1614647Z27Y7T?autoplay=1',
'md5': '16746bfc28c42049492385c989b26c4a',
'info_dict': {
'id': '161464',
'ext': 'mp4',
'title': 'Nervenkitzel Achterbahn',
'alt_title': 'Karussellbauer in Deutschland',
'description': 'md5:ffe7b1cc59a01f585e0569949aef73cc',
'release_year': 2005,
'creator': 'SPIEGEL TV',
'thumbnail': r're:^https?://.*\.jpg$',
'duration': 2761,
'timestamp': 1394021479,
'upload_date': '20140305',
},
'params': {
'format': 'bestvideo',
'skip_download': True,
},
}
@staticmethod
def _extract_urls(webpage):
# Reference:
# 1. https://nx-s.akamaized.net/files/201510/44.pdf
# iFrame Embed Integration
return [mobj.group('url') for mobj in re.finditer(
2017-07-16 00:23:19 +02:00
r'<iframe[^>]+\bsrc=(["\'])(?P<url>(?:https?:)?//embed\.nexx(?:\.cloud|cdn\.com)/\d+/(?:(?!\1).)+)\1',
webpage)]
def _real_extract(self, url):
embed_id = self._match_id(url)
webpage = self._download_webpage(url, embed_id)
return self.url_result(NexxIE._extract_url(webpage), ie=NexxIE.ie_key())