youtube_dl/extractor/myvideo.py

   1 from __future__ import unicode_literals
   2
   3 import binascii
   4 import base64
   5 import hashlib
   6 import re
   7 import json
   8
   9 from .common import InfoExtractor
  10 from ..utils import (
  11     compat_ord,
  12     compat_urllib_parse,
  13     compat_urllib_request,
  14
  15     ExtractorError,
  16 )
  17
  18
  19 class MyVideoIE(InfoExtractor):
  20     _VALID_URL = r'http://(?:www\.)?myvideo\.de/(?:[^/]+/)?watch/(?P<id>[0-9]+)/[^?/]+.*'
  21     IE_NAME = 'myvideo'
  22     _TEST = {
  23         'url': 'http://www.myvideo.de/watch/8229274/bowling_fail_or_win',
  24         'md5': '2d2753e8130479ba2cb7e0a37002053e',
  25         'info_dict': {
  26             'id': '8229274',
  27             'ext': 'flv',
  28             'title': 'bowling-fail-or-win',
  29         }
  30     }
  31
  32     # Original Code from: https://github.com/dersphere/plugin.video.myvideo_de.git
  33     # Released into the Public Domain by Tristan Fischer on 2013-05-19
  34     # https://github.com/rg3/youtube-dl/pull/842
  35     def __rc4crypt(self,data, key):
  36         x = 0
  37         box = list(range(256))
  38         for i in list(range(256)):
  39             x = (x + box[i] + compat_ord(key[i % len(key)])) % 256
  40             box[i], box[x] = box[x], box[i]
  41         x = 0
  42         y = 0
  43         out = ''
  44         for char in data:
  45             x = (x + 1) % 256
  46             y = (y + box[x]) % 256
  47             box[x], box[y] = box[y], box[x]
  48             out += chr(compat_ord(char) ^ box[(box[x] + box[y]) % 256])
  49         return out
  50
  51     def __md5(self,s):
  52         return hashlib.md5(s).hexdigest().encode()
  53
  54     def _real_extract(self,url):
  55         mobj = re.match(self._VALID_URL, url)
  56         video_id = mobj.group('id')
  57
  58         GK = (
  59           b'WXpnME1EZGhNRGhpTTJNM01XVmhOREU0WldNNVpHTTJOakpt'
  60           b'TW1FMU5tVTBNR05pWkRaa05XRXhNVFJoWVRVd1ptSXhaVEV3'
  61           b'TnpsbA0KTVRkbU1tSTRNdz09'
  62         )
  63
  64         # Get video webpage
  65         webpage_url = 'http://www.myvideo.de/watch/%s' % video_id
  66         webpage = self._download_webpage(webpage_url, video_id)
  67
  68         mobj = re.search('source src=\'(.+?)[.]([^.]+)\'', webpage)
  69         if mobj is not None:
  70             self.report_extraction(video_id)
  71             video_url = mobj.group(1) + '.flv'
  72
  73             video_title = self._html_search_regex('<title>([^<]+)</title>',
  74                 webpage, 'title')
  75
  76             return {
  77                 'id': video_id,
  78                 'url': video_url,
  79                 'title': video_title,
  80             }
  81
  82         mobj = re.search(r'data-video-service="/service/data/video/%s/config' % video_id, webpage)
  83         if mobj is not None:
  84             request = compat_urllib_request.Request('http://www.myvideo.de/service/data/video/%s/config' % video_id, '')
  85             response = self._download_webpage(request, video_id,
  86                                               'Downloading video info')
  87             info = json.loads(base64.b64decode(response).decode('utf-8'))
  88             return {
  89                 'id': video_id,
  90                 'title': info['title'],
  91                 'url': info['streaming_url'].replace('rtmpe', 'rtmpt'),
  92                 'play_path': info['filename'],
  93                 'ext': 'flv',
  94                 'thumbnail': info['thumbnail'][0]['url'],
  95             }
  96
  97         # try encxml
  98         mobj = re.search('var flashvars={(.+?)}', webpage)
  99         if mobj is None:
 100             raise ExtractorError('Unable to extract video')
 101
 102         params = {}
 103         encxml = ''
 104         sec = mobj.group(1)
 105         for (a, b) in re.findall('(.+?):\'(.+?)\',?', sec):
 106             if not a == '_encxml':
 107                 params[a] = b
 108             else:
 109                 encxml = compat_urllib_parse.unquote(b)
 110         if not params.get('domain'):
 111             params['domain'] = 'www.myvideo.de'
 112         xmldata_url = '%s?%s' % (encxml, compat_urllib_parse.urlencode(params))
 113         if 'flash_playertype=MTV' in xmldata_url:
 114             self._downloader.report_warning('avoiding MTV player')
 115             xmldata_url = (
 116                 'http://www.myvideo.de/dynamic/get_player_video_xml.php'
 117                 '?flash_playertype=D&ID=%s&_countlimit=4&autorun=yes'
 118             ) % video_id
 119
 120         # get enc data
 121         enc_data = self._download_webpage(xmldata_url, video_id).split('=')[1]
 122         enc_data_b = binascii.unhexlify(enc_data)
 123         sk = self.__md5(
 124             base64.b64decode(base64.b64decode(GK)) +
 125             self.__md5(
 126                 str(video_id).encode('utf-8')
 127             )
 128         )
 129         dec_data = self.__rc4crypt(enc_data_b, sk)
 130
 131         # extracting infos
 132         self.report_extraction(video_id)
 133
 134         video_url = None
 135         mobj = re.search('connectionurl=\'(.*?)\'', dec_data)
 136         if mobj:
 137             video_url = compat_urllib_parse.unquote(mobj.group(1))
 138             if 'myvideo2flash' in video_url:
 139                 self.report_warning(
 140                     'Rewriting URL to use unencrypted rtmp:// ...',
 141                     video_id)
 142                 video_url = video_url.replace('rtmpe://', 'rtmp://')
 143
 144         if not video_url:
 145             # extract non rtmp videos
 146             mobj = re.search('path=\'(http.*?)\' source=\'(.*?)\'', dec_data)
 147             if mobj is None:
 148                 raise ExtractorError('unable to extract url')
 149             video_url = compat_urllib_parse.unquote(mobj.group(1)) + compat_urllib_parse.unquote(mobj.group(2))
 150
 151         video_file = self._search_regex('source=\'(.*?)\'', dec_data, 'video file')
 152         video_file = compat_urllib_parse.unquote(video_file)
 153
 154         if not video_file.endswith('f4m'):
 155             ppath, prefix = video_file.split('.')
 156             video_playpath = '%s:%s' % (prefix, ppath)
 157         else:
 158             video_playpath = ''
 159
 160         video_swfobj = self._search_regex('swfobject.embedSWF\(\'(.+?)\'', webpage, 'swfobj')
 161         video_swfobj = compat_urllib_parse.unquote(video_swfobj)
 162
 163         video_title = self._html_search_regex("<h1(?: class='globalHd')?>(.*?)</h1>",
 164             webpage, 'title')
 165
 166         return {
 167             'id': video_id,
 168             'url': video_url,
 169             'tc_url': video_url,
 170             'title': video_title,
 171             'ext': 'flv',
 172             'play_path': video_playpath,
 173             'player_url': video_swfobj,
 174         }
 175