youtube_dl/extractor/theplatform.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import json
   5 import time
   6 import hmac
   7 import binascii
   8 import hashlib
   9
  10
  11 from .common import InfoExtractor
  12 from ..compat import (
  13     compat_str,
  14 )
  15 from ..utils import (
  16     determine_ext,
  17     ExtractorError,
  18     xpath_with_ns,
  19     unsmuggle_url,
  20     int_or_none,
  21 )
  22
  23 _x = lambda p: xpath_with_ns(p, {'smil': 'http://www.w3.org/2005/SMIL21/Language'})
  24
  25
  26 class ThePlatformIE(InfoExtractor):
  27     _VALID_URL = r'''(?x)
  28         (?:https?://(?:link|player)\.theplatform\.com/[sp]/(?P<provider_id>[^/]+)/
  29            (?P<config>(?:[^/\?]+/(?:swf|config)|onsite)/select/)?
  30          |theplatform:)(?P<id>[^/\?&]+)'''
  31
  32     _TESTS = [{
  33         # from http://www.metacafe.com/watch/cb-e9I_cZgTgIPd/blackberrys_big_bold_z30/
  34         'url': 'http://link.theplatform.com/s/dJ5BDC/e9I_cZgTgIPd/meta.smil?format=smil&Tracking=true&mbr=true',
  35         'info_dict': {
  36             'id': 'e9I_cZgTgIPd',
  37             'ext': 'flv',
  38             'title': 'Blackberry\'s big, bold Z30',
  39             'description': 'The Z30 is Blackberry\'s biggest, baddest mobile messaging device yet.',
  40             'duration': 247,
  41         },
  42         'params': {
  43             # rtmp download
  44             'skip_download': True,
  45         },
  46     }, {
  47         # from http://www.cnet.com/videos/tesla-model-s-a-second-step-towards-a-cleaner-motoring-future/
  48         'url': 'http://link.theplatform.com/s/kYEXFC/22d_qsQ6MIRT',
  49         'info_dict': {
  50             'id': '22d_qsQ6MIRT',
  51             'ext': 'flv',
  52             'description': 'md5:ac330c9258c04f9d7512cf26b9595409',
  53             'title': 'Tesla Model S: A second step towards a cleaner motoring future',
  54         },
  55         'params': {
  56             # rtmp download
  57             'skip_download': True,
  58         }
  59     }]
  60
  61     @staticmethod
  62     def _sign_url(url, sig_key, sig_secret, life=600, include_qs=False):
  63         flags = '10' if include_qs else '00'
  64         expiration_date = '%x' % (int(time.time()) + life)
  65
  66         def str_to_hex(str):
  67             return binascii.b2a_hex(str.encode('ascii')).decode('ascii')
  68
  69         def hex_to_str(hex):
  70             return binascii.a2b_hex(hex)
  71
  72         relative_path = url.split('http://link.theplatform.com/s/')[1].split('?')[0]
  73         clear_text = hex_to_str(flags + expiration_date + str_to_hex(relative_path))
  74         checksum = hmac.new(sig_key.encode('ascii'), clear_text, hashlib.sha1).hexdigest()
  75         sig = flags + expiration_date + checksum + str_to_hex(sig_secret)
  76         return '%s&sig=%s' % (url, sig)
  77
  78     def _real_extract(self, url):
  79         url, smuggled_data = unsmuggle_url(url, {})
  80
  81         mobj = re.match(self._VALID_URL, url)
  82         provider_id = mobj.group('provider_id')
  83         video_id = mobj.group('id')
  84
  85         if not provider_id:
  86             provider_id = 'dJ5BDC'
  87
  88         if smuggled_data.get('force_smil_url', False):
  89             smil_url = url
  90         elif mobj.group('config'):
  91             config_url = url + '&form=json'
  92             config_url = config_url.replace('swf/', 'config/')
  93             config_url = config_url.replace('onsite/', 'onsite/config/')
  94             config = self._download_json(config_url, video_id, 'Downloading config')
  95             smil_url = config['releaseUrl'] + '&format=SMIL&formats=MPEG4&manifest=f4m'
  96         else:
  97             smil_url = ('http://link.theplatform.com/s/{0}/{1}/meta.smil?'
  98                         'format=smil&mbr=true'.format(provider_id, video_id))
  99
 100         sig = smuggled_data.get('sig')
 101         if sig:
 102             smil_url = self._sign_url(smil_url, sig['key'], sig['secret'])
 103
 104         meta = self._download_xml(smil_url, video_id)
 105         try:
 106             error_msg = next(
 107                 n.attrib['abstract']
 108                 for n in meta.findall(_x('.//smil:ref'))
 109                 if n.attrib.get('title') == 'Geographic Restriction' or n.attrib.get('title') == 'Expired')
 110         except StopIteration:
 111             pass
 112         else:
 113             raise ExtractorError(error_msg, expected=True)
 114
 115         info_url = 'http://link.theplatform.com/s/{0}/{1}?format=preview'.format(provider_id, video_id)
 116         info_json = self._download_webpage(info_url, video_id)
 117         info = json.loads(info_json)
 118
 119         subtitles = {}
 120         captions = info.get('captions')
 121         if isinstance(captions, list):
 122             for caption in captions:
 123                 lang, src, mime = caption.get('lang', 'en'), caption.get('src'), caption.get('type')
 124                 subtitles[lang] = [{
 125                     'ext': 'srt' if mime == 'text/srt' else 'ttml',
 126                     'url': src,
 127                 }]
 128
 129         head = meta.find(_x('smil:head'))
 130         body = meta.find(_x('smil:body'))
 131
 132         f4m_node = body.find(_x('smil:seq//smil:video'))
 133         if f4m_node is None:
 134             f4m_node = body.find(_x('smil:seq/smil:video'))
 135         if f4m_node is not None and '.f4m' in f4m_node.attrib['src']:
 136             f4m_url = f4m_node.attrib['src']
 137             if 'manifest.f4m?' not in f4m_url:
 138                 f4m_url += '?'
 139             # the parameters are from syfy.com, other sites may use others,
 140             # they also work for nbc.com
 141             f4m_url += '&g=UXWGVKRWHFSP&hdcore=3.0.3'
 142             formats = self._extract_f4m_formats(f4m_url, video_id)
 143         else:
 144             formats = []
 145             switch = body.find(_x('smil:switch'))
 146             if switch is None:
 147                 switch = body.find(_x('smil:par//smil:switch'))
 148             if switch is None:
 149                 switch = body.find(_x('smil:par/smil:switch'))
 150             if switch is None:
 151                 switch = body.find(_x('smil:par'))
 152             if switch is not None:
 153                 base_url = head.find(_x('smil:meta')).attrib['base']
 154                 for f in switch.findall(_x('smil:video')):
 155                     attr = f.attrib
 156                     width = int_or_none(attr.get('width'))
 157                     height = int_or_none(attr.get('height'))
 158                     vbr = int_or_none(attr.get('system-bitrate'), 1000)
 159                     format_id = '%dx%d_%dk' % (width, height, vbr)
 160                     formats.append({
 161                         'format_id': format_id,
 162                         'url': base_url,
 163                         'play_path': 'mp4:' + attr['src'],
 164                         'ext': 'flv',
 165                         'width': width,
 166                         'height': height,
 167                         'vbr': vbr,
 168                     })
 169             else:
 170                 switch = body.find(_x('smil:seq//smil:switch'))
 171                 if switch is None:
 172                     switch = body.find(_x('smil:seq/smil:switch'))
 173                 for f in switch.findall(_x('smil:video')):
 174                     attr = f.attrib
 175                     vbr = int_or_none(attr.get('system-bitrate'), 1000)
 176                     ext = determine_ext(attr['src'])
 177                     if ext == 'once':
 178                         ext = 'mp4'
 179                     formats.append({
 180                         'format_id': compat_str(vbr),
 181                         'url': attr['src'],
 182                         'vbr': vbr,
 183                         'ext': ext,
 184                     })
 185             self._sort_formats(formats)
 186
 187         return {
 188             'id': video_id,
 189             'title': info['title'],
 190             'subtitles': subtitles,
 191             'formats': formats,
 192             'description': info['description'],
 193             'thumbnail': info['defaultThumbnailUrl'],
 194             'duration': int_or_none(info.get('duration'), 1000),
 195         }