_ Git - youtube-dl/blob - 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<media>(?:[^/]+/)+select/media/)|(?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         'url': 'https://player.theplatform.com/p/D6x-PC/pulse_preview/embed/select/media/yMBg9E8KFxZD',
  61         'info_dict': {
  62             'id': 'yMBg9E8KFxZD',
  63             'ext': 'mp4',
  64             'description': 'md5:644ad9188d655b742f942bf2e06b002d',
  65             'title': 'HIGHLIGHTS: USA bag first ever series Cup win',
  66         }
  67     }, {
  68         'url': 'http://player.theplatform.com/p/NnzsPC/widget/select/media/4Y0TlYUr_ZT7',
  69         'only_matching': True,
  70     }]
  71
  72     @staticmethod
  73     def _sign_url(url, sig_key, sig_secret, life=600, include_qs=False):
  74         flags = '10' if include_qs else '00'
  75         expiration_date = '%x' % (int(time.time()) + life)
  76
  77         def str_to_hex(str):
  78             return binascii.b2a_hex(str.encode('ascii')).decode('ascii')
  79
  80         def hex_to_str(hex):
  81             return binascii.a2b_hex(hex)
  82
  83         relative_path = url.split('http://link.theplatform.com/s/')[1].split('?')[0]
  84         clear_text = hex_to_str(flags + expiration_date + str_to_hex(relative_path))
  85         checksum = hmac.new(sig_key.encode('ascii'), clear_text, hashlib.sha1).hexdigest()
  86         sig = flags + expiration_date + checksum + str_to_hex(sig_secret)
  87         return '%s&sig=%s' % (url, sig)
  88
  89     def _real_extract(self, url):
  90         url, smuggled_data = unsmuggle_url(url, {})
  91
  92         mobj = re.match(self._VALID_URL, url)
  93         provider_id = mobj.group('provider_id')
  94         video_id = mobj.group('id')
  95
  96         if not provider_id:
  97             provider_id = 'dJ5BDC'
  98
  99         path = provider_id
 100         if mobj.group('media'):
 101             path += '/media'
 102         path += '/' + video_id
 103
 104         if smuggled_data.get('force_smil_url', False):
 105             smil_url = url
 106         elif mobj.group('config'):
 107             config_url = url + '&form=json'
 108             config_url = config_url.replace('swf/', 'config/')
 109             config_url = config_url.replace('onsite/', 'onsite/config/')
 110             config = self._download_json(config_url, video_id, 'Downloading config')
 111             if 'releaseUrl' in config:
 112                 release_url = config['releaseUrl']
 113             else:
 114                 release_url = 'http://link.theplatform.com/s/%s?mbr=true' % path
 115             smil_url = release_url + '&format=SMIL&formats=MPEG4&manifest=f4m'
 116         else:
 117             smil_url = 'http://link.theplatform.com/s/%s/meta.smil?format=smil&mbr=true' % path
 118
 119         sig = smuggled_data.get('sig')
 120         if sig:
 121             smil_url = self._sign_url(smil_url, sig['key'], sig['secret'])
 122
 123         meta = self._download_xml(smil_url, video_id)
 124         try:
 125             error_msg = next(
 126                 n.attrib['abstract']
 127                 for n in meta.findall(_x('.//smil:ref'))
 128                 if n.attrib.get('title') == 'Geographic Restriction' or n.attrib.get('title') == 'Expired')
 129         except StopIteration:
 130             pass
 131         else:
 132             raise ExtractorError(error_msg, expected=True)
 133
 134         info_url = 'http://link.theplatform.com/s/%s?format=preview' % path
 135         info_json = self._download_webpage(info_url, video_id)
 136         info = json.loads(info_json)
 137
 138         subtitles = {}
 139         captions = info.get('captions')
 140         if isinstance(captions, list):
 141             for caption in captions:
 142                 lang, src, mime = caption.get('lang', 'en'), caption.get('src'), caption.get('type')
 143                 subtitles[lang] = [{
 144                     'ext': 'srt' if mime == 'text/srt' else 'ttml',
 145                     'url': src,
 146                 }]
 147
 148         head = meta.find(_x('smil:head'))
 149         body = meta.find(_x('smil:body'))
 150
 151         f4m_node = body.find(_x('smil:seq//smil:video'))
 152         if f4m_node is None:
 153             f4m_node = body.find(_x('smil:seq/smil:video'))
 154         if f4m_node is not None and '.f4m' in f4m_node.attrib['src']:
 155             f4m_url = f4m_node.attrib['src']
 156             if 'manifest.f4m?' not in f4m_url:
 157                 f4m_url += '?'
 158             # the parameters are from syfy.com, other sites may use others,
 159             # they also work for nbc.com
 160             f4m_url += '&g=UXWGVKRWHFSP&hdcore=3.0.3'
 161             formats = self._extract_f4m_formats(f4m_url, video_id)
 162         else:
 163             formats = []
 164             switch = body.find(_x('smil:switch'))
 165             if switch is None:
 166                 switch = body.find(_x('smil:par//smil:switch'))
 167             if switch is None:
 168                 switch = body.find(_x('smil:par/smil:switch'))
 169             if switch is None:
 170                 switch = body.find(_x('smil:par'))
 171             if switch is not None:
 172                 base_url = head.find(_x('smil:meta')).attrib['base']
 173                 for f in switch.findall(_x('smil:video')):
 174                     attr = f.attrib
 175                     width = int_or_none(attr.get('width'))
 176                     height = int_or_none(attr.get('height'))
 177                     vbr = int_or_none(attr.get('system-bitrate'), 1000)
 178                     format_id = '%dx%d_%dk' % (width, height, vbr)
 179                     formats.append({
 180                         'format_id': format_id,
 181                         'url': base_url,
 182                         'play_path': 'mp4:' + attr['src'],
 183                         'ext': 'flv',
 184                         'width': width,
 185                         'height': height,
 186                         'vbr': vbr,
 187                     })
 188             else:
 189                 switch = body.find(_x('smil:seq//smil:switch'))
 190                 if switch is None:
 191                     switch = body.find(_x('smil:seq/smil:switch'))
 192                 for f in switch.findall(_x('smil:video')):
 193                     attr = f.attrib
 194                     vbr = int_or_none(attr.get('system-bitrate'), 1000)
 195                     ext = determine_ext(attr['src'])
 196                     if ext == 'once':
 197                         ext = 'mp4'
 198                     formats.append({
 199                         'format_id': compat_str(vbr),
 200                         'url': attr['src'],
 201                         'vbr': vbr,
 202                         'ext': ext,
 203                     })
 204             self._sort_formats(formats)
 205
 206         return {
 207             'id': video_id,
 208             'title': info['title'],
 209             'subtitles': subtitles,
 210             'formats': formats,
 211             'description': info['description'],
 212             'thumbnail': info['defaultThumbnailUrl'],
 213             'duration': int_or_none(info.get('duration'), 1000),
 214         }