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