Raphaël G. Git Repositories - youtubedl/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 )
  21
  22 _x = lambda p: xpath_with_ns(p, {'smil': 'http://www.w3.org/2005/SMIL21/Language'})
  23
  24
  25 class ThePlatformIE(InfoExtractor):
  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, mime = caption.get('lang', 'en'), caption.get('src'), caption.get('type')
 110                 subtitles[lang] = [{
 111                     'ext': 'srt' if mime == 'text/srt' else 'ttml',
 112                     'url': src,
 113                 }]
 114
 115         head = meta.find(_x('smil:head'))
 116         body = meta.find(_x('smil:body'))
 117
 118         f4m_node = body.find(_x('smil:seq//smil:video'))
 119         if f4m_node is not None and '.f4m' in f4m_node.attrib['src']:
 120             f4m_url = f4m_node.attrib['src']
 121             if 'manifest.f4m?' not in f4m_url:
 122                 f4m_url += '?'
 123             # the parameters are from syfy.com, other sites may use others,
 124             # they also work for nbc.com
 125             f4m_url += '&g=UXWGVKRWHFSP&hdcore=3.0.3'
 126             formats = self._extract_f4m_formats(f4m_url, video_id)
 127         else:
 128             formats = []
 129             switch = body.find(_x('smil:switch'))
 130             if switch is not None:
 131                 base_url = head.find(_x('smil:meta')).attrib['base']
 132                 for f in switch.findall(_x('smil:video')):
 133                     attr = f.attrib
 134                     width = int(attr['width'])
 135                     height = int(attr['height'])
 136                     vbr = int(attr['system-bitrate']) // 1000
 137                     format_id = '%dx%d_%dk' % (width, height, vbr)
 138                     formats.append({
 139                         'format_id': format_id,
 140                         'url': base_url,
 141                         'play_path': 'mp4:' + attr['src'],
 142                         'ext': 'flv',
 143                         'width': width,
 144                         'height': height,
 145                         'vbr': vbr,
 146                     })
 147             else:
 148                 switch = body.find(_x('smil:seq//smil:switch'))
 149                 for f in switch.findall(_x('smil:video')):
 150                     attr = f.attrib
 151                     vbr = int(attr['system-bitrate']) // 1000
 152                     ext = determine_ext(attr['src'])
 153                     if ext == 'once':
 154                         ext = 'mp4'
 155                     formats.append({
 156                         'format_id': compat_str(vbr),
 157                         'url': attr['src'],
 158                         'vbr': vbr,
 159                         'ext': ext,
 160                     })
 161             self._sort_formats(formats)
 162
 163         return {
 164             'id': video_id,
 165             'title': info['title'],
 166             'subtitles': subtitles,
 167             'formats': formats,
 168             'description': info['description'],
 169             'thumbnail': info['defaultThumbnailUrl'],
 170             'duration': info['duration'] // 1000,
 171         }