youtube_dl/extractor/crackle.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals, division
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..utils import int_or_none
   8
   9
  10 class CrackleIE(InfoExtractor):
  11     _VALID_URL = r'(?:crackle:|https?://(?:www\.)?crackle\.com/(?:playlist/\d+/|(?:[^/]+/)+))(?P<id>\d+)'
  12     _TEST = {
  13         'url': 'http://www.crackle.com/comedians-in-cars-getting-coffee/2498934',
  14         'info_dict': {
  15             'id': '2498934',
  16             'ext': 'mp4',
  17             'title': 'Everybody Respects A Bloody Nose',
  18             'description': 'Jerry is kaffeeklatsching in L.A. with funnyman J.B. Smoove (Saturday Night Live, Real Husbands of Hollywood). They’re headed for brew at 10 Speed Coffee in a 1964 Studebaker Avanti.',
  19             'thumbnail': 're:^https?://.*\.jpg',
  20             'duration': 906,
  21             'series': 'Comedians In Cars Getting Coffee',
  22             'season_number': 8,
  23             'episode_number': 4,
  24             'subtitles': {
  25                 'en-US': [{
  26                     'ext': 'ttml',
  27                 }]
  28             },
  29         },
  30         'params': {
  31             # m3u8 download
  32             'skip_download': True,
  33         }
  34     }
  35
  36     # extracted from http://legacyweb-us.crackle.com/flash/ReferrerRedirect.ashx
  37     _MEDIA_FILE_SLOTS = {
  38         'c544.flv': {
  39             'width': 544,
  40             'height': 306,
  41         },
  42         '360p.mp4': {
  43             'width': 640,
  44             'height': 360,
  45         },
  46         '480p.mp4': {
  47             'width': 852,
  48             'height': 478,
  49         },
  50         '480p_1mbps.mp4': {
  51             'width': 852,
  52             'height': 478,
  53         },
  54     }
  55
  56     def _real_extract(self, url):
  57         video_id = self._match_id(url)
  58
  59         config_doc = self._download_xml(
  60             'http://legacyweb-us.crackle.com/flash/QueryReferrer.ashx?site=16',
  61             video_id, 'Downloading config')
  62
  63         item = self._download_xml(
  64             'http://legacyweb-us.crackle.com/app/revamp/vidwallcache.aspx?flags=-1&fm=%s' % video_id,
  65             video_id).find('i')
  66         title = item.attrib['t']
  67
  68         subtitles = {}
  69         formats = self._extract_m3u8_formats(
  70             'http://content.uplynk.com/ext/%s/%s.m3u8' % (config_doc.attrib['strUplynkOwnerId'], video_id),
  71             video_id, 'mp4', m3u8_id='hls', fatal=None)
  72         path = item.attrib.get('p')
  73         if path:
  74             http_base_url = 'http://ahttp.crackle.com/' + path
  75             for mfs_path, mfs_info in self._MEDIA_FILE_SLOTS.items():
  76                 formats.append({
  77                     'url': http_base_url + mfs_path,
  78                     'format_id': 'http-' + mfs_path.split('.')[0],
  79                     'width': mfs_info['width'],
  80                     'height': mfs_info['height'],
  81                 })
  82             for cc in item.findall('cc'):
  83                 locale = cc.attrib.get('l')
  84                 v = cc.attrib.get('v')
  85                 if locale and v:
  86                     if locale not in subtitles:
  87                         subtitles[locale] = []
  88                     subtitles[locale] = [{
  89                         'url': '%s/%s%s_%s.xml' % (config_doc.attrib['strSubtitleServer'], path, locale, v),
  90                         'ext': 'ttml',
  91                     }]
  92         self._sort_formats(formats, ('width', 'height', 'tbr', 'format_id'))
  93
  94         media_details = self._download_json(
  95             'https://web-api-us.crackle.com/Service.svc/details/media/%s/TW?format=json' % video_id,
  96             video_id, fatal=False)
  97         thumbnails = []
  98         if media_details:
  99             for key, value in media_details.items():
 100                 mobj = re.match('^Thumbnail_(\d+)x(\d+)$', key)
 101                 if mobj:
 102                     width, height = list(map(int, mobj.groups()))
 103                     thumbnails.append({
 104                         'id': '%dp' % height,
 105                         'url': value,
 106                         'width': width,
 107                         'height': height,
 108                     })
 109
 110         return {
 111             'id': video_id,
 112             'title': title,
 113             'description': item.attrib.get('d'),
 114             'duration': int(item.attrib.get('r'), 16) / 1000 if item.attrib.get('r') else None,
 115             'series': item.attrib.get('sn'),
 116             'season_number': int_or_none(item.attrib.get('se')),
 117             'episode_number': int_or_none(item.attrib.get('ep')),
 118             'thumbnails': thumbnails,
 119             'subtitles': subtitles,
 120             'formats': formats,
 121         }