Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/nbc.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import json
   5
   6 from .common import InfoExtractor
   7 from ..utils import find_xpath_attr, compat_str
   8
   9
  10 class NBCIE(InfoExtractor):
  11     _VALID_URL = r'http://www\.nbc\.com/[^/]+/video/[^/]+/(?P<id>n?\d+)'
  12
  13     _TEST = {
  14         'url': 'http://www.nbc.com/chicago-fire/video/i-am-a-firefighter/2734188',
  15         'md5': '54d0fbc33e0b853a65d7b4de5c06d64e',
  16         'info_dict': {
  17             'id': 'u1RInQZRN7QJ',
  18             'ext': 'flv',
  19             'title': 'I Am a Firefighter',
  20             'description': 'An emergency puts Dawson\'sf irefighter skills to the ultimate test in this four-part digital series.',
  21         },
  22     }
  23
  24     def _real_extract(self, url):
  25         mobj = re.match(self._VALID_URL, url)
  26         video_id = mobj.group('id')
  27         webpage = self._download_webpage(url, video_id)
  28         theplatform_url = self._search_regex('class="video-player video-player-full" data-mpx-url="(.*?)"', webpage, 'theplatform url')
  29         if theplatform_url.startswith('//'):
  30             theplatform_url = 'http:' + theplatform_url
  31         return self.url_result(theplatform_url)
  32
  33
  34 class NBCNewsIE(InfoExtractor):
  35     _VALID_URL = r'''(?x)https?://www\.nbcnews\.com/
  36         ((video/.+?/(?P<id>\d+))|
  37         (feature/[^/]+/(?P<title>.+)))
  38         '''
  39
  40     _TESTS = [
  41         {
  42             'url': 'http://www.nbcnews.com/video/nbc-news/52753292',
  43             'md5': '47abaac93c6eaf9ad37ee6c4463a5179',
  44             'info_dict': {
  45                 'id': '52753292',
  46                 'ext': 'flv',
  47                 'title': 'Crew emerges after four-month Mars food study',
  48                 'description': 'md5:24e632ffac72b35f8b67a12d1b6ddfc1',
  49             },
  50         },
  51         {
  52             'url': 'http://www.nbcnews.com/feature/edward-snowden-interview/how-twitter-reacted-snowden-interview-n117236',
  53             'md5': 'b2421750c9f260783721d898f4c42063',
  54             'info_dict': {
  55                 'id': 'I1wpAI_zmhsQ',
  56                 'ext': 'flv',
  57                 'title': 'How Twitter Reacted To The Snowden Interview',
  58                 'description': 'md5:65a0bd5d76fe114f3c2727aa3a81fe64',
  59             },
  60             'add_ie': ['ThePlatform'],
  61         },
  62     ]
  63
  64     def _real_extract(self, url):
  65         mobj = re.match(self._VALID_URL, url)
  66         video_id = mobj.group('id')
  67         if video_id is not None:
  68             all_info = self._download_xml('http://www.nbcnews.com/id/%s/displaymode/1219' % video_id, video_id)
  69             info = all_info.find('video')
  70
  71             return {
  72                 'id': video_id,
  73                 'title': info.find('headline').text,
  74                 'ext': 'flv',
  75                 'url': find_xpath_attr(info, 'media', 'type', 'flashVideo').text,
  76                 'description': compat_str(info.find('caption').text),
  77                 'thumbnail': find_xpath_attr(info, 'media', 'type', 'thumbnail').text,
  78             }
  79         else:
  80             # "feature" pages use theplatform.com
  81             title = mobj.group('title')
  82             webpage = self._download_webpage(url, title)
  83             bootstrap_json = self._search_regex(
  84                 r'var bootstrapJson = ({.+})\s*$', webpage, 'bootstrap json',
  85                 flags=re.MULTILINE)
  86             bootstrap = json.loads(bootstrap_json)
  87             info = bootstrap['results'][0]['video']
  88             playlist_url = info['fallbackPlaylistUrl'] + '?form=MPXNBCNewsAPI'
  89             mpxid = info['mpxId']
  90             all_videos = self._download_json(playlist_url, title)['videos']
  91             # The response contains additional videos
  92             info = next(v for v in all_videos if v['mpxId'] == mpxid)
  93
  94             return {
  95                 '_type': 'url',
  96                 # We get the best quality video
  97                 'url': info['videoAssets'][-1]['publicUrl'],
  98                 'ie_key': 'ThePlatform',
  99             }