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 (
   8     compat_str,
   9     ExtractorError,
  10     find_xpath_attr,
  11 )
  12
  13
  14 class NBCIE(InfoExtractor):
  15     _VALID_URL = r'http://www\.nbc\.com/[^/]+/video/[^/]+/(?P<id>n?\d+)'
  16
  17     _TEST = {
  18         'url': 'http://www.nbc.com/chicago-fire/video/i-am-a-firefighter/2734188',
  19         'md5': '54d0fbc33e0b853a65d7b4de5c06d64e',
  20         'info_dict': {
  21             'id': 'u1RInQZRN7QJ',
  22             'ext': 'flv',
  23             'title': 'I Am a Firefighter',
  24             'description': 'An emergency puts Dawson\'sf irefighter skills to the ultimate test in this four-part digital series.',
  25         },
  26     }
  27
  28     def _real_extract(self, url):
  29         mobj = re.match(self._VALID_URL, url)
  30         video_id = mobj.group('id')
  31         webpage = self._download_webpage(url, video_id)
  32         theplatform_url = self._search_regex('class="video-player video-player-full" data-mpx-url="(.*?)"', webpage, 'theplatform url')
  33         if theplatform_url.startswith('//'):
  34             theplatform_url = 'http:' + theplatform_url
  35         return self.url_result(theplatform_url)
  36
  37
  38 class NBCNewsIE(InfoExtractor):
  39     _VALID_URL = r'''(?x)https?://www\.nbcnews\.com/
  40         ((video/.+?/(?P<id>\d+))|
  41         (feature/[^/]+/(?P<title>.+)))
  42         '''
  43
  44     _TESTS = [
  45         {
  46             'url': 'http://www.nbcnews.com/video/nbc-news/52753292',
  47             'md5': '47abaac93c6eaf9ad37ee6c4463a5179',
  48             'info_dict': {
  49                 'id': '52753292',
  50                 'ext': 'flv',
  51                 'title': 'Crew emerges after four-month Mars food study',
  52                 'description': 'md5:24e632ffac72b35f8b67a12d1b6ddfc1',
  53             },
  54         },
  55         {
  56             'url': 'http://www.nbcnews.com/feature/edward-snowden-interview/how-twitter-reacted-snowden-interview-n117236',
  57             'md5': 'b2421750c9f260783721d898f4c42063',
  58             'info_dict': {
  59                 'id': 'I1wpAI_zmhsQ',
  60                 'ext': 'flv',
  61                 'title': 'How Twitter Reacted To The Snowden Interview',
  62                 'description': 'md5:65a0bd5d76fe114f3c2727aa3a81fe64',
  63             },
  64             'add_ie': ['ThePlatform'],
  65         },
  66     ]
  67
  68     def _real_extract(self, url):
  69         mobj = re.match(self._VALID_URL, url)
  70         video_id = mobj.group('id')
  71         if video_id is not None:
  72             all_info = self._download_xml('http://www.nbcnews.com/id/%s/displaymode/1219' % video_id, video_id)
  73             info = all_info.find('video')
  74
  75             return {
  76                 'id': video_id,
  77                 'title': info.find('headline').text,
  78                 'ext': 'flv',
  79                 'url': find_xpath_attr(info, 'media', 'type', 'flashVideo').text,
  80                 'description': compat_str(info.find('caption').text),
  81                 'thumbnail': find_xpath_attr(info, 'media', 'type', 'thumbnail').text,
  82             }
  83         else:
  84             # "feature" pages use theplatform.com
  85             title = mobj.group('title')
  86             webpage = self._download_webpage(url, title)
  87             bootstrap_json = self._search_regex(
  88                 r'var bootstrapJson = ({.+})\s*$', webpage, 'bootstrap json',
  89                 flags=re.MULTILINE)
  90             bootstrap = json.loads(bootstrap_json)
  91             info = bootstrap['results'][0]['video']
  92             mpxid = info['mpxId']
  93
  94             base_urls = [
  95                 info['fallbackPlaylistUrl'],
  96                 info['associatedPlaylistUrl'],
  97             ]
  98
  99             for base_url in base_urls:
 100                 playlist_url = base_url + '?form=MPXNBCNewsAPI'
 101                 all_videos = self._download_json(playlist_url, title)['videos']
 102
 103                 try:
 104                     info = next(v for v in all_videos if v['mpxId'] == mpxid)
 105                     break
 106                 except StopIteration:
 107                     continue
 108
 109             if info is None:
 110                 raise ExtractorError('Could not find video in playlists')
 111
 112             return {
 113                 '_type': 'url',
 114                 # We get the best quality video
 115                 'url': info['videoAssets'][-1]['publicUrl'],
 116                 'ie_key': 'ThePlatform',
 117             }