_ Git - youtube-dl/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 ..compat import (
   8     compat_str,
   9 )
  10 from ..utils import (
  11     ExtractorError,
  12     find_xpath_attr,
  13 )
  14
  15
  16 class NBCIE(InfoExtractor):
  17     _VALID_URL = r'http://www\.nbc\.com/(?:[^/]+/)+(?P<id>n?\d+)'
  18
  19     _TESTS = [
  20         {
  21             'url': 'http://www.nbc.com/chicago-fire/video/i-am-a-firefighter/2734188',
  22             # md5 checksum is not stable
  23             'info_dict': {
  24                 'id': 'bTmnLCvIbaaH',
  25                 'ext': 'flv',
  26                 'title': 'I Am a Firefighter',
  27                 'description': 'An emergency puts Dawson\'sf irefighter skills to the ultimate test in this four-part digital series.',
  28             },
  29         },
  30         {
  31             'url': 'http://www.nbc.com/the-tonight-show/episodes/176',
  32             'info_dict': {
  33                 'id': 'XwU9KZkp98TH',
  34                 'ext': 'flv',
  35                 'title': 'Ricky Gervais, Steven Van Zandt, ILoveMakonnen',
  36                 'description': 'A brand new episode of The Tonight Show welcomes Ricky Gervais, Steven Van Zandt and ILoveMakonnen.',
  37             },
  38             'skip': 'Only works from US',
  39         },
  40     ]
  41
  42     def _real_extract(self, url):
  43         video_id = self._match_id(url)
  44         webpage = self._download_webpage(url, video_id)
  45         theplatform_url = self._search_regex(
  46             '(?:class="video-player video-player-full" data-mpx-url|class="player" src)="(.*?)"',
  47             webpage, 'theplatform url').replace('_no_endcard', '')
  48         if theplatform_url.startswith('//'):
  49             theplatform_url = 'http:' + theplatform_url
  50         return self.url_result(theplatform_url)
  51
  52
  53 class NBCNewsIE(InfoExtractor):
  54     _VALID_URL = r'''(?x)https?://www\.nbcnews\.com/
  55         ((video/.+?/(?P<id>\d+))|
  56         (feature/[^/]+/(?P<title>.+)))
  57         '''
  58
  59     _TESTS = [
  60         {
  61             'url': 'http://www.nbcnews.com/video/nbc-news/52753292',
  62             'md5': '47abaac93c6eaf9ad37ee6c4463a5179',
  63             'info_dict': {
  64                 'id': '52753292',
  65                 'ext': 'flv',
  66                 'title': 'Crew emerges after four-month Mars food study',
  67                 'description': 'md5:24e632ffac72b35f8b67a12d1b6ddfc1',
  68             },
  69         },
  70         {
  71             'url': 'http://www.nbcnews.com/feature/edward-snowden-interview/how-twitter-reacted-snowden-interview-n117236',
  72             'md5': 'b2421750c9f260783721d898f4c42063',
  73             'info_dict': {
  74                 'id': 'I1wpAI_zmhsQ',
  75                 'ext': 'mp4',
  76                 'title': 'How Twitter Reacted To The Snowden Interview',
  77                 'description': 'md5:65a0bd5d76fe114f3c2727aa3a81fe64',
  78             },
  79             'add_ie': ['ThePlatform'],
  80         },
  81     ]
  82
  83     def _real_extract(self, url):
  84         mobj = re.match(self._VALID_URL, url)
  85         video_id = mobj.group('id')
  86         if video_id is not None:
  87             all_info = self._download_xml('http://www.nbcnews.com/id/%s/displaymode/1219' % video_id, video_id)
  88             info = all_info.find('video')
  89
  90             return {
  91                 'id': video_id,
  92                 'title': info.find('headline').text,
  93                 'ext': 'flv',
  94                 'url': find_xpath_attr(info, 'media', 'type', 'flashVideo').text,
  95                 'description': compat_str(info.find('caption').text),
  96                 'thumbnail': find_xpath_attr(info, 'media', 'type', 'thumbnail').text,
  97             }
  98         else:
  99             # "feature" pages use theplatform.com
 100             title = mobj.group('title')
 101             webpage = self._download_webpage(url, title)
 102             bootstrap_json = self._search_regex(
 103                 r'var bootstrapJson = ({.+})\s*$', webpage, 'bootstrap json',
 104                 flags=re.MULTILINE)
 105             bootstrap = json.loads(bootstrap_json)
 106             info = bootstrap['results'][0]['video']
 107             mpxid = info['mpxId']
 108
 109             base_urls = [
 110                 info['fallbackPlaylistUrl'],
 111                 info['associatedPlaylistUrl'],
 112             ]
 113
 114             for base_url in base_urls:
 115                 if not base_url:
 116                     continue
 117                 playlist_url = base_url + '?form=MPXNBCNewsAPI'
 118                 all_videos = self._download_json(playlist_url, title)['videos']
 119
 120                 try:
 121                     info = next(v for v in all_videos if v['mpxId'] == mpxid)
 122                     break
 123                 except StopIteration:
 124                     continue
 125
 126             if info is None:
 127                 raise ExtractorError('Could not find video in playlists')
 128
 129             return {
 130                 '_type': 'url',
 131                 # We get the best quality video
 132                 'url': info['videoAssets'][-1]['publicUrl'],
 133                 'ie_key': 'ThePlatform',
 134             }