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