_ 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     compat_HTTPError,
  10 )
  11 from ..utils import (
  12     ExtractorError,
  13     find_xpath_attr,
  14 )
  15
  16
  17 class NBCIE(InfoExtractor):
  18     _VALID_URL = r'http://www\.nbc\.com/(?:[^/]+/)+(?P<id>n?\d+)'
  19
  20     _TESTS = [
  21         {
  22             'url': 'http://www.nbc.com/chicago-fire/video/i-am-a-firefighter/2734188',
  23             # md5 checksum is not stable
  24             'info_dict': {
  25                 'id': 'bTmnLCvIbaaH',
  26                 'ext': 'flv',
  27                 'title': 'I Am a Firefighter',
  28                 'description': 'An emergency puts Dawson\'sf irefighter skills to the ultimate test in this four-part digital series.',
  29             },
  30         },
  31         {
  32             'url': 'http://www.nbc.com/the-tonight-show/episodes/176',
  33             'info_dict': {
  34                 'id': 'XwU9KZkp98TH',
  35                 'ext': 'flv',
  36                 'title': 'Ricky Gervais, Steven Van Zandt, ILoveMakonnen',
  37                 'description': 'A brand new episode of The Tonight Show welcomes Ricky Gervais, Steven Van Zandt and ILoveMakonnen.',
  38             },
  39             'skip': 'Only works from US',
  40         },
  41     ]
  42
  43     def _real_extract(self, url):
  44         video_id = self._match_id(url)
  45         webpage = self._download_webpage(url, video_id)
  46         theplatform_url = self._search_regex(
  47             '(?:class="video-player video-player-full" data-mpx-url|class="player" src)="(.*?)"',
  48             webpage, 'theplatform url').replace('_no_endcard', '')
  49         if theplatform_url.startswith('//'):
  50             theplatform_url = 'http:' + theplatform_url
  51         return self.url_result(theplatform_url)
  52
  53
  54 class NBCNewsIE(InfoExtractor):
  55     _VALID_URL = r'''(?x)https?://www\.nbcnews\.com/
  56         ((video/.+?/(?P<id>\d+))|
  57         (feature/[^/]+/(?P<title>.+)))
  58         '''
  59
  60     _TESTS = [
  61         {
  62             'url': 'http://www.nbcnews.com/video/nbc-news/52753292',
  63             'md5': '47abaac93c6eaf9ad37ee6c4463a5179',
  64             'info_dict': {
  65                 'id': '52753292',
  66                 'ext': 'flv',
  67                 'title': 'Crew emerges after four-month Mars food study',
  68                 'description': 'md5:24e632ffac72b35f8b67a12d1b6ddfc1',
  69             },
  70         },
  71         {
  72             'url': 'http://www.nbcnews.com/feature/edward-snowden-interview/how-twitter-reacted-snowden-interview-n117236',
  73             'md5': 'b2421750c9f260783721d898f4c42063',
  74             'info_dict': {
  75                 'id': 'I1wpAI_zmhsQ',
  76                 'ext': 'mp4',
  77                 'title': 'How Twitter Reacted To The Snowden Interview',
  78                 'description': 'md5:65a0bd5d76fe114f3c2727aa3a81fe64',
  79             },
  80             'add_ie': ['ThePlatform'],
  81         },
  82         {
  83             'url': 'http://www.nbcnews.com/feature/dateline-full-episodes/full-episode-family-business-n285156',
  84             'md5': 'fdbf39ab73a72df5896b6234ff98518a',
  85             'info_dict': {
  86                 'id': 'Wjf9EDR3A_60',
  87                 'ext': 'mp4',
  88                 'title': 'FULL EPISODE: Family Business',
  89                 'description': 'md5:757988edbaae9d7be1d585eb5d55cc04',
  90             },
  91         },
  92     ]
  93
  94     def _real_extract(self, url):
  95         mobj = re.match(self._VALID_URL, url)
  96         video_id = mobj.group('id')
  97         if video_id is not None:
  98             all_info = self._download_xml('http://www.nbcnews.com/id/%s/displaymode/1219' % video_id, video_id)
  99             info = all_info.find('video')
 100
 101             return {
 102                 'id': video_id,
 103                 'title': info.find('headline').text,
 104                 'ext': 'flv',
 105                 'url': find_xpath_attr(info, 'media', 'type', 'flashVideo').text,
 106                 'description': compat_str(info.find('caption').text),
 107                 'thumbnail': find_xpath_attr(info, 'media', 'type', 'thumbnail').text,
 108             }
 109         else:
 110             # "feature" pages use theplatform.com
 111             title = mobj.group('title')
 112             webpage = self._download_webpage(url, title)
 113             bootstrap_json = self._search_regex(
 114                 r'var bootstrapJson = ({.+})\s*$', webpage, 'bootstrap json',
 115                 flags=re.MULTILINE)
 116             bootstrap = json.loads(bootstrap_json)
 117             info = bootstrap['results'][0]['video']
 118             mpxid = info['mpxId']
 119
 120             base_urls = [
 121                 info['fallbackPlaylistUrl'],
 122                 info['associatedPlaylistUrl'],
 123             ]
 124
 125             for base_url in base_urls:
 126                 if not base_url:
 127                     continue
 128                 playlist_url = base_url + '?form=MPXNBCNewsAPI'
 129
 130                 try:
 131                     all_videos = self._download_json(playlist_url, title)
 132                 except ExtractorError as ee:
 133                     if isinstance(ee.cause, compat_HTTPError):
 134                         continue
 135                     raise
 136
 137                 if not all_videos or 'videos' not in all_videos:
 138                     continue
 139
 140                 try:
 141                     info = next(v for v in all_videos['videos'] if v['mpxId'] == mpxid)
 142                     break
 143                 except StopIteration:
 144                     continue
 145
 146             if info is None:
 147                 raise ExtractorError('Could not find video in playlists')
 148
 149             return {
 150                 '_type': 'url',
 151                 # We get the best quality video
 152                 'url': info['videoAssets'][-1]['publicUrl'],
 153                 'ie_key': 'ThePlatform',
 154             }