[nbcnews] Add support for /feature/* pages (closes #3007)

youtube_dl/extractor/nbc.py

@@ -1,6 +1,7 @@
 from __future__ import unicode_literals
 
 import re
+import json
 
 from .common import InfoExtractor
 from ..utils import find_xpath_attr, compat_str
@@ -31,30 +32,68 @@ class NBCIE(InfoExtractor):
 
 
 class NBCNewsIE(InfoExtractor):
-    _VALID_URL = r'https?://www\.nbcnews\.com/video/.+?/(?P<id>\d+)'
+    _VALID_URL = r'''(?x)https?://www\.nbcnews\.com/
+        ((video/.+?/(?P<id>\d+))|
+        (feature/[^/]+/(?P<title>.+)))
+        '''
 
-    _TEST = {
+    _TESTS = [
-        'url': 'http://www.nbcnews.com/video/nbc-news/52753292',
+        {
-        'md5': '47abaac93c6eaf9ad37ee6c4463a5179',
+            'url': 'http://www.nbcnews.com/video/nbc-news/52753292',
-        'info_dict': {
+            'md5': '47abaac93c6eaf9ad37ee6c4463a5179',
-            'id': '52753292',
+            'info_dict': {
-            'ext': 'flv',
+                'id': '52753292',
-            'title': 'Crew emerges after four-month Mars food study',
+                'ext': 'flv',
-            'description': 'md5:24e632ffac72b35f8b67a12d1b6ddfc1',
+                'title': 'Crew emerges after four-month Mars food study',
+                'description': 'md5:24e632ffac72b35f8b67a12d1b6ddfc1',
+            },
         },
-    }
+        {
+            'url': 'http://www.nbcnews.com/feature/edward-snowden-interview/how-twitter-reacted-snowden-interview-n117236',
+            'md5': 'b2421750c9f260783721d898f4c42063',
+            'info_dict': {
+                'id': 'I1wpAI_zmhsQ',
+                'ext': 'flv',
+                'title': 'How Twitter Reacted To The Snowden Interview',
+                'description': 'md5:65a0bd5d76fe114f3c2727aa3a81fe64',
+            },
+            'add_ie': ['ThePlatform'],
+        },
+    ]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('id')
-        all_info = self._download_xml('http://www.nbcnews.com/id/%s/displaymode/1219' % video_id, video_id)
+        if video_id is not None:
-        info = all_info.find('video')
+            all_info = self._download_xml('http://www.nbcnews.com/id/%s/displaymode/1219' % video_id, video_id)
+            info = all_info.find('video')
 
-        return {
+            return {
-            'id': video_id,
+                'id': video_id,
-            'title': info.find('headline').text,
+                'title': info.find('headline').text,
-            'ext': 'flv',
+                'ext': 'flv',
-            'url': find_xpath_attr(info, 'media', 'type', 'flashVideo').text,
+                'url': find_xpath_attr(info, 'media', 'type', 'flashVideo').text,
-            'description': compat_str(info.find('caption').text),
+                'description': compat_str(info.find('caption').text),
-            'thumbnail': find_xpath_attr(info, 'media', 'type', 'thumbnail').text,
+                'thumbnail': find_xpath_attr(info, 'media', 'type', 'thumbnail').text,
-        }
+            }
+        else:
+            # "feature" pages use theplatform.com
+            title = mobj.group('title')
+            webpage = self._download_webpage(url, title)
+            bootstrap_json = self._search_regex(
+                r'var bootstrapJson = ({.+})\s*$', webpage, 'bootstrap json',
+                flags=re.MULTILINE)
+            bootstrap = json.loads(bootstrap_json)
+            info = bootstrap['results'][0]['video']
+            playlist_url = info['fallbackPlaylistUrl'] + '?form=MPXNBCNewsAPI'
+            mpxid = info['mpxId']
+            all_videos = self._download_json(playlist_url, title)['videos']
+            # The response contains additional videos
+            info = next(v for v in all_videos if v['mpxId'] == mpxid)
+
+            return {
+                '_type': 'url',
+                # We get the best quality video
+                'url': info['videoAssets'][-1]['publicUrl'],
+                'ie_key': 'ThePlatform',
+            }