debian/control: Add recommends on aria2 | wget | curl to use external downloaders.

[youtubedl] / youtube_dl / extractor / nbc.py

diff --git a/youtube_dl/extractor/nbc.py b/youtube_dl/extractor/nbc.py
index aa34665d1669f32ab31a02618c58ef9c4b130fe2..f840f65321997078859ab5f74682969ec4499359 100644 (file)
--- a/youtube_dl/extractor/nbc.py
+++ b/youtube_dl/extractor/nbc.py
@@ -4,28 +4,48 @@ import re
 import json
 
 from .common import InfoExtractor
-from ..utils import find_xpath_attr, compat_str
+from ..compat import (
+    compat_str,
+    compat_HTTPError,
+)
+from ..utils import (
+    ExtractorError,
+    find_xpath_attr,
+)
 
 
 class NBCIE(InfoExtractor):
-    _VALID_URL = r'http://www\.nbc\.com/[^/]+/video/[^/]+/(?P<id>n?\d+)'
-
-    _TEST = {
-        'url': 'http://www.nbc.com/chicago-fire/video/i-am-a-firefighter/2734188',
-        'md5': '54d0fbc33e0b853a65d7b4de5c06d64e',
-        'info_dict': {
-            'id': 'u1RInQZRN7QJ',
-            'ext': 'flv',
-            'title': 'I Am a Firefighter',
-            'description': 'An emergency puts Dawson\'sf irefighter skills to the ultimate test in this four-part digital series.',
+    _VALID_URL = r'http://www\.nbc\.com/(?:[^/]+/)+(?P<id>n?\d+)'
+
+    _TESTS = [
+        {
+            'url': 'http://www.nbc.com/chicago-fire/video/i-am-a-firefighter/2734188',
+            # md5 checksum is not stable
+            'info_dict': {
+                'id': 'bTmnLCvIbaaH',
+                'ext': 'flv',
+                'title': 'I Am a Firefighter',
+                'description': 'An emergency puts Dawson\'sf irefighter skills to the ultimate test in this four-part digital series.',
+            },
         },
-    }
+        {
+            'url': 'http://www.nbc.com/the-tonight-show/episodes/176',
+            'info_dict': {
+                'id': 'XwU9KZkp98TH',
+                'ext': 'flv',
+                'title': 'Ricky Gervais, Steven Van Zandt, ILoveMakonnen',
+                'description': 'A brand new episode of The Tonight Show welcomes Ricky Gervais, Steven Van Zandt and ILoveMakonnen.',
+            },
+            'skip': 'Only works from US',
+        },
+    ]
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
+        video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
-        theplatform_url = self._search_regex('class="video-player video-player-full" data-mpx-url="(.*?)"', webpage, 'theplatform url')
+        theplatform_url = self._search_regex(
+            '(?:class="video-player video-player-full" data-mpx-url|class="player" src)="(.*?)"',
+            webpage, 'theplatform url').replace('_no_endcard', '')
         if theplatform_url.startswith('//'):
             theplatform_url = 'http:' + theplatform_url
         return self.url_result(theplatform_url)
@@ -53,12 +73,22 @@ class NBCNewsIE(InfoExtractor):
             'md5': 'b2421750c9f260783721d898f4c42063',
             'info_dict': {
                 'id': 'I1wpAI_zmhsQ',
-                'ext': 'flv',
+                'ext': 'mp4',
                 'title': 'How Twitter Reacted To The Snowden Interview',
                 'description': 'md5:65a0bd5d76fe114f3c2727aa3a81fe64',
             },
             'add_ie': ['ThePlatform'],
         },
+        {
+            'url': 'http://www.nbcnews.com/feature/dateline-full-episodes/full-episode-family-business-n285156',
+            'md5': 'fdbf39ab73a72df5896b6234ff98518a',
+            'info_dict': {
+                'id': 'Wjf9EDR3A_60',
+                'ext': 'mp4',
+                'title': 'FULL EPISODE: Family Business',
+                'description': 'md5:757988edbaae9d7be1d585eb5d55cc04',
+            },
+        },
     ]
 
     def _real_extract(self, url):
@@ -85,11 +115,36 @@ class NBCNewsIE(InfoExtractor):
                 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)
+
+            base_urls = [
+                info['fallbackPlaylistUrl'],
+                info['associatedPlaylistUrl'],
+            ]
+
+            for base_url in base_urls:
+                if not base_url:
+                    continue
+                playlist_url = base_url + '?form=MPXNBCNewsAPI'
+
+                try:
+                    all_videos = self._download_json(playlist_url, title)
+                except ExtractorError as ee:
+                    if isinstance(ee.cause, compat_HTTPError):
+                        continue
+                    raise
+
+                if not all_videos or 'videos' not in all_videos:
+                    continue
+
+                try:
+                    info = next(v for v in all_videos['videos'] if v['mpxId'] == mpxid)
+                    break
+                except StopIteration:
+                    continue
+
+            if info is None:
+                raise ExtractorError('Could not find video in playlists')
 
             return {
                 '_type': 'url',