[npo] Improve formats extraction (Closes #3782)

youtube_dl/extractor/npo.py

@@ -5,6 +5,7 @@ import re
 from .common import InfoExtractor
 from ..utils import (
     unified_strdate,
+    parse_duration,
     qualities,
 )
 
@@ -13,7 +14,8 @@ class NPOIE(InfoExtractor):
     IE_NAME = 'npo.nl'
     _VALID_URL = r'https?://www\.npo\.nl/[^/]+/[^/]+/(?P<id>[^/?]+)'
 
-    _TEST = {
+    _TESTS = [
+        {
             'url': 'http://www.npo.nl/nieuwsuur/22-06-2014/VPWON_1220719',
             'md5': '4b3f9c429157ec4775f2c9cb7b911016',
             'info_dict': {
@@ -23,7 +25,32 @@ class NPOIE(InfoExtractor):
                 'description': 'Dagelijks tussen tien en elf: nieuws, sport en achtergronden.',
                 'upload_date': '20140622',
             },
+        },
+        {
+            'url': 'http://www.npo.nl/de-mega-mike-mega-thomas-show/27-02-2009/VARA_101191800',
+            'md5': 'da50a5787dbfc1603c4ad80f31c5120b',
+            'info_dict': {
+                'id': 'VARA_101191800',
+                'ext': 'm4v',
+                'title': 'De Mega Mike & Mega Thomas show',
+                'description': 'md5:3b74c97fc9d6901d5a665aac0e5400f4',
+                'upload_date': '20090227',
+                'duration': 2400,
+            },
+        },
+        {
+            'url': 'http://www.npo.nl/tegenlicht/25-02-2013/VPWON_1169289',
+            'md5': 'f8065e4e5a7824068ed3c7e783178f2c',
+            'info_dict': {
+                'id': 'VPWON_1169289',
+                'ext': 'm4v',
+                'title': 'Tegenlicht',
+                'description': 'md5:d6476bceb17a8c103c76c3b708f05dd1',
+                'upload_date': '20130225',
+                'duration': 3000,
+            },
         }
+    ]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
@@ -43,19 +70,28 @@ class NPOIE(InfoExtractor):
         token = self._search_regex(r'npoplayer\.token = "(.+?)"', token_page, 'token')
 
         formats = []
-        quality = qualities(['adaptive', 'h264_sb', 'h264_bb', 'h264_std'])
+        quality = qualities(['adaptive', 'wmv_sb', 'h264_sb', 'wmv_bb', 'h264_bb', 'wvc1_std', 'h264_std'])
         for format_id in metadata['pubopties']:
-            streams_info = self._download_json(
+            format_info = self._download_json(
                 'http://ida.omroep.nl/odi/?prid=%s&puboptions=%s&adaptive=yes&token=%s' % (video_id, format_id, token),
-                video_id, 'Downloading %s streams info' % format_id)
+                video_id, 'Downloading %s JSON' % format_id)
-            stream_info = self._download_json(
+            if format_info.get('error_code', 0) or format_info.get('errorcode', 0):
-                streams_info['streams'][0] + '&type=json',
+                continue
-                video_id, 'Downloading %s stream info' % format_id)
+            streams = format_info.get('streams')
+            if streams:
+                video_info = self._download_json(
+                    streams[0] + '&type=json',
+                    video_id, 'Downloading %s stream JSON' % format_id)
+            else:
+                video_info = format_info
+            video_url = video_info.get('url')
+            if not video_url:
+                continue
             if format_id == 'adaptive':
-                formats.extend(self._extract_m3u8_formats(stream_info['url'], video_id))
+                formats.extend(self._extract_m3u8_formats(video_url, video_id))
             else:
                 formats.append({
-                    'url': stream_info['url'],
+                    'url': video_url,
                     'format_id': format_id,
                     'quality': quality(format_id),
                 })
@@ -65,7 +101,8 @@ class NPOIE(InfoExtractor):
             'id': video_id,
             'title': metadata['titel'],
             'description': metadata['info'],
-            'thumbnail': metadata['images'][-1]['url'],
+            'thumbnail': metadata.get('images', [{'url': None}])[-1]['url'],
-            'upload_date': unified_strdate(metadata['gidsdatum']),
+            'upload_date': unified_strdate(metadata.get('gidsdatum')),
+            'duration': parse_duration(metadata.get('tijdsduur')),
             'formats': formats,
         }