[swrmediathek] improve extraction

youtube_dl/extractor/swrmediathek.py

@@ -1,10 +1,12 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
-from ..utils import parse_duration
+from ..utils import (
+    parse_duration,
+    int_or_none,
+    determine_protocol,
+)
 
 
 class SWRMediathekIE(InfoExtractor):
@@ -38,6 +40,7 @@ class SWRMediathekIE(InfoExtractor):
             'uploader': 'SWR Fernsehen',
             'uploader_id': '990030',
         },
+        '_skip': 'redirect to http://swrmediathek.de/index.htm?hinweis=swrlink',
     }, {
         'url': 'http://swrmediathek.de/player.htm?show=bba23e10-cb93-11e3-bf7f-0026b975f2e6',
         'md5': '4382e4ef2c9d7ce6852535fa867a0dd3',
@@ -51,54 +54,62 @@ class SWRMediathekIE(InfoExtractor):
             'upload_date': '20140520',
             'uploader': 'SWR 2',
             'uploader_id': '284670',
-        }
+        },
+        '_skip': 'redirect to http://swrmediathek.de/index.htm?hinweis=swrlink',
     }]
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
+        video_id = self._match_id(url)
 
         video = self._download_json(
-            'http://swrmediathek.de/AjaxEntry?ekey=%s' % video_id, video_id, 'Downloading video JSON')
+            'http://swrmediathek.de/AjaxEntry?ekey=%s' % video_id,
+            video_id, 'Downloading video JSON')
 
         attr = video['attr']
-        media_type = attr['entry_etype']
+        title = attr['entry_title']
+        media_type = attr.get('entry_etype')
 
         formats = []
-        for entry in video['sub']:
-            if entry['name'] != 'entry_media':
+        for entry in video.get('sub', []):
+            if entry.get('name') != 'entry_media':
                 continue
 
-            entry_attr = entry['attr']
-            codec = entry_attr['val0']
-            quality = int(entry_attr['val1'])
-
-            fmt = {
-                'url': entry_attr['val2'],
-                'quality': quality,
-            }
-
-            if media_type == 'Video':
-                fmt.update({
-                    'format_note': ['144p', '288p', '544p', '720p'][quality - 1],
-                    'vcodec': codec,
+            entry_attr = entry.get('attr', {})
+            f_url = entry_attr.get('val2')
+            if not f_url:
+                continue
+            codec = entry_attr.get('val0')
+            if codec == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    f_url, video_id, 'mp4', 'm3u8_native',
+                    m3u8_id='hls', fatal=False))
+            elif codec == 'f4m':
+                formats.extend(self._extract_f4m_formats(
+                    f_url + '?hdcore=3.7.0', video_id,
+                    f4m_id='hds', fatal=False))
+            else:
+                formats.append({
+                    'format_id': determine_protocol({'url': f_url}),
+                    'url': f_url,
+                    'quality': int_or_none(entry_attr.get('val1')),
+                    'vcodec': codec if media_type == 'Video' else 'none',
+                    'acodec': codec if media_type == 'Audio' else None,
                 })
-            elif media_type == 'Audio':
-                fmt.update({
-                    'acodec': codec,
-                })
-            formats.append(fmt)
-
         self._sort_formats(formats)
 
+        upload_date = None
+        entry_pdatet = attr.get('entry_pdatet')
+        if entry_pdatet:
+            upload_date = entry_pdatet[:-4]
+
         return {
             'id': video_id,
-            'title': attr['entry_title'],
-            'description': attr['entry_descl'],
-            'thumbnail': attr['entry_image_16_9'],
-            'duration': parse_duration(attr['entry_durat']),
-            'upload_date': attr['entry_pdatet'][:-4],
-            'uploader': attr['channel_title'],
-            'uploader_id': attr['channel_idkey'],
+            'title': title,
+            'description': attr.get('entry_descl'),
+            'thumbnail': attr.get('entry_image_16_9'),
+            'duration': parse_duration(attr.get('entry_durat')),
+            'upload_date': upload_date,
+            'uploader': attr.get('channel_title'),
+            'uploader_id': attr.get('channel_idkey'),
             'formats': formats,
         }