[srmediathek] improve extraction

youtube_dl/extractor/ard.py

@@ -110,13 +110,19 @@ class ARDMediathekIE(InfoExtractor):
                 server = stream.get('_server')
                 for stream_url in stream_urls:
                     ext = determine_ext(stream_url)
+                    if quality != 'auto' and ext in ('f4m', 'm3u8'):
+                        continue
                     if ext == 'f4m':
-                        formats.extend(self._extract_f4m_formats(
+                        f4m_formats = self._extract_f4m_formats(
                             stream_url + '?hdcore=3.1.1&plugin=aasp-3.1.1.69.124',
-                            video_id, preference=-1, f4m_id='hds'))
+                            video_id, preference=-1, f4m_id='hds', fatal=False)
+                        if f4m_formats:
+                            formats.extend(f4m_formats)
                     elif ext == 'm3u8':
-                        formats.extend(self._extract_m3u8_formats(
+                        m3u8_formats = self._extract_m3u8_formats(
-                            stream_url, video_id, 'mp4', preference=1, m3u8_id='hls'))
+                            stream_url, video_id, 'mp4', preference=1, m3u8_id='hls', fatal=False)
+                        if m3u8_formats:
+                            formats.extend(m3u8_formats)
                     else:
                         if server and server.startswith('rtmp'):
                             f = {

youtube_dl/extractor/srmediathek.py

@@ -1,17 +1,18 @@
 # encoding: utf-8
 from __future__ import unicode_literals
 
-import json
+from .ard import ARDMediathekIE
+from ..utils import (
+    ExtractorError,
+    get_element_by_attribute,
+)
 
-from .common import InfoExtractor
-from ..utils import js_to_json
 
-
+class SRMediathekIE(ARDMediathekIE):
-class SRMediathekIE(InfoExtractor):
     IE_DESC = 'Saarländischer Rundfunk'
     _VALID_URL = r'https?://sr-mediathek\.sr-online\.de/index\.php\?.*?&id=(?P<id>[0-9]+)'
 
-    _TEST = {
+    _TESTS = [{
         'url': 'http://sr-mediathek.sr-online.de/index.php?seite=7&id=28455',
         'info_dict': {
             'id': '28455',
@@ -20,24 +21,36 @@ class SRMediathekIE(InfoExtractor):
             'description': 'Ringen: KSV Köllerbach gegen Aachen-Walheim; Frauen-Fußball: 1. FC Saarbrücken gegen Sindelfingen; Motorsport: Rallye in Losheim; dazu: Interview mit Timo Bernhard; Turnen: TG Saar; Reitsport: Deutscher Voltigier-Pokal; Badminton: Interview mit Michael Fuchs ',
             'thumbnail': 're:^https?://.*\.jpg$',
         },
-    }
+        'skip': 'no longer available',
+    }, {
+        'url': 'http://sr-mediathek.sr-online.de/index.php?seite=7&id=37682',
+        'info_dict': {
+            'id': '37682',
+            'ext': 'mp4',
+            'title': 'Love, Cakes and Rock\'n\'Roll',
+            'description': 'md5:18bf9763631c7d326c22603681e1123d',
+        },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
+        'expected_warnings': ['Unable to download f4m manifest']
+    }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
 
-        murls = json.loads(js_to_json(self._search_regex(
+        if '>Der gew&uuml;nschte Beitrag ist leider nicht mehr verf&uuml;gbar.<' in webpage:
-            r'var mediaURLs\s*=\s*(.*?);\n', webpage, 'video URLs')))
+            raise ExtractorError('Video %s is no longer available' % video_id, expected=True)
-        formats = [{'url': murl} for murl in murls]
-        self._sort_formats(formats)
-
-        title = json.loads(js_to_json(self._search_regex(
-            r'var mediaTitles\s*=\s*(.*?);\n', webpage, 'title')))[0]
 
-        return {
+        media_collection_url = self._search_regex(
+            r'data-mediacollection-ardplayer="([^"]+)"', webpage, 'media collection url')
+        info = self._extract_media_info(media_collection_url, webpage, video_id)
+        info.update({
             'id': video_id,
-            'title': title,
+            'title': get_element_by_attribute('class', 'ardplayer-title', webpage),
-            'formats': formats,
             'description': self._og_search_description(webpage),
             'thumbnail': self._og_search_thumbnail(webpage),
-        }
+        })
+        return info