X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/575f75429e2e479957d006a26940f4a5a51347ce..27375a0e586f091b3366c9963a15304fee569a05:/youtube_dl/extractor/mtv.py

diff --git a/youtube_dl/extractor/mtv.py b/youtube_dl/extractor/mtv.py
index 8acea14..7a3b57a 100644
--- a/youtube_dl/extractor/mtv.py
+++ b/youtube_dl/extractor/mtv.py
@@ -50,8 +50,7 @@ class MTVServicesInfoExtractor(InfoExtractor):
         thumb_node = itemdoc.find(search_path)
         if thumb_node is None:
             return None
-        else:
-            return thumb_node.attrib['url']
+        return thumb_node.get('url') or thumb_node.text or None
 
     def _extract_mobile_video_formats(self, mtvn_id):
         webpage_url = self._MOBILE_TEMPLATE % mtvn_id
@@ -83,7 +82,7 @@ class MTVServicesInfoExtractor(InfoExtractor):
                 hls_url = rendition.find('./src').text
                 formats.extend(self._extract_m3u8_formats(
                     hls_url, video_id, ext='mp4', entry_protocol='m3u8_native',
-                    m3u8_id='hls'))
+                    m3u8_id='hls', fatal=False))
             else:
                 # fms
                 try:
@@ -106,7 +105,8 @@ class MTVServicesInfoExtractor(InfoExtractor):
                     }])
                 except (KeyError, TypeError):
                     raise ExtractorError('Invalid rendition field.')
-        self._sort_formats(formats)
+        if formats:
+            self._sort_formats(formats)
         return formats
 
     def _extract_subtitles(self, mdoc, mtvn_id):
@@ -115,10 +115,17 @@ class MTVServicesInfoExtractor(InfoExtractor):
             if transcript.get('kind') != 'captions':
                 continue
             lang = transcript.get('srclang')
-            subtitles[lang] = [{
-                'url': compat_str(typographic.get('src')),
-                'ext': typographic.get('format')
-            } for typographic in transcript.findall('./typographic')]
+            for typographic in transcript.findall('./typographic'):
+                sub_src = typographic.get('src')
+                if not sub_src:
+                    continue
+                ext = typographic.get('format')
+                if ext == 'cea-608':
+                    ext = 'scc'
+                subtitles.setdefault(lang, []).append({
+                    'url': compat_str(sub_src),
+                    'ext': ext
+                })
         return subtitles
 
     def _get_video_info(self, itemdoc, use_hls=True):
@@ -133,8 +140,11 @@ class MTVServicesInfoExtractor(InfoExtractor):
             mediagen_url += 'acceptMethods='
             mediagen_url += 'hls' if use_hls else 'fms'
 
-        mediagen_doc = self._download_xml(mediagen_url, video_id,
-                                          'Downloading video urls')
+        mediagen_doc = self._download_xml(
+            mediagen_url, video_id, 'Downloading video urls', fatal=False)
+
+        if mediagen_doc is False:
+            return None
 
         item = mediagen_doc.find('./video/item')
         if item is not None and item.get('type') == 'text':
@@ -174,6 +184,13 @@ class MTVServicesInfoExtractor(InfoExtractor):
 
         formats = self._extract_video_formats(mediagen_doc, mtvn_id, video_id)
 
+        # Some parts of complete video may be missing (e.g. missing Act 3 in
+        # http://www.southpark.de/alle-episoden/s14e01-sexual-healing)
+        if not formats:
+            return None
+
+        self._sort_formats(formats)
+
         return {
             'title': title,
             'formats': formats,
@@ -205,9 +222,14 @@ class MTVServicesInfoExtractor(InfoExtractor):
         title = xpath_text(idoc, './channel/title')
         description = xpath_text(idoc, './channel/description')
 
+        entries = []
+        for item in idoc.findall('.//item'):
+            info = self._get_video_info(item, use_hls)
+            if info:
+                entries.append(info)
+
         return self.playlist_result(
-            [self._get_video_info(item, use_hls) for item in idoc.findall('.//item')],
-            playlist_title=title, playlist_description=description)
+            entries, playlist_title=title, playlist_description=description)
 
     def _extract_triforce_mgid(self, webpage, data_zone=None, video_id=None):
         triforce_feed = self._parse_json(self._search_regex(
@@ -243,7 +265,7 @@ class MTVServicesInfoExtractor(InfoExtractor):
 
         if mgid is None or ':' not in mgid:
             mgid = self._search_regex(
-                [r'data-mgid="(.*?)"', r'swfobject.embedSWF\(".*?(mgid:.*?)"'],
+                [r'data-mgid="(.*?)"', r'swfobject\.embedSWF\(".*?(mgid:.*?)"'],
                 webpage, 'mgid', default=None)
 
         if not mgid: