[lecturio] Improve subtitles extraction (closes #18488)
authorSergey M․ <dstftw@gmail.com>
Tue, 11 Dec 2018 16:45:02 +0000 (23:45 +0700)
committerSergey M․ <dstftw@gmail.com>
Tue, 11 Dec 2018 16:45:02 +0000 (23:45 +0700)
youtube_dl/extractor/lecturio.py

index 62ff28e028b2443fcbfe74da821daa33481844d9..0f1265cdfc93e5387fa6d7458d0fde3162880980 100644 (file)
@@ -136,9 +136,15 @@ class LecturioIE(LecturioBaseIE):
             cc_url = url_or_none(cc_url)
             if not cc_url:
                 continue
-            sub_dict = automatic_captions if 'auto-translated' in cc_label else subtitles
             lang = self._search_regex(
-                r'/([a-z]{2})_', cc_url, 'lang', default=cc_label.split()[0])
+                r'/([a-z]{2})_', cc_url, 'lang',
+                default=cc_label.split()[0] if cc_label else 'en')
+            original_lang = self._search_regex(
+                r'/[a-z]{2}_([a-z]{2})_', cc_url, 'original lang',
+                default=None)
+            sub_dict = (automatic_captions
+                        if 'auto-translated' in cc_label or original_lang
+                        else subtitles)
             sub_dict.setdefault(self._CC_LANGS.get(lang, lang), []).append({
                 'url': cc_url,
             })