Fix BabelLanguage to handle both 2-letter and 3-letter codes

The babelfish library requires different constructors for different
ISO language code formats:
- fromalpha2() for 2-letter codes (e.g., 'en')
- direct constructor for 3-letter codes (e.g., 'eng')

This fixes the OCR conversion error when language codes are in
different formats.

Author: mikeSGman

fastflix/widgets/background_tasks.py

@@ -242,11 +242,20 @@ def _convert_sup_to_srt(self, sup_filepath: str) -> bool:
             sup_path = Path(sup_filepath)
             media = Mkv(sup_filepath)
 
-            # Configure options for pgsrip - use 3-letter language code (e.g., "eng", "spa", "fra")
-            # BabelLanguage expects 3-letter ISO 639-2 codes
+            # Configure options for pgsrip
+            # BabelLanguage needs different constructors for 2-letter vs 3-letter codes
             try:
+                # Detect if language code is 2-letter or 3-letter
+                if len(self.language) == 2:
+                    babel_lang = BabelLanguage.fromalpha2(self.language)
+                elif len(self.language) == 3:
+                    babel_lang = BabelLanguage(self.language)
+                else:
+                    # Try as language name
+                    babel_lang = BabelLanguage.fromname(self.language)
+
                 options = Options(
-                    languages={BabelLanguage(self.language)},  # Use 3-letter code directly
+                    languages={babel_lang},
                     overwrite=True,  # Overwrite existing .srt files
                 )
             except Exception: