Update app.py

app.py

@@ -47,6 +47,14 @@ import webrtcvad
 from pydub import AudioSegment
 from pydub.silence import split_on_silence
 import soundfile as sf
+import langcodes
+
+# ISO 639-3 → ISO 639-1
+def iso_639_3_to_1(code3):
+    try:
+        return langcodes.Language.get(code3).language
+    except:
+        return 'en'
 
 logger = logging.getLogger(__name__)
 
@@ -380,7 +388,7 @@ def transcribe_segments_with_scribe(full_audio_path, segments):
                         logger.info(f"Segment {i+1} discarded: cleaned text is empty.")
 
                     if "language_code" in scribe_result and detected_language == "unknown":
-                        detected_language = scribe_result["language_code"]
+                        detected_language = iso_639_3_to_1(scribe_result["language_code"])
 
             except Exception as e:
                 logger.error(f"Error processing segment {i+1}: {e}")
@@ -1509,4 +1517,4 @@ def build_interface():
 tts_model = None
 # Launch the Gradio interface
 demo = build_interface()
-demo.launch()
+demo.launch()