Update websocket_handler.py

websocket_handler.py

@@ -600,10 +600,6 @@ async def generate_and_stream_tts(
         # Generate audio
         audio_data = await tts_provider.synthesize(text)
         
-        # Debug log to check audio data
-        log_debug(f"Audio data type: {type(audio_data)}, length: {len(audio_data)}")
-        log_debug(f"First 10 bytes: {audio_data[:10]}")
-        
         # Change state to playing
         await session.change_state(ConversationState.PLAYING_AUDIO)
         await websocket.send_json({
@@ -612,46 +608,34 @@ async def generate_and_stream_tts(
             "to": "playing_audio"
         })
         
+        # Convert audio to base64 for transmission
+        audio_base64 = base64.b64encode(audio_data).decode('utf-8')
+        
         # Stream audio in chunks
         chunk_size = session.audio_chunk_size
-        total_chunks = (len(audio_data) + chunk_size - 1) // chunk_size
+        total_length = len(audio_base64)
+        total_chunks = (total_length + chunk_size - 1) // chunk_size
         
-        for i in range(0, len(audio_data), chunk_size):
+        for i in range(0, total_length, chunk_size):
             # Check for cancellation
             if asyncio.current_task().cancelled():
                 break
                 
-            chunk = audio_data[i:i + chunk_size]
+            chunk = audio_base64[i:i + chunk_size]
             chunk_index = i // chunk_size
             
-            # IMPORTANT: Ensure chunk is bytes before encoding
-            if isinstance(chunk, str):
-                # If chunk is already a string, it might be base64 already
-                chunk_base64 = chunk
-            else:
-                # Convert bytes to base64
-                chunk_base64 = base64.b64encode(chunk).decode('utf-8')
-            
             await websocket.send_json({
                 "type": "tts_audio",
-                "data": chunk_base64,
+                "data": chunk,
                 "chunk_index": chunk_index,
                 "total_chunks": total_chunks,
                 "is_last": chunk_index == total_chunks - 1,
-                "mime_type": "audio/mpeg"
+                "mime_type": "audio/mpeg"  # MP3 format için
             })
             
             # Small delay to prevent overwhelming the client
             await asyncio.sleep(0.01)
         
-        # Send state back to idle after completion
-        await session.change_state(ConversationState.IDLE)
-        await websocket.send_json({
-            "type": "state_change",
-            "from": "playing_audio",
-            "to": "idle"
-        })
-        
         log_info(
             f"TTS streaming completed",
             session_id=session.session.session_id,
@@ -666,7 +650,6 @@ async def generate_and_stream_tts(
         log_error(
             f"TTS generation error",
             error=str(e),
-            traceback=traceback.format_exc(),
             session_id=session.session.session_id
         )
         await websocket.send_json({