Update app.py

app.py

@@ -1,4 +1,4 @@
-from fastapi import FastAPI, Form
+from fastapi import FastAPI, Form, HTTPException
 from fastapi.responses import JSONResponse, FileResponse
 import uvicorn
 from pydantic import BaseModel
@@ -9,6 +9,11 @@ import base64
 from asr import transcribe, ASR_LANGUAGES
 from tts import synthesize, TTS_LANGUAGES
 from lid import identify
+import logging
+
+# Configure logging
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
 
 app = FastAPI(title="MMS: Scaling Speech Technology to 1000+ languages")
 
@@ -23,39 +28,59 @@ class AudioRequest(BaseModel):
 
 @app.post("/transcribe")
 async def transcribe_audio(request: AudioRequest):
-    audio_bytes = base64.b64decode(request.audio)
-    audio_array, sample_rate = sf.read(io.BytesIO(audio_bytes))
-    
-    result = transcribe(audio_array, request.language)
-    return JSONResponse(content={"transcription": result})
+    try:
+        audio_bytes = base64.b64decode(request.audio)
+        audio_array, sample_rate = sf.read(io.BytesIO(audio_bytes))
+        
+        result = transcribe(audio_array, request.language)
+        return JSONResponse(content={"transcription": result})
+    except Exception as e:
+        logger.error(f"Error in transcribe_audio: {str(e)}")
+        raise HTTPException(status_code=500, detail=f"An error occurred: {str(e)}")
 
 @app.post("/synthesize")
 async def synthesize_speech(request: TTSRequest):
-    audio, filtered_text = synthesize(request.text, request.language, request.speed)
-    
-    # Convert numpy array to bytes
-    buffer = io.BytesIO()
-    sf.write(buffer, audio, 22050, format='wav')
-    buffer.seek(0)
-    
-    return FileResponse(
-        buffer, 
-        media_type="audio/wav", 
-        headers={"Content-Disposition": "attachment; filename=synthesized_audio.wav"}
-    )
+    try:
+        audio, filtered_text = synthesize(request.text, request.language, request.speed)
+        
+        # Convert numpy array to bytes
+        buffer = io.BytesIO()
+        sf.write(buffer, audio, 22050, format='wav')
+        buffer.seek(0)
+        
+        return FileResponse(
+            buffer, 
+            media_type="audio/wav", 
+            headers={"Content-Disposition": "attachment; filename=synthesized_audio.wav"}
+        )
+    except Exception as e:
+        logger.error(f"Error in synthesize_speech: {str(e)}")
+        raise HTTPException(status_code=500, detail=f"An error occurred: {str(e)}")
 
 @app.post("/identify")
 async def identify_language(request: AudioRequest):
-    audio_bytes = base64.b64decode(request.audio)
-    audio_array, sample_rate = sf.read(io.BytesIO(audio_bytes))
-    
-    result = identify(audio_array)
-    return JSONResponse(content={"language_identification": result})
+    try:
+        audio_bytes = base64.b64decode(request.audio)
+        audio_array, sample_rate = sf.read(io.BytesIO(audio_bytes))
+        
+        result = identify(audio_array)
+        return JSONResponse(content={"language_identification": result})
+    except Exception as e:
+        logger.error(f"Error in identify_language: {str(e)}")
+        raise HTTPException(status_code=500, detail=f"An error occurred: {str(e)}")
 
 @app.get("/asr_languages")
 async def get_asr_languages():
-    return JSONResponse(content=ASR_LANGUAGES)
+    try:
+        return JSONResponse(content=ASR_LANGUAGES)
+    except Exception as e:
+        logger.error(f"Error in get_asr_languages: {str(e)}")
+        raise HTTPException(status_code=500, detail=f"An error occurred: {str(e)}")
 
 @app.get("/tts_languages")
 async def get_tts_languages():
-    return JSONResponse(content=TTS_LANGUAGES)
+    try:
+        return JSONResponse(content=TTS_LANGUAGES)
+    except Exception as e:
+        logger.error(f"Error in get_tts_languages: {str(e)}")
+        raise HTTPException(status_code=500, detail=f"An error occurred: {str(e)}")