Fix bugs

app.py

@@ -4,7 +4,8 @@ import numpy as np
 from PIL import Image
 import pytesseract
 from gtts import gTTS
-import io
+import tempfile
+import os
 
 def preprocess(image):
     img = np.array(image)
@@ -30,11 +31,10 @@ def extract_text_and_speak(image):
     
     # Create TTS only if text exists
     if text and text != "No readable text found.":
-        tts = gTTS(text)
-        buffer = io.BytesIO()
-        tts.write_to_fp(buffer)
-        buffer.seek(0)
-        audio_file = buffer
+        with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmpfile:
+            tts = gTTS(text)
+            tts.save(tmpfile.name)
+            audio_file = tmpfile.name
     else:
         audio_file = None
         if not text: