Update app.py

app.py

@@ -157,20 +157,35 @@ def process_audio_to_text(audio_path, inputlanguage="English", outputlanguage="E
 def process_text_to_audio(text, translatefrom="English", translateto="English", filename_prefix="audio"):
     """
     Convert text input to audio using the Gradio client.
-    Ensure the audio file is correctly saved and returned as a file path.
+    Ensure the audio file is correctly saved and returned as a file path or binary data.
     """
-    # Generate audio from text
-    audio_response = audio_client.predict(
-        text,
-        translatefrom,  
-        translateto, 
-        api_name="/t2st"
-    )
-    text_hash = hashlib.md5(text.encode('utf-8')).hexdigest()
-    filename = f"{filename_prefix}_{text_hash}.wav"
-    audio_file_path = save_audio_data_to_file(audio_response[0], filename=filename)
-    
-    return audio_response[0], audio_file_path
+    try:
+        # Generate audio from text
+        audio_response = audio_client.predict(
+            text,
+            translatefrom,  
+            translateto, 
+            api_name="/t2st"
+        )
+        if "error" in audio_response:
+            raise ValueError(f"API Error: {audio_response['error']}")
+
+        # Assuming audio_response[0] is a URL or file path to the generated audio
+        audio_url = audio_response[0]
+        response = requests.get(audio_url)
+        audio_data = response.content  # This should be binary data
+
+        # Generate a unique filename based on the text's hash
+        text_hash = hashlib.md5(text.encode('utf-8')).hexdigest()
+        filename = f"{filename_prefix}_{text_hash}.wav"
+        # Save the audio data to a new file
+        new_audio_file_path = save_audio_data_to_file(audio_data, filename=filename)
+
+        # Return the path to the saved audio file
+        return new_audio_file_path
+    except Exception as e:
+        print(f"Error processing text to audio: {e}")
+        return None
 
 def save_audio_data_to_file(audio_data, directory="audio_files", filename="output_audio.wav"):
     """
@@ -328,17 +343,17 @@ outputs = [
 ]
 
 def update_outputs(inputlanguage, target_language, audio, image, text, file):
-    processed_text, audio_output, top_phrases, translations, audio_outputs = process_input(
+    processed_text, audio_output_path, top_phrases, translations, audio_outputs = process_input(
         image=image, file=file, audio=audio, text=text, 
         translateto=target_language, translatefrom=inputlanguage 
     )
-    audio_output = gr.Audio(file_path=audio_output_path)
-    audio_outputs_components = [(gr.Audio(file_path=ao[0]), gr.Audio(file_path=ao[1])) for ao in audio_outputs]
-    output_tuple = (processed_text, audio_output)
+
+    audio_outputs_components = [(ao[0], ao[1]) for ao in audio_outputs]  
+    output_tuple = (processed_text, audio_output_path) 
     for i in range(len(top_phrases)):
         output_tuple += (top_phrases[i], translations[i]) + audio_outputs_components[i]
     while len(output_tuple) < 14: 
-        output_tuple += ("", "", gr.Audio(), gr.Audio())
+        output_tuple += ("", "", "", "")
     return output_tuple
 
 def interface_func(inputlanguage, target_language, audio, image, text, file):