process audio logic

app.py

@@ -151,18 +151,43 @@ def process_audio_to_text(audio_path, inputlanguage="English", outputlanguage="E
     print("Audio Result: ", result)
     return result[0]
 
-def process_text_to_audio(text, translatefrom="English", translateto="English"):
+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.
     """
-    audio_client = Client(SEAMLESSM4T)
-    result = audio_client.predict(
+    # Generate audio from text
+    audio_response = audio_client.predict(
         text,
         translatefrom,  
         translateto, 
         api_name="/t2st"
     )
-    return result[0] 
+    filename = f"{filename_prefix}_{hash(text)}.wav"
+    audio_file_path = save_audio_data_to_file(audio_response, filename=filename)
+    return audio_file_path
+
+def save_audio_data_to_file(audio_data, directory="audio_files", filename="output_audio.wav"):
+    """
+    Save audio data to a file and return the file path.
+    """
+    Path(directory).mkdir(parents=True, exist_ok=True)
+    file_path = os.path.join(directory, filename)
+    with open(file_path, 'wb') as file:
+        file.write(audio_data)
+    return file_path
+
+# Ensure the function that reads the audio file checks if the path is a file
+def read_audio_file(file_path):
+    """
+    Read and return the audio file content if the path is a file.
+    """
+    if os.path.isfile(file_path):
+        with open(file_path, 'rb') as file:
+            return file.read()
+    else:
+        raise ValueError(f"Expected a file path, got a directory: {file_path}")
+
 
 def initialize_ocr_models():
     """
@@ -221,11 +246,11 @@ def process_input(image=None, file=None, audio=None, text="", translateto = "Eng
         audio_text = long_audio_processor.process_long_audio(audio, inputlanguage=translatefrom, outputlanguage=translateto)
         final_text += "\n" + audio_text
 
-    final_text_with_producetext = final_text + producetext
+    final_text_with_producetext = final_text + producetext.format(target_language=translateto)
 
     response = co.generate(
         model='c4ai-aya',
-        prompt=final_text_with_producetext.format(target_language=translateto),
+        prompt=final_text_with_producetext,
         max_tokens=1024,
         temperature=0.5
     )