Update app.py

app.py

@@ -55,18 +55,19 @@ def translate_text(text, target_language):
     except Exception as e:
         return None, f"Translation error: {str(e)}"
 
-# Function to convert text to speech using Kokoro or gTTS
-def text_to_speech(text, language, tts_engine):
+# Function to convert text to speech using Kokoro or gTTS based on language
+def text_to_speech(text, language):
     try:
-        if tts_engine == "Kokoro" and language in KOKORO_LANGUAGES:
+        # Check if the language is supported by Kokoro
+        if language in KOKORO_LANGUAGES:
             # Use Kokoro TTS
             lang_code = KOKORO_LANGUAGES[language]
             pipeline = KPipeline(lang_code=lang_code)
             generator = pipeline(text, voice="af_heart", speed=1, split_pattern=r'\n+')
             audio_data = None
             for i, (gs, ps, audio) in enumerate(generator):
-                audio_data = audio  # Use the last generated audio segment
-                break  # Only take the first segment for simplicity
+                audio_data = audio  # Use the first segment
+                break
             if audio_data is None:
                 raise ValueError("No audio generated by Kokoro")
             with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as fp:
@@ -84,7 +85,7 @@ def text_to_speech(text, language, tts_engine):
         return None, f"TTS error: {str(e)}"
 
 # Main function to process audio input and return outputs
-def process_audio(audio_file, target_language, tts_engine):
+def process_audio(audio_file, target_language):
     if audio_file is None:
         return "Please upload an audio file or record audio.", None, None, None
     
@@ -96,7 +97,7 @@ def process_audio(audio_file, target_language, tts_engine):
     if error:
         return error, transcription, None, None
     
-    audio_output, error = text_to_speech(translated_text, target_language, tts_engine)
+    audio_output, error = text_to_speech(translated_text, target_language)
     if error:
         return error, transcription, translated_text, None
     
@@ -105,23 +106,17 @@ def process_audio(audio_file, target_language, tts_engine):
 # Gradio interface
 with gr.Blocks(title="AI Audio Translator") as demo:
     gr.Markdown("# AI Audio Translator")
-    gr.Markdown("Upload an audio file or record via microphone, select a target language and TTS engine, and get the transcription, translation, and translated audio!")
+    gr.Markdown("Upload an audio file or record via microphone, select a target language, and get the transcription, translation, and translated audio! Uses Kokoro TTS for supported languages, otherwise gTTS.")
     
     supported_langs = list(set(list(KOKORO_LANGUAGES.keys()) + list({v: k for k, v in lang.tts_langs().items()}.keys())))
     
     with gr.Row():
         audio_input = gr.Audio(sources=["upload", "microphone"], type="filepath", label="Input Audio")
-        with gr.Column():
-            target_lang = gr.Dropdown(
-                choices=sorted(supported_langs),
-                value="Spanish",
-                label="Target Language"
-            )
-            tts_engine = gr.Radio(
-                choices=["Kokoro", "gTTS"],
-                value="gTTS",
-                label="Text-to-Speech Engine"
-            )
+        target_lang = gr.Dropdown(
+            choices=sorted(supported_langs),
+            value="Spanish",
+            label="Target Language"
+        )
     
     submit_btn = gr.Button("Translate")
     
@@ -133,7 +128,7 @@ with gr.Blocks(title="AI Audio Translator") as demo:
     
     submit_btn.click(
         fn=process_audio,
-        inputs=[audio_input, target_lang, tts_engine],
+        inputs=[audio_input, target_lang],
         outputs=[error_output, transcription_output, translation_output, audio_output]
     )