Update app.py

app.py

@@ -280,122 +280,6 @@ def process_video_with_gpt(video_input, user_prompt):
     )
     return response.choices[0].message.content
 
-
-
-def process_tts(text, voice="alloy"):
-    """
-    Process text-to-speech using OpenAI's TTS API
-    Voices available: alloy, echo, fable, onyx, nova, shimmer
-    """
-    try:
-        response = openai_client.audio.speech.create(
-            model="tts-1",
-            voice=voice,
-            input=text
-        )
-        
-        # Generate a unique filename
-        filename = generate_filename("tts_output", "mp3")
-        
-        # Save the audio file
-        response.stream_to_file(filename)
-        
-        # Create audio player HTML
-        audio_html = f"""
-        <audio controls>
-            <source src="data:audio/mp3;base64,{base64.b64encode(open(filename, 'rb').read()).decode()}" type="audio/mp3">
-            Your browser does not support the audio element.
-        </audio>
-        """
-        
-        return filename, audio_html
-    except Exception as e:
-        st.error(f"TTS Error: {str(e)}")
-        return None, None
-
-def update_chat_interface():
-    """Update the chat interface to include voice selection and TTS playback"""
-    # Add voice selection to sidebar
-    st.sidebar.markdown("### 🗣️ TTS Voice Settings")
-    selected_voice = st.sidebar.selectbox(
-        "Choose TTS Voice:",
-        ["alloy", "echo", "fable", "onyx", "nova", "shimmer"],
-        help="Select the voice for text-to-speech playback"
-    )
-    
-    # Store the selected voice in session state
-    if "selected_voice" not in st.session_state:
-        st.session_state.selected_voice = selected_voice
-
-# Modify the chat processing functions to include TTS
-def process_with_gpt(text_input):
-    """Process text with GPT-4o and add TTS."""
-    if text_input:
-        st.session_state.messages.append({"role": "user", "content": text_input})
-        
-        with st.chat_message("user"):
-            st.markdown(text_input)
-        
-        with st.chat_message("assistant"):
-            completion = openai_client.chat.completions.create(
-                model=st.session_state["openai_model"],
-                messages=[
-                    {"role": m["role"], "content": m["content"]}
-                    for m in st.session_state.messages
-                ],
-                stream=False
-            )
-            return_text = completion.choices[0].message.content
-            st.write("GPT-4o: " + return_text)
-            
-            # Add TTS playback
-            filename, audio_html = process_tts(return_text, st.session_state.selected_voice)
-            if audio_html:
-                st.markdown(audio_html, unsafe_allow_html=True)
-            
-            # Original file handling
-            filename = generate_filename("GPT-4o: " + return_text, "md")
-            create_file(filename, text_input, return_text)
-            st.session_state.messages.append({"role": "assistant", "content": return_text})
-            return return_text
-
-def process_with_claude(text_input):
-    """Process text with Claude and add TTS."""
-    if text_input:
-        with st.chat_message("user"):
-            st.markdown(text_input)
-
-        with st.chat_message("assistant"):
-            response = claude_client.messages.create(
-                model="claude-3-sonnet-20240229",
-                max_tokens=1000,
-                messages=[
-                    {"role": "user", "content": text_input}
-                ]
-            )
-            response_text = response.content[0].text
-            st.write("Claude: " + response_text)
-            
-            # Add TTS playback
-            filename, audio_html = process_tts(response_text, st.session_state.selected_voice)
-            if audio_html:
-                st.markdown(audio_html, unsafe_allow_html=True)
-            
-            # Original file handling
-            filename = generate_filename("Claude: " + response_text, "md")
-            create_file(filename, text_input, response_text)
-            
-            st.session_state.chat_history.append({
-                "user": text_input,
-                "claude": response_text
-            })
-            return response_text
-
-
-
-
-
-
     
 def extract_urls(text):
     try:
@@ -751,9 +635,11 @@ speech_recognition_html = """
             const output = document.getElementById('output');
             let fullTranscript = '';
             let lastUpdateTime = Date.now();
+
             // Configure recognition
             recognition.continuous = true;
             recognition.interimResults = true;
+
             // Function to start recognition
             const startRecognition = () => {
                 try {
@@ -763,21 +649,24 @@ speech_recognition_html = """
                     stopButton.disabled = false;
                 } catch (e) {
                     console.error(e);
-                    #status.textContent = '';
                     status.textContent = 'Error: ' + e.message;
                 }
             };
+
             // Auto-start on load
             window.addEventListener('load', () => {
                 setTimeout(startRecognition, 1000);
             });
+
             startButton.onclick = startRecognition;
+
             stopButton.onclick = () => {
                 recognition.stop();
                 status.textContent = 'Stopped';
                 startButton.disabled = false;
                 stopButton.disabled = true;
             };
+
             clearButton.onclick = () => {
                 fullTranscript = '';
                 output.textContent = '';
@@ -785,9 +674,11 @@ speech_recognition_html = """
                     type: 'clear_transcript',
                 }, '*');
             };
+
             recognition.onresult = (event) => {
                 let interimTranscript = '';
                 let finalTranscript = '';
+
                 for (let i = event.resultIndex; i < event.results.length; i++) {
                     const transcript = event.results[i][0].transcript;
                     if (event.results[i].isFinal) {
@@ -796,6 +687,7 @@ speech_recognition_html = """
                         interimTranscript += transcript;
                     }
                 }
+
                 if (finalTranscript || (Date.now() - lastUpdateTime > 5000)) {
                     if (finalTranscript) {
                         fullTranscript += finalTranscript;
@@ -805,10 +697,14 @@ speech_recognition_html = """
                     }
                     lastUpdateTime = Date.now();
                 }            
+
                 output.textContent = fullTranscript + (interimTranscript ? '... ' + interimTranscript : '');
                 output.scrollTop = output.scrollHeight;
+
                 document.getElementById('streamlit-data').value = fullTranscript;
+
             };
+
             recognition.onend = () => {
                 if (!stopButton.disabled) {
                     try {
@@ -822,6 +718,7 @@ speech_recognition_html = """
                     }
                 }
             };
+
             recognition.onerror = (event) => {
                 console.error('Recognition error:', event.error);
                 status.textContent = 'Error: ' + event.error;
@@ -986,104 +883,45 @@ def main():
         transcript = st.components.v1.html(speech_recognition_html, height=400)
         
         # Update session state if there's new data
-        showTranscript=False
-        if showTranscript:
-            #transcript is not None and transcript != "":
+        if transcript is not None and transcript != "":
             st.session_state.voice_transcript = transcript
-            
-            # Display the transcript in a Streamlit text area
-            st.markdown("### Processed Voice Input:")
-            st.text_area("Voice Transcript", st.session_state.voice_transcript, height=100)
-            
-            # Add functionality to process the transcript
-            if st.button("Process Transcript"):
-                st.subheader("AI Response to Transcript")
-                gpt_response = process_with_gpt(st.session_state.voice_transcript)
-                st.markdown(gpt_response)
-            
-            # Option to clear the transcript
-            if st.button("Clear Transcript"):
-                st.session_state.voice_transcript = ""
-                st.rerun()
-                
-    
-            # Buttons to process the transcript
-            if st.button("Search with GPT"):
-                st.subheader("GPT-4o Response")
-                gpt_response = process_with_gpt(st.session_state.voice_transcript)
-                st.markdown(gpt_response)
-    
-            if st.button("Search with Claude"):
-                st.subheader("Claude Response")
-                claude_response = process_with_claude(st.session_state.voice_transcript)
-                st.markdown(claude_response)
-    
-            if st.button("Search ArXiv"):
-                st.subheader("ArXiv Search Results")
-                arxiv_results = perform_ai_lookup(st.session_state.voice_transcript)
-                st.markdown(arxiv_results)
-                
-            
-            # Display last voice input
-            if st.session_state.last_voice_input:
-                st.text_area("Last Voice Input:", st.session_state.last_voice_input, height=100)
-
-
-        # Model Selection
-        model_choice = st.sidebar.radio(
-            "Choose AI Model:",
-            ["GPT-4o", "Claude-3", "GPT+Claude+Arxiv"]
-        )
-        
-        # Chat Interface
-        user_input = st.text_area("Message:", height=100)
         
-        if st.button("Send 📨"):
-            if user_input:
-                if model_choice == "GPT-4o":
-                    gpt_response = process_with_gpt(user_input)
-                elif model_choice == "Claude-3":
-                    claude_response = process_with_claude(user_input)
-                else:  # Both
-                    col1, col2, col3 = st.columns(3)
-                    with col2:
-                        st.subheader("Claude-3.5 Sonnet:")
-                        try:
-                            claude_response = process_with_claude(user_input)
-                        except:
-                            st.write('Claude 3.5 Sonnet out of tokens.')
-                    with col1:
-                        st.subheader("GPT-4o Omni:")
-                        try:
-                            gpt_response = process_with_gpt(user_input)  
-                        except:
-                            st.write('GPT 4o out of tokens')
-                    with col3:
-                        st.subheader("Arxiv and Mistral Research:")
-                        with st.spinner("Searching ArXiv..."):
-                            #results = search_arxiv(user_input)
-                            results = perform_ai_lookup(user_input)
-                            
-                            st.markdown(results)
-                            
-        # Display Chat History
-        st.subheader("Chat History 📜")
-        tab1, tab2 = st.tabs(["Claude History", "GPT-4o History"])
+        # Display the transcript in a Streamlit text area
+        st.markdown("### Processed Voice Input:")
+        st.text_area("Voice Transcript", st.session_state.voice_transcript, height=100)
         
-        with tab1:
-            for chat in st.session_state.chat_history:
-                st.text_area("You:", chat["user"], height=100)
-                st.text_area("Claude:", chat["claude"], height=200)
-                st.markdown(chat["claude"])
+        # Add functionality to process the transcript
+        if st.button("Process Transcript"):
+            st.subheader("AI Response to Transcript")
+            gpt_response = process_with_gpt(st.session_state.voice_transcript)
+            st.markdown(gpt_response)
         
-        with tab2:
-            for message in st.session_state.messages:
-                with st.chat_message(message["role"]):
-                    st.markdown(message["content"])
-
-
-# -------------------------------------------------------  *************************  --->
+        # Option to clear the transcript
+        if st.button("Clear Transcript"):
+            st.session_state.voice_transcript = ""
+            st.rerun()
+            
 
+        # Buttons to process the transcript
+        if st.button("Search with GPT"):
+            st.subheader("GPT-4o Response")
+            gpt_response = process_with_gpt(st.session_state.voice_transcript)
+            st.markdown(gpt_response)
+
+        if st.button("Search with Claude"):
+            st.subheader("Claude Response")
+            claude_response = process_with_claude(st.session_state.voice_transcript)
+            st.markdown(claude_response)
+
+        if st.button("Search ArXiv"):
+            st.subheader("ArXiv Search Results")
+            arxiv_results = perform_ai_lookup(st.session_state.voice_transcript)
+            st.markdown(arxiv_results)
+            
+        
+        # Display last voice input
+        if st.session_state.last_voice_input:
+            st.text_area("Last Voice Input:", st.session_state.last_voice_input, height=100)
     
     
     if tab_main == "💬 Chat":