Update app.py

app.py

@@ -3,19 +3,21 @@ from PIL import Image
 import io
 import base64
 import uuid
-
-# Assuming google.generativeai is correctly imported as genai and the API key is set
+from gtts import gTTS
 import google.generativeai as genai
 
-api_key = "AIzaSyC70u1sN87IkoxOoIj4XCAPw97ae2LZwNM"  # Replace with your actual API key
+# Set your API key
+api_key = "YOUR_ACTUAL_API_KEY"  # Replace with your actual API key
 genai.configure(api_key=api_key)
 
+# Configure the generative AI model
 generation_config = genai.GenerationConfig(
     temperature=0.9,
     max_output_tokens=3000
 )
 
-safety_settings = SAFETY_SETTINGS = [
+# Safety settings configuration
+safety_settings = [
     {
         "category": "HARM_CATEGORY_DANGEROUS_CONTENT",
         "threshold": "BLOCK_NONE",
@@ -34,6 +36,7 @@ safety_settings = SAFETY_SETTINGS = [
     },
 ]
 
+# Initialize session state
 if 'chat_history' not in st.session_state:
     st.session_state['chat_history'] = []
 if 'file_uploader_key' not in st.session_state:
@@ -43,6 +46,7 @@ if 'use_vision_model' not in st.session_state:
 
 st.title("Gemini Chatbot")
 
+# Helper functions for image processing and chat history management
 def get_image_base64(image):
     image = image.convert("RGB")
     buffered = io.BytesIO()
@@ -55,29 +59,30 @@ def clear_conversation():
     st.session_state['file_uploader_key'] = str(uuid.uuid4())
     st.session_state['use_vision_model'] = False
 
-def send_message():
+def display_chat_history():
+    for entry in st.session_state['chat_history']:
+        role = entry["role"]
+        parts = entry["parts"][0]
+        if 'text' in parts:
+            st.markdown(f"{role.title()}: {parts['text']}")
+        elif 'data' in parts:
+            st.image(Image.open(io.BytesIO(base64.b64decode(parts['data']))), caption='Uploaded Image')
+
+def get_chat_history_str():
+    chat_history_str = "\n".join(
+        f"{entry['role'].title()}: {part['text']}" if 'text' in part
+        else f"{entry['role'].title()}: (Image)"
+        for entry in st.session_state['chat_history']
+        for part in entry['parts']
+    )
+    return chat_history_str
+
+# Send message function with TTS integration
+def send_message(tts=False):
     user_input = st.session_state.user_input
     uploaded_files = st.session_state.uploaded_files
-    if uploaded_files:
-        st.session_state['use_vision_model'] = True
-    prompts = []
-    for entry in st.session_state['chat_history']:
-        for part in entry['parts']:
-            if 'text' in part:
-                prompts.append(part['text'])
-            elif 'data' in part:
-                prompts.append("[Image]")
-    if user_input:
-        prompts.append(user_input)
-        st.session_state['chat_history'].append({"role": "user", "parts": [{"text": user_input}]})
-    if uploaded_files:
-        for uploaded_file in uploaded_files:
-            base64_image = get_image_base64(Image.open(uploaded_file))
-            prompts.append("[Image]")
-            st.session_state['chat_history'].append({
-                "role": "user",
-                "parts": [{"mime_type": uploaded_file.type, "data": base64_image}]
-            })
+    # Your existing code that processes user input and uploaded files...
+
     model_name = 'gemini-pro-vision' if st.session_state['use_vision_model'] else 'gemini-pro'
     model = genai.GenerativeModel(
         model_name=model_name,
@@ -97,31 +102,30 @@ def send_message():
     response_text = response.text if hasattr(response, "text") else "No response text found."
     if response_text:
         st.session_state['chat_history'].append({"role": "model", "parts": [{"text": response_text}]})
+        
+        # If TTS is enabled, convert the response text to speech
+        if tts:
+            tts = gTTS(text=response_text, lang='en')
+            tts_file = BytesIO()
+            tts.write_to_fp(tts_file)
+            tts_file.seek(0)
+            st.audio(tts_file, format='audio/mp3')
+
+    # Clear the input fields after sending the message
     st.session_state.user_input = ''
     st.session_state.uploaded_files = []
     st.session_state.file_uploader_key = str(uuid.uuid4())
-    display_chat_history()
-
-def display_chat_history():
-    for entry in st.session_state['chat_history']:
-        role = entry["role"]
-        parts = entry["parts"][0]
-        if 'text' in parts:
-            st.markdown(f"{role.title()}: {parts['text']}")
-        elif 'data' in parts:
-            st.image(Image.open(io.BytesIO(base64.b64decode(parts['data']))), caption='Uploaded Image')
 
-def get_chat_history_str():
-    chat_history_str = "\n".join(
-        f"{entry['role'].title()}: {part['text']}" if 'text' in part 
-        else f"{entry['role'].title()}: (Image)"
-        for entry in st.session_state['chat_history'] 
-        for part in entry['parts']
-    )
-    return chat_history_str
+    # Display the updated chat history
+    display_chat_history()
 
-user_input = st.text_area("Enter your message here:", key="user_input")
+# User input text area
+user_input = st.text_area(
+    "Enter your message here:",
+    key="user_input"
+)
 
+# File uploader for images
 uploaded_files = st.file_uploader(
     "Upload images:",
     type=["png", "jpg", "jpeg"],
@@ -129,16 +133,22 @@ uploaded_files = st.file_uploader(
     key=st.session_state.file_uploader_key
 )
 
-send_button = st.button("Send", on_click=send_message)
+# Send message button
+send_button = st.button(
+    "Send",
+    on_click=send_message,
+    args=(False,)  # TTS disabled by default when clicking the Send button
+)
 
+# Clear conversation button
 clear_button = st.button("Clear Conversation", on_click=clear_conversation)
 
-# Function to download the chat history
+# Function to download the chat history as a text file
 def download_chat_history():
     chat_history_str = get_chat_history_str()
     return chat_history_str
 
-# Add a button to download the chat history as a text file
+# Download button for the chat history
 download_button = st.download_button(
     label="Download Chat",
     data=download_chat_history(),
@@ -147,4 +157,20 @@ download_button = st.download_button(
 )
 
 # Ensure the file_uploader widget state is tied to the randomly generated key
-st.session_state.uploaded_files = uploaded_files
+st.session_state.uploaded_files = uploaded_files
+
+# JavaScript to capture the Ctrl+Enter event and trigger a button click
+    st.markdown(
+        """
+        <script>
+        // Use jQuery to capture the Ctrl+Enter event and click the 'Send' button
+        $(document).on('keydown', 'textarea', function(e) {
+            if (e.keyCode == 13 && e.ctrlKey) {
+                $('button:contains("Send")').click();
+                e.preventDefault();
+            }
+        });
+        </script>
+        """,
+        unsafe_allow_html=True
+    )