Streamlit Chat Widget

The streamlit_chat_widget is a custom chat input component designed for Streamlit applications, allowing users to send text and audio messages seamlessly. This widget can be integrated into Streamlit apps, making it ideal for building conversational AI interfaces, voice assistants, or any chat-based applications the component was developed by Mohammed Bahageel artificial Intelligence developer.

Features

• Text Input: Users can type text messages and send them.
• Audio Recording: Built-in microphone functionality to capture audio messages.
• Fixed Positioning: The widget remains fixed at the bottom of the Streamlit app, similar to Streamlit's st.chat_input.

Installation

Install the package directly from PyPI:

pip install --U streamlit-chat-widget

Setup

Step 1: Import and Initialize the Component

To use the component, import it into your Streamlit app. The widget will display at the bottom of the screen by default.

Step 2: Set Up the Frontend Component

If you want to modify or rebuild the React component, you can find the source code in the frontend directory.

1. Navigate to the frontend folder:

   cd streamlit_chat_widget/frontend
   

2. Install dependencies:

   npm install
   

3. Build the frontend:

   npm run build
   

4. Run the component locally (for development):

   npm start
   

Usage

Here’s a sample code to get you started with streamlit_chat_widget:

import streamlit as st
from streamlit_chat_widget import chat_input_widget

def main():
    st.title("My Custom Chat Application")
    
    # Initialize chat history in Streamlit session state
    if "chat_history" not in st.session_state:
        st.session_state.chat_history = []

    # Display the chat history
    for message in st.session_state.chat_history:
        st.write(message)

    # Display the chat input widget at the bottom
    user_input = chat_input_widget()

    # Process the user's input from the widget
     if user_input:
        if "text" in user_input:
            user_text =user_input["text"]
            st.session_state.chat_history.append(f"You: {user_text}")
        elif "audioFile" in user_input:
            audio_bytes = bytes(user_input["audioFile"])
            st.audio(audio_bytes)

if __name__ == "__main__":
    main()

Example

To create a simple app with the chat widget:

import os
import streamlit as st
from streamlit_chat_widget import chat_input_widget

def app():
    st.title("Doctor AI Assistant")
    
    if "chat_history" not in st.session_state:
        st.session_state.chat_history = [
            "Hello! I am your Doctor AI Assistant. How can I help you?"
        ]

    # Display chat history
    for message in st.session_state.chat_history:
        st.write(message)

    # Display the chat input widget at the bottom
    user_input = chat_input_widget()

    if user_input:
        if "text" in user_input:
            user_text =user_input["text"]
            st.session_state.chat_history.append(f"You: {user_text}")
        elif "audioFile" in user_input:
            audio_bytes = bytes(user_input["audioFile"])
            st.audio(audio_bytes)

if __name__ == "__main__":
    app()

CSS Customization

To style the widget or customize its appearance, add custom CSS in your Streamlit app. Here’s a basic example:

st.markdown("""
<style>
.chat-container {
    position: fixed;
    bottom: 0;
    width: 100%;
    max-width: 600px;
    margin: 0 auto;
}
</style>
""", unsafe_allow_html=True)
 

Additional Customization:

To postion the chat input widget in a fixed postion at the bottom of streamlit application use streamlit floating containers

import streamlit as st
from streamlit_float import *
from streamlit_chat_widget import chat_input_widget

def app():  
    float_init()
    footer_container = st.container()
    with footer_container:
        user_input = chat_input_widget()
    footer_container.float(
        "display:flex; align-items:center;justify-content:center; overflow:hidden visible;flex-direction:column; position:fixed;bottom:15px;"
    )
    
if __name__ == "__main__":
    app()

License

This project is licensed under the MIT License. See the LICENSE file for more information.