Streamlit Keywords

A Streamlit custom component that provides enhanced functionality over the standard streamlit.multiselect. Users can enter free-form text which will appear as separate keyword chips.

Author: Tyler House (@tahouse)

Demo

Features

• 🏷️ Keyword Entry: Input multiple keywords as chips/tags.
• 📝 Free-Form Text: Users can enter free-form text, which will be tokenized into individual keywords.
• 🎨 Customizable: Customize the label, placeholder text, and set a maximum number of keywords.
• 🔢 Keyword Limit: Optionally limit the maximum number of keywords.
• 🎨 Theme Integration: Automatic integration with Streamlit's theming (dark/light modes).
• 🖥️ Seamless Integration: Integrates seamlessly with Streamlit apps.
• 🗑️ Keyword Removal: Easily remove keywords by clicking the remove icon.

Installation

pip install streamlit-keywords

Usage

import streamlit as st
from streamlit_keywords import keywords_input

# Create a keyword input
keywords = keywords_input(
    value=["example", "keywords"],
    label="Enter Keywords",
    text="Type a keyword and press Enter",
    max_keywords=5,
    key="unique_keywords_input"
)

# Display the entered keywords
st.write("Entered keywords:", keywords)

Examples

Here are some usage patterns to help you integrate streamlit_keywords into your Streamlit app. See examples/example.py for more usage examples.

Basic Keyword Input

import streamlit as st
from streamlit_keywords import keywords_input

st.title("Keyword Input Example")

keywords = keywords_input(
    label="Keywords",
    text="Add keywords and press Enter",
    key="basic_keyword_input"
)

st.write("You have entered:", keywords)

Keyword Input with Maximum Limit

import streamlit as st
from streamlit_keywords import keywords_input

keywords = keywords_input(
    label="Enter up to 5 keywords",
    max_keywords=5,
    key="keyword_input_max5"
)

st.write("Keywords:", keywords)

if keywords and len(keywords) >= 5:
    st.warning("You have reached the maximum number of keywords.")

Keyword Input with Default Values

import streamlit as st
from streamlit_keywords import keywords_input

initial_keywords = ["Streamlit", "Python", "Data Science"]

keywords = keywords_input(
    value=initial_keywords,
    label="Edit your keywords",
    key="keyword_input_with_defaults"
)

st.write("Current keywords:", keywords)

Component API

keywords_input()

Display a Streamlit component for entering keywords.

Signature:

def keywords_input(
    value: list = [],
    label: str = "Keywords Input",
    text: str = "Type a keyword and press Enter",
    max_keywords: Optional[int] = None,
    key: Optional[str] = None,
) -> List[str]:

Parameters:

• value (List[str], optional):
  Initial list of keywords. Defaults to [].

• label (str, optional):
  Label for the input component. Defaults to "Keywords Input".

• text (str, optional):
  Instructions or placeholder text displayed in the input field. Defaults to "Type a keyword and press Enter".

• max_keywords (Optional[int], optional):
  Maximum number of keywords allowed. If None, there is no limit. Defaults to None.

• key (str, optional):
  An optional string to use as the unique key for the widget. It's recommended to assign a unique key to prevent the component from remounting on script reruns. Defaults to None.

Returns:

• List[str]:
  A list of entered keywords.

Development

This repository uses the Streamlit Component template system. If you want to modify or develop the component, follow these steps.

Prerequisites

• Python: Version 3.7 or newer.
• Node.js and npm: Required for frontend development.

Backend Development

Install Python Dependencies

Clone the repository and install the development dependencies.

# Clone the repository
git clone https://github.com/tahouse/streamlit-keywords.git
cd streamlit-keywords

# Install Python dependencies
pip install -e ".[devel]"

Frontend Development

The frontend of this component is built using React and TypeScript, leveraging Material-UI components.

Install Node Dependencies

Navigate to the streamlit_keywords/frontend directory and install the dependencies as specified in package.json.

cd streamlit_keywords/frontend
npm install

package.json dependencies:

{
  "dependencies": {
    "@emotion/react": "^11.13.3",
    "@emotion/styled": "^11.13.0",
    "@mui/icons-material": "^6.1.6",
    "@mui/material": "^6.1.6",
    "react": "^17.0.2",
    "react-dom": "^17.0.2",
    "streamlit-component-lib": "^2.0.0"
  },
  "devDependencies": {
    "@babel/plugin-proposal-private-property-in-object": "^7.21.11",
    "@types/node": "^22.0.0",
    "@types/react": "^17.0.2",
    "@types/react-dom": "^17.0.2",
    "node": "^22.0.0",
    "react-scripts": "^5.0.1",
    "typescript": "^4.2.0"
  }
}

Running in Development Mode

To run the component in development mode with live reloading:

1. Start the frontend development server:

   npm start
   

   This will start the React development server.

2. Set the _RELEASE flag:

   Ensure that _RELEASE = False is set in your __init__.py. This tells the component to use the local development server.

3. Run your Streamlit app:

   In a separate terminal, navigate back to the root directory and run your Streamlit app.

   cd ../../
   streamlit run your_app.py
   

Building for Production

When you're ready to build the component for production:

1. Build the frontend:

   npm run build
   

   This will generate the production-ready frontend assets.

2. Update _RELEASE flag:

   Set _RELEASE = True in __init__.py.

   _RELEASE = True
   

3. Your component is now ready:

   The component is now ready for use in production environments.

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

Acknowledgemnts

This project draws inspiration from multiple sources:

• Streamlit Multiselect: The standard Streamlit library's multiselect component served as initial inspiration, though this project extends functionality to allow free text entry.

• streamlit-tags: Special thanks to Gagan Bhatia (@gagan3012) for streamlit-tags, which was what I initially used in my application. This project addresses mobile usability limitations in streamlit-tags where the enter button functioned only as a tab key.

• Material UI: This component is built using Material UI's Autocomplete component, which provided the foundation for the user interface.

The decision to build streamlit-keywords came from a need for a modern, mobile-friendly keyword input component with up-to-date dependencies, improved mobile user experience, and theming that closely matches Streamlit's native multiselect component.