Client and Viewer
To demonstrate potential applications of our public transit service, we developed two Python packages.
Client
The public-transit-client package was created to encapsulate the entire interaction with the Spring REST API, offering pythonic functions and typed class objects for both request and response parameters. This package was built using poetry and published to PyPI, adhering to best practices, including the inclusion of the py.typed file to support type hinting. The client is designed for both frontend applications that integrate with the routing service and researchers who need to perform bulk routing requests directly from Python without relying on a graphical user interface.
Viewer
The second package, public-transit-viewer, is a mockup frontend application built using Streamlit, an open-source Python framework designed for quickly building and deploying data-driven web applications with minimal effort. It leverages the public-transit-client package to communicate with the Java-based public transit service. The viewer provides a graphical interface to easily request connections and visualize results, making it useful for both demonstration and debugging purposes. The viewer is published as image on the GitHub Container Registry for easy deployment.
While Streamlit allowed us to rapidly develop the application, its limited potential for customization and linear, non-modular code structure makes it less ideal for a fully custom, branded frontend. For projects requiring a more powerful and customizable user interface, we would recommend using a JavaScript framework. However, given the objectives of this study, Streamlit's simplicity and rapid development capabilities made it an adequate choice for our needs.