Author: This is a demo package showing how to run a web app through Jupyter Server Proxy.
This mechanism allows a user to run a separate web app through JupyterHub. To try it, install this package into the same Python environment which is used to launch your single-user server:
sudo -E pip install git+https://github.com/arang129/aita.git sudo -E pip install --upgrade git+https://github.com/arang129/aita.git sudo -E pip uninstall aita
If your server is already running use the JupyterHub control panel
(/hub/home) to stop and start it. You should have a new 'hello' option in
the 'New' menu (classic notebook) or the launcher (Jupyterlab). You can also
go directly to https://(your-jhub-server)/user-redirect/hello/ .
This is meant as a starting point for building useful applications to run in Jupyter Server Proxy. See the JSP docs and especially the examples page for more information.
For real web applications in Python, you will want a web framework rather than
the low-level http.server module. There are many choices, but Tornado (used by Jupyter) and Flask are two well known ones.
Security: This example uses a Unix socket between Jupyter and the proxied application, which is a new option in Jupyter Server Proxy 4.0. This is set up so that only the user running the application can connect to it. If you choose to use a TCP socket instead, pay attention to whether other users can connect to it and what it might allow them to do.