Run Python code from JavaScript in a Jupyter notebook

jupyter_javascript_snippet.js

// %%javascript
window.executePython = function(python) {
    return new Promise((resolve, reject) => {
        var callbacks = {
            iopub: {
                output: (data) => resolve(data.content.text.trim())
            }
        };
        Jupyter.notebook.kernel.execute(`print(${python})`, callbacks);    
    });
}



// Use it in any Jupyter JS/HTML cell like this
%%javascript
window.executePython("1 + 1")
    .then(result => console.log(result)); // Logs 2

// You can access any defined object/method in the notebook
// I suggest writing a function that returns your data as JSON and just calling the function.

Hi, thanks for posting the snippet, I am surprised there aren't more examples of this online. Would you be able to explain what's going in the executePython function for a JavaScript beginner? I am wondering if there is a need for a promise or if the code could made shorter?

Actually this doesn't work in JUpyter-Lab:

...
VM466:8 Uncaught (in promise) ReferenceError: Jupyter is not defined
...

Somehow Jupyter object is not visible from inside the notebook. Would you have any ideas these days how to communicate with the Kernel (ipython or other) from Javascript ?

cheers