Install

Make sure you have Jupyter Notebook or Jupyter Lab installed.

Then, install the package with:

npm install -g typescript-jupyter-kernel

Once npm has finished, run the following command:

ts-kernel install

You can now check if the installation succeeded:

jupyter kernelspec list

# Available kernels:
#  typescript        ~/.local/share/jupyter/kernels/typescript

Features

Code completion: using the TAB key, driven by the official Typescript LSP Server (same used by VSCode)

Use NPM dependencies: simply navigate to your notebook's directory and add a package.json file

cd /path/to/notebook
npm init
npm install --save short-unique-id

Fully written in Typescript: the kernel is written entirely in Typescript 🦾

SWC Optimized: the kernel compiles your code with SWC, making it blazing fast!

Able to render HTML: the kernel can render HTML in the notebook

Custom output command: the kernel includes a handy jupyter.out function so your notebooks aren't riddled with unsightly console.log calls

Fully compatible: works both in JupyterLab and Jupyter Notebook

Convenience jupyter functionalities

We have implemented a handy jupyter object containing several useful functions:

• dom

  This is a virtual dom (JSDOM) which allows you to build html in the notebook.

• html

  This is an alias for the included D3 Selection library which allows powerful HTML manipulation.

• render

  This is a function that renders the html in the notebook.

• out

  This is a function that prints a given string to the notebook between <pre> tags, useful for printing the raw output of a function or process.

• escape

  This is a function that escapes the given string so it can be safely printed in the notebook.

• data

  This is an alias for the included D3 DSV library which allows to read data from/to CSV, TSV or JSON files/strings.

• table

  This is a function that renders a table in the notebook from data using one of the data functions.

• load

  You can use the load helper to load either a js or css file into the notebook.