Egison Kernel for Jupyter

A Jupyter kernel for the Egison programming language (version 5).

Provides syntax highlighting in Jupyter notebooks via a CodeMirror mode that supports Egison 5 syntax including type classes, inductive data types, type annotations, and pattern matching.

How to Install

1. Create a Python virtual environment and install dependencies

cd egison-jupyter
python3 -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install .

2. Install Jupyter and the Egison kernel

pip install jupyter
python -m egison_kernel.install

3. Install the syntax highlighting extension

pip install jupyterlab
cd jupyterlab-egison
pip install -e .

4. Rebuild the syntax highlighting extension (after editing)

If you modify the CodeMirror mode source (jupyterlab-egison/src/index.ts), rebuild and reinstall the extension, then restart Jupyter:

cd jupyterlab-egison
jlpm build:lib && jlpm build:labextension:dev
pip install -e .

Then restart Jupyter Notebook (Ctrl+C to stop, then jupyter notebook again).

How to Use

Make sure the virtual environment is activated, then start Jupyter:

source .venv/bin/activate
jupyter notebook
# In the notebook interface, select Egison from the 'New' menu

Syntax Highlighting

The CodeMirror mode supports Egison 5 syntax:

• Type system keywords: class, instance, inductive, extends, declare
• Definition keywords: def, let, in, where
• Pattern matching: match, matchAll, matchDFS, matchAllDFS, as, with, loop, forall
• Built-in types: Integer, MathExpr, Float, Bool, Char, String, IO, Tensor, Vector, Matrix, DiffForm, Matcher, Pattern, List
• Pattern variables ($x), value patterns (#x)
• Comments: line comments (--) and nested block comments ({- -})
• Tensor index notation and mathematical symbols
• Greek letters are treated as regular identifiers (no special highlighting)

Requirements

• Egison 5.x installed and available in PATH
• Python 3
• Jupyter Notebook or JupyterLab

Acknowledgement

I learned how to implement a Jupyter kernel from bash_kernel.

I thank Shunsuke Gotoh for his article on the initial Python program of this kernel.