d2-widget 0.0.1rc2

An AnyWidget for displaying declarative diagrams written in D2

D2 Widget

Bring the power of D2 to Python notebooks.

d2-widget is an AnyWidget for displaying declarative diagrams written in D2.

Installation

pip install d2-widget

or with uv:

uv add d2-widget

Usage

The following examples demonstrate how to use Widget with increasing complexity.

Basic Usage

The simplest way to use Widget is to pass a D2 diagram as a string to the constructor.

from d2_widget import Widget

Widget("x -> y")

Inline Configuration

You can add direction and layout settings directly in the D2 markup.

from d2_widget import Widget

Widget("""
direction: right
x -> y
""")

Compile Options

You can specify compile options using the second argument to the constructor. You can read about the semantics of the options in the D2 documentation.

from d2_widget import Widget

Widget("""
direction: right
x -> y
""",
  {
      "themeID": 200,  # ID of the "Dark mauve" theme
      "pad": 0,        # Disable padding
      "sketch": True,  # Enable sketch mode
  },
)

Accessing the SVG

You can access the generated SVG using the svg attribute.

from d2_widget import Widget

w = Widget("x -> y")
w.svg

%%d2 Cell Magic

You can use the %%d2 cell magic to display a D2 diagram in a Jupyter notebook.

First, you need to load the extension:

%load_ext d2_widget

Then, you can use the %%d2 cell magic to display a D2 diagram. You can pass compile options to the cell magic using keyword arguments.

%%d2 sketch=True themeID=200
direction: right
x -> y
y -> z { style.animated: true }
z -> x

Development

We recommend using uv for development. It will automatically manage virtual environments and dependencies for you.

uv run jupyter lab example.ipynb

Alternatively, create and manage your own virtual environment:

python -m venv .venv
source .venv/bin/activate
pip install -e ".[dev]"
jupyter lab example.ipynb

The widget front-end code bundles it's JavaScript dependencies. After setting up Python, make sure to install these dependencies locally:

pnpm install

While developing, you can run the following in a separate terminal to automatically rebuild JavaScript as you make changes:

pnpm dev

Open example.ipynb in JupyterLab, VS Code, or your favorite editor to start developing. Changes made in js/ will be reflected in the notebook.