A cookiecutter template for creating a JupyterLab and Jupyter Notebook extension for rendering specific mime types and file extensions.

• jupyterlab_json
• jupyterlab_geojson
• jupyterlab_plotly

Install cookiecutter:

pip install cookiecutter

Use cookiecutter to generate a package:

cookiecutter https://github.com/jupyterlab/mimerender-cookiecutter

The cookiecutter will prompt you with the following questions and generate a project according to your responses:

• author_name: Your full name. This will be used in the generated Python and npm packages.
• author_email: Your email address. This will be used in the generated Python and npm packages.
• mime_type: A valid mime type (e.g. application/json, application/geo+json). This will be used to render output data of this mime type with your extension.
• mime_short_name: A display name (no spaces) for your mime type (e.g. JSON, GeoJSON). This will be used in the generated Python and npm packages, README, and class names.
• file_type: OPTIONAL A valid file type (e.g. json, geojson). This will be used to open files of this type with your extension.
• extension_name: Your JupyterLab and Jupyter Notebook extension name (e.g. jupyerlab_json, jupyerlab_geojson).

The project is divided into 2 top-level directories, one for each extension (lab and notebook).

In most cases, you will only need to edit the OutputWidget._render method in labextension/src/output.js (for rendering output data of a specific mime type) and the DocWidget.onUpdateRequest method in labextension/src/doc.js (if your extension should render files of a specific type).

• extension_name
  • extension_name: The Python package
    • static: Compiled Javascript for the notebook extension.
    • __init__.py: Exports paths and metadata of lab and notebook extensions and exports an optional display method that can be imported into a notebook and used to easily display data using this renderer
  • src - The extension source.
    • embed.js: Entry point for embedded widget
    • extension.js: Integration point with Jupyter Notebook
    • index.js: Entry point for the Jupyter Notebook extension
    • plugin.js: Entry point for the JupyterLab extension
    • renderer.js: Methods for rendering output data of mime_type defined in prompts

This cookiecutter will change over time to adapt to API changes and best practices, so you should be able to rebase your extension from an updated version of the cookiecutter. The following is the recommended workflow:

• Create a branch called cookiecutter that will serve as the base for another branch called develop. All feature branches shoud be based on develop and merged into master.
• When a new version of mimerender-cookiecutter is available:
  • Checkout cookiecutter branch: git checkout cookiecutter
  • Re-run the cookiecutter command from the repository root: cookiecutter https://github.com/jupyterlab/mimerender-cookiecutter [--checkout BRANCH_NAME] --output-dir .. --config-file .cookiecutter.yaml --no-input --overwrite-if-exists
    • --checkout is optional and specifies a branch of mimerender-cookiecutter to checkout (e.g. react).
    • --output-dir .. allows us to run the cookiecutter in the repo root vs. the parent directory because cookiecutter will always render output in a child directory of the output-dir.
    • --config-file .cookiecutter.yaml will run the cookiecutter with the original inputs which are persisted to .cookiecutter.yaml in the repo root.
    • --no-input will skip the cookiecutter prompts.
    • --overwrite-if-exists will replace our previous cookiecutter output with the most up-to-date output.
  • Commit these changes to cookiecutter with a message like "mimerender-cookiecutter [VERSION_NUMBER]": git commit -m "mimerender-cookiecutter [VERSION_NUMBER]"
  • Checkout the develop branch: git checkout develop
  • Rebase it from cookiecutter: git rebase cookiecutter
    • You may encounter some merge conflicts, so resolve them and complete the rebase using git rebase --continue
  • Publish a new version of your extension by following the instructions in RELEASE.md.

We suggest that extension names start with jupyterlab_ and use underscores if needed to improve readability, such as jupyterlab_myextension.