Contribute to Jupyter Book¶

Welcome to the jupyter-book repository! We’re excited you’re here and want to contribute. ✨

Development guidelines¶

For information about development conventions, practices, and infrastructure, please see the executablebooks/ development guidelines.

Getting started¶

To get started with Jupyter Book’s codebase, take the following steps:

Clone and install the package¶

git clone https://github.com/executablebooks/jupyter-book
cd jupyter-book

Next, install:

pip install -e .[testing,code_style,sphinx]

This will install Jupyter Book locally, along with the packages needed to test it as well as packages for ensuring code style.

Install the pre-commit hooks¶

Jupyter Book uses pre-commit to ensure code style and quality before a commit is made. This ensures that the look and feel of Jupyter Book remain consistent over time and across developers. pre-commit is installed when you install jupyter book with pip install -e .[code_style].

To enable pre-commit for your clone, run the following from the repository root:

pre-commit install

From now on, when you make a commit to Jupyter Book, pre-commit will ensure that your code looks correct according to a few checks.

Run the tests¶

For code tests, jupyter-book uses pytest). You may run all the tests, or only ones that do not require additional installations, with the following command:

>> pytest -m 'not requires_chrome and not requires_tex'

You can alternatively use tox, to run the tests in multiple isolated environments, and also without the need for the initial dependencies install (see the tox.ini file for available test environments and further explanation):

>> tox -e py38-sphinx3 -- -m 'not requires_chrome and not requires_tex'

Either will both run the Jupyter Book test suite, except for the PDF tests. This is because running the PDF generation tests require a full Latex environment, which you may not have set up.

Note

Jupyter Book makes use of pytest-xdist for running tests in parallel. You can take advantage of this by running tests with the -n argument followed by the number of CPUs you would like to use. For example: pytest -n 4. This makes the tests run much faster.

To test PDF generation¶

If you’d like to test (or try out) the generation of PDFs, take the following steps:

To generate PDFs via HTML, make sure you install Jupyter Book with pip install -e .[pdfhtml]. This will install pyppeteer, which runs a headless chrome session to convert your book to PDF. Next, follow the installation instructions at Build a PDF from your book HTML. You should then be able to build your book’s PDF through HTML.

To generate PDFs via Latex, make sure you install a working latex distribution locally. Do so by following the instructions in Build a PDF using Latex.

If you have installed the requirements for both HTML and Latex generation, you should be able to run the full test suite with pytest.

Repository Structure of Jupyter Book¶

This section covers the general structure of the Jupyter Book repository, and explains which pieces are where.

The Jupyter Book repository contains two main pieces:

MyST markdown¶

Jupyter Book supports an extended version of Jupyter Markdown called “MyST markdown”. For information about the myst syntax and how to use it, see the myst parser documentation.

The command-line tool and Python package¶

This is used to help create and build books. It can be found at ./jupyter_book.

The commands/ folder has the CLI. This is the interface for users to create, build, and control their book via the command-line.
The sphinx.py module builds the books.
The yaml.py module handles configuration.
The toc.py module prepares the table of contents.

The other modules handle more specific functionality in Jupyter Book - see their module docstrings for more information.

The template Jupyter Book¶

Jupyter Book comes bundled with a small template book to show off content. This can be immediately built with jupyter-book build. It can be found at jupyter_book/book_template.

An example¶

Here are a few examples of how this code gets used to help you get started.

when somebody runs jupyter-book create mybook/, the create.py module is used to generate an empty template using the template in jupyter_book/book_template/.
when somebody runs jupyter-book build mybook/, the build.py module to loop through your page content files, and uses the page/ module to convert each one into HTML and places it in mybook/_build.

Hopefully this explanation gets you situated and helps you understand how the pieces all fit together. If you have any questions, feel free to open an issue asking for help!

Other major tools in the Jupyter Book stack¶

Jupyter Book depends on a collection of open source tools in the Python / Sphinx ecosystem. These tools do much of the heavy lifting of Jupyter Book, and are where many improvements and changes will need to be. Here is a list of the major tools and what kinds of functionality they support:

The Sphinx Documentation engine is used to build book outputs. This relies on a number of extensions that are developed by Jupyter Book.
MyST markdown is parsed into Sphinx by the MyST Parser.
The MyST-NB package parses Jupyter Notebooks into Sphinx and also controls some parts of notebook execution. It also provides inserting code outputs into content.
Jupyter-Cache manages the execution and cacheing of notebook content at build time. It is controlled by MyST-NB.
The Sphinx-Book-Theme defines the look and feel of Jupyter Book, and is where most of the CSS rules are defined.