jp-babel: A babel Kernel for the Jupyter Notebook

jp-babel is an npm package that implements a babel kernel for the Jupyter notebook). A Jupyter notebook combines the creation of rich-text documents (including equations, plots and videos) with the execution of code in a number of programming languages.

The execution of code is carried out by means of a kernel that implements the Jupyter messaging protocol. There are kernels available for Python, Julia, Ruby, Haskell and many other languages.

Proof-of-Concept and Goals

jp-babel came to existence prompted by a number of requests from IJavascript users. See this issue for further details.

By publishing jp-babel I'm seeking to:

• provide users with a "usable" babel kernel that with some support may become a fully featured kernel,

• reuse the IJavascript code and develop a Node.js library to implement Jupyter kernels for other languages.

Installation

Please, refer to the installation notes for IJavascript.

For example, in Ubuntu 16.04, you can run:

sudo apt-get install nodejs-legacy npm ipython ipython-notebook
sudo npm install -g jp-babel

babelrc

By default, jp-babel is set to use the preset babel-preset-env. Users can customise their .babelrc. If they do so, they must ensure all the referenced plugins and presets can be required from within the jp-babel session.

Usage

jp-Babel provides 5 executables: jp-babel-install, jp-babel-notebook, jp-babel-console, jp-babel-kernel and jp-babel. Their purpose and basic use is described in the sections below. Please, refer to the usage notes for further details.

jp-babel-install: jp-Babel kernel spec installer

'jp-babel-install` registers the jp-Babel kernel with Jupyter, so that other tools (e.g. the Jupyter notebook) can invoke it. The following command flags are recognised:

--debug                   enable debug messages
--help                    show this help
--hide-undefined          do not show undefined results
--install=[local|global]  install kernel for current user or globally
--protocol=version        set messaging protocol version, e.g. 5.0
--show-undefined          show undefined results
--spec-path=[none|full]   set whether kernel spec uses full paths
--startup-script=path     run script on kernel startup
                          (path can be a file or a folder)
--version                 show kernel version
--versions                show kernel and library versions
--working-dir=path        set kernel working directory
                          (default = current working directory)

jp-babel-notebook: jp-Babel notebook

After running jp-babel-install, Jupyter notebook users can invoke the Jupyter notebook as usual. jp-babel-notebook is provided for convenience to users of the IPython notebook prior to version 3. jp-babel-notebook is a wrapper around ipython notebook. It extends the command flags accepted by ipython notebook with the following:

--help                       show jp-Babel and notebook help
--jp-debug                   enable debug messages
--jp-help                    show this help
--jp-hide-undefined          do not show undefined results
--jp-install=[local|global]  install kernel for current user or globally
--jp-protocol=version        set protocol version, e.g. 5.0
--jp-show-undefined          show undefined results
--jp-spec-path=[none|full]   set whether kernel spec uses full paths
--jp-startup-script=path     run script on startup
                             (path can be a file or a folder)
--jp-working-dir=path        set kernel working directory
                             (default = current working directory)
--version                    show kernel version
--versions                   show kernel and library versions

jp-babel-console: jp-Babel console

jp-babel-console is provided for convenience to users as a wrapper around jupyter console --kernel=babel. The following command flags are recognised:

--help                       show jp-Babel and notebook help
--jp-debug                   enable debug messages
--jp-help                    show this help
--jp-hide-undefined          do not show undefined results
--jp-install=[local|global]  install kernel for current user or globally
--jp-protocol=version        set protocol version, e.g. 5.0
--jp-show-undefined          show undefined results
--jp-spec-path=[none|full]   set whether kernel spec uses full paths
--jp-startup-script=path     run script on startup
                             (path can be a file or a folder)
--jp-working-dir=path        set kernel working directory
                             (default = current working directory)
--version                    show kernel version
--versions                   show kernel and library versions

jp-babel-kernel: jp-Babel kernel

jp-babel-kernel is the executable invoked by Jupyter tools (e.g. the notebook) and that appears in the kernel spec that jp-babel-install creates for jp-Babel. You won't need this command, unless you want to create a custom kernel spec.

Usage:
    jp-babel-kernel [options] connection_file

Options:
    --debug                           enable debug messages
    --hide-undefined                  do not show undefined results
    --protocol=Major[.minor[.patch]]  set protocol version, e.g. 5.0
    --session-working-dir=path        set session working directory
    --show-undefined                  show undefined results
    --startup-script=path             run script on startup
                                      (path can be a file or a folder)

jp-babel: Deprecated CLI

jp-babel is provided for backwards-compatibility. It will be removed in the next major-version update. Please, use jp-babel-install or jp-babel-notebook instead.

Troubleshooting

• jp-babel takes 10-15 seconds on the first execution after kernel start (Issue #4): This issue has been seen on a system using npm 2. As suggested here, running npm dedupe on the folder where jp-babel was installed helps reduce this time. Alternatively, using npm 3 to install jp-babel also helps.

Contributions

First of all, thank you for taking the time to contribute. The maintenance of IJavascript is currently my priority. I would really appreciate some help. Please, read CONTRIBUTING and use the issue tracker for any contributions: support requests, bug reports, enhancement requests, pull requests, submission of tutorials...

TO DO

• Add tests