Thank you for contributing! EIN development will work best in a UNIX environment.
Fork the repo on github. Clone the fork to your home directory.
Install cask. Run make dist to ensure correct cask functionality.
Run make test to ensure a correct baseline. You need Jupyter, R and Julia
executables to be available through your $PATH for this (and also,
matplotlib and kernels for R and Julia installed and visible for the
aforementioned Jupyter).
Remove the MELPA-installed EIN by deleting the package directory (on my system,
it's ~/.emacs.d/elpa/ein-20190122.1341) or running M-x package-delete.
In your init.el or .emacs, add the following:
(add-to-list 'load-path "~/emacs-ipython-notebook/lisp")
(load "ein-autoloads")
Now whatever changes you make to the repo will be reflected in new emacs instances.
M-x ein:dev-start-debug activates full logging and backtrace on error.
make quick runs a syntax check and the unit tests. It is far quicker than the
laborious make test.
Located in ~/emacs-ipython-notebook/test.
If you add a feature, we encourage writing an integration test.
cask exec ecukes is the bulk of make test. Ecukes is our friend and
guardian. We follow its opinionated file structure in
~/emacs-ipython-notebook/features.
To run say just the login tests, cask exec ecukes --tags "@login".