A Django email backend for hijacking email sending in a staging/test environment. It extends
the default SMTP backend to intercept outgoing emails and instead send them
to a single email address that you specify. It does not intercept emails going to the site admins
(as defined by the ADMINS
setting) so it will not interfere with 500 error emails.
You can also configure the hijacked emails to be logged through standard Python logging. Mixin classes are provided to use the same hijack logic for any existing email backend such as django-ses.
- Python >= 3.6
- Django >= 2.2 (supported versions)
To install django-email-bandit via pip:
pip install django-email-bandit
Add django-email-bandit to your installed apps:
INSTALLED_APPS = ( ... 'bandit', ... )
For your test environment you should enable the backend:
EMAIL_BACKEND = 'bandit.backends.smtp.HijackSMTPBackend'
and set the email which will receive all of the emails:
BANDIT_EMAIL = '[email protected]'
or even multiple addresses:
BANDIT_EMAIL = ['[email protected]', '[email protected]']
It's also possible to whitelist certain email addresses and domains:
BANDIT_WHITELIST = [ '[email protected]', # Just this specific email address 'example.net' # All email addresses @example.net ]
Full project documentation is on Read the Docs.
We use Github Actions to lint (using pre-commit, black, isort, and flake8), test (using tox and tox-gh-actions), calculate coverage (using coverage), and build documentation (using sphinx).
We have a local script to do these actions locally, named maintain.sh
:
$ ./maintain.sh
A Github Action workflow also builds and pushes a new package to PyPI whenever a new
Release is created in Github. This uses a project-specific PyPI token, as described in
the PyPI documentation here. That token has been
saved in the PYPI_PASSWORD
settings for this repo, but has not been saved anywhere
else so if it is needed for any reason, the current one should be deleted and a new one
generated.
As always, be sure to bump the version in bandit/__init__.py
before creating a
Release, so that the proper version gets pushed to PyPI.
If you have questions, issues or requests for improvements please let us know on Github.
Development sponsored by Caktus Consulting Group, LLC.