django-janitor allows you to use bleach to clean HTML stored in a Model's field.

django, html, html-parsing, python, whitelist
pip install django-janitor==0.5.1



django-janitor allows you to use bleach to clean HTML stored in arbitrary Models.

This is useful when you've got content stored in a 3rd-party app, but you'd like to use a whitelist for allowed HTML tags.


  • You don't need to edit an existing app (django-janitor listens for a Models' pre_save signal)
  • Allowed HTML tags, attributes, and css are specified in a whitelist
  • Whitelists are managed in Django's Admin
  • There are (sensible) default whitelists included


  • Requires Bleach >= 1.4
  • Works with Django 1.4 to 1.6.5


To install the most recent release:

pip install django-janitor

To install from the current repository:

pip install git+

Then, add janitor to your installed apps, and run syncdb or run the south migrations:

python migrate janitor


Visit the Janitor app in Django's Admin, and create a new Field sanitizer. Then select the Model and specify the fieldname which should be cleaned. After you set up the whitelists for Tags, Attributes, etc, save the Field sanitizer.

From now on, when the Model is saved, it's content in will be cleaned using bleach.

Here's a Screenshot:



There are a few tests in janitor/tests. You can run these with:

python tests janitor

These tests dynamically add a sample app/model to INSTALLED_APPS, then call syncdb.

Management Commands

There are a few managment commands avaialable to make it easier to use django-janitor. The first is clean_all which will look at all of the models that have a related Field Sanitizer, calling the models' save method to trigger the pre_save signal (which forces the fields to be cleaned):

python clean_all

This is useful if you've created a FieldSanitizer for a model with existing content.

The second management command is clean_model, which works in a similar fashion, but allows you to specify an app and a model:

python clean_model myapp.MyModel

Finally, list_html_elements and list_html_elements_for_model exist to help you discover what HTML tags are being used in existing content. While these commands do require that a FieldSanitizer be configured for existing Models, they may be used to help you decide which tags to include in a whitelist.

You should run these commands before using clean_all or clean_model to see what sort of data exists before it's cleaned:

python list_html_elements


python list_html_elements_for_model myapp.MyModel