Django application for moderation of model instances.
Provides convenience methods and an admin interface for moderating instances of registered Django models.
django-gatekeeper is a project of Sunlight Labs (c) 2008. Writen by Jeremy Carbaugh <jcarbaugh@sunlightfoundation.com>
All code is under a BSD-style license, see LICENSE for details.
Homepage: http://pypi.python.org/pypi/django-gatekeeper/
Source: http://github.com/sunlightlabs/django-gatekeeper/
python >= 2.4
django >= 1.0
To install run
python setup.py install
which will install the application into python's site-packages directory.
Add to INSTALLED_APPS:
gatekeeper
Be sure to place gatekeeper above any application which contains models that will be moderated.
Add to MIDDLEWARE_CLASSES:
gatekeeper.middleware.GatekeeperMiddleware
Add to urls.py:
url(r'^admin/gatekeeper/', include('gatekeeper.urls')),
>>> from django.db import models >>> import gatekeeper >>> class MyModel(models.Model): ... pass >>> gatekeeper.register(MyModel)
gatekeeper.register
optionally takes some extra parameters for advanced configuration, see advanced register options for details.
Include the following in urls.py BEFORE the default admin:
url(r'^admin/gatekeeper/', include('gatekeeper.urls')),
Models registered with gatekeeper.register
are given a custom manager that adds several extra fields. (As documented in advanced register options all of these fields can be renamed via special parameters to gatekeeper.register
but for simplicity they will be referred to by their default names throughout this document.)
Every model instance accessed via the custom manager ('objects' by default) has a few extra read-only fields tacked on:
- moderation_status - current moderation status (APPROVED_STATUS, PENDING_STATUS, or REJECTED_STATUS)
- flagged - flag status (True or False)
In addition, moderation_object
is provided as a shortcut to access the actual ModeratedObject
instance. The first access of this attribute incurs a database hit but the instance is then cached for the object lifetime.
The custom manager returns a special GatekeeperQuerySet with a few extra filters:
>>> MyModel.objects.all().approved() # approved by moderator >>> MyModel.objects.all().pending() # pending in moderation queue >>> MyModel.objects.all().rejected() # rejected by moderator >>> MyModel.objects.all().flagged() # flagged >>> MyModel.objects.all().not_flagged() # not flagged
These are implemented on the GatekeeperQuerySet itself so that they can be chained:
>>> MyModel.objects.filter(spam=eggs).rejected().flagged()
When a moderated object instance is deleted, the associated ModeratedObject instance is deleted as well.
It can be hassle to have to manually moderate objects when there is a simple ruleset used to determine how an object will be moderated. In order to use auto-moderation, the following needs to be added to settings.py:
GATEKEEPER_ENABLE_AUTOMODERATION = True
Gatekeeper provides two methods of auto-moderation. First, if the user that saves a moderated object has permission to moderate that object, it will be automatically approved. This will always happen if GATEKEEPER_ENABLE_AUTOMODERATION is set to true in settings.py. The second form of auto-moderation allows a moderation method to be written. This method should return True to approve, False to reject, or None to pass on for manual moderation. The auto-moderation function is pass as an argument when registering a Model.
>>> class MyModel(models.Model): ... pass >>> def myautomod(obj): ... pass >>> gatekeeper.register(MyModel, auto_moderator=myautomod)
If the auto-moderation function returns None or is not specified for a model, the first form of auto-moderation will be attempted.
By default, moderated model instances will be marked as pending and placed on the moderation queue when created. This behavior can be overridden by specifying GATEKEEPER_DEFAULT_STATUS in settings.py.
gatekeeper.PENDING_STATUS
- mark objects as pending and place on the moderation queuegatekeeper.APPROVED_STATUS
- mark objects as approved and bypass the moderation queuegatekeeper.PENDING_STATUS
- mark objects as rejected and bypass the moderation queue
Flagging is provided as a simple mechanism to allow for users to flag content generally for further review. By default flagging an object does not change it's moderation status, but GATEKEEPER_STATUS_ON_FLAG is available toalter this behavior. If GATEKEEPER_STATUS_ON_FLAG is set to one of the status constants (see Default Moderation) the given status will be set for an object when ModeratedObject.flag
is called.
If gatekeeper is added to an existing application, objects already in the database will not be registered with gatekeeper. You can register existing objects with gatekeeper by passing true to the import_unmoderated
parameter of the registration method. The imported objects will be set to the state specified by GATEKEEPER_DEFAULT_STATUS in settings.py or pending if GATEKEEPER_DEFAULT_STATUS is not set.
>>> gatekeeper.register(MyModel, import_unmoderated=True)
Gatekeep will send a notification email to a list of recipients when a new object is placed on the moderation queue. Specify GATEKEEPER_MODERATOR_LIST in settings.py to enable notifications.
GATEKEEPER_MODERATOR_LIST = ['moderator@mysite.com','admin@yoursite.com']
Many applications will want to execute certain tasks once an object is moderated. Gatekeeper provides a signal that is fired when an object is manually or automatically moderated.
gatekeeper.post_moderation
gatekeeper.register
takes a few optional parameters:
- manager_name:
- name of manager to add/replace on model (defaults to objects)
- status_name:
- name of moderation status field to add to model instances (defaults to
moderation_status
) - flagged_name:
- name of flagged field to add to model instances (defaults to
flagged
) - moderation_object_name:
- name of moderation_object accessor to add to model instances (defaults to
moderation_object
) - base_manager:
- Manager for contributed manager placed at
manager_name
to inherit from. Default behavior is to attempt to inherit from same class as the manager already in place or fall back tomodels.Manager
if no manager exists.