Skip to main content

Object-level permissions across multiple models for Django.

Project description

https://travis-ci.org/Rediker-Software/olp.svg?branch=master

Object-level permissions (OLP)

Django has supported object-level permissions since Django 1.2, but it requires a custom authentication backend to implement it. There are tons of implementations out there, but most of them are meant to handle specific situations.

Features

Like most other implementations, OLP is meant to handle a few specific cases which we encountered.

Different types of “groups”

Django has a concept of “users” and “groups” which the authentication system is built upon. These groups only consist of a single model and cannot be built upon to have different, distinct types. OLP allows you to have multiple models contribute to an individual user’s permissions, even on the object level.

The different models can have independent permissions, and do not need to be directly attached to a user. If they are attached to a user (and specified in the settings), they can contribute to a user’s overall permissions.

Direct integration

OLP is directly integrated with Django’s authentication backends. While it is not required to use OLP, it is recommended as it makes it easier to use OLP.

Normalization of methods

Only the user and groups models have has_perm functions, and that becomes a quick limitation when working with other models. OLP patches the other models (as specified in the settings) and gives them has_perm methods which act the same way as the ones given to users and groups.

In order to make assigning and removing permissions as easy as checking them, OLP will also patch all of the models with assign_perm and remove_perm methods.

Settings

OLP determines the settings from the Django settings file using the key OLP_SETTINGS. This must be a dictionary which contains the key models, as shown below.

OLP_SETTINGS = {
    "models": (
    ),
}

The models key should contain a tuple of tuples containing the string path to the model and the queryset filter used to filter by a user.

Example

example/settings.py

OLP_SETTINGS = {
    "models": (
        ("django.contrib.auth.models.Group", "users"),
    ),
}

example/models.py

from django.db import models

class Apple(models.Model):
    owner = models.ForeignKey("auth.User")

    class Meta:
        permissions = (
            ("can_see_apple", "User can see the apple."),
        )

python shell

>>> from olp.utils import patch_models
>>> patch_models()
>>> from example.models import Apple
>>> from django.contrib.auth import Group, User
>>> user = User.objects.all()[0]
>>> group = Group.objects.all()[0]
>>> apple = Apple(owner=user)
>>> apple.save()
>>> user.has_perm("example.can_see_apple", apple)
False
>>> user.assign_perm("can_see_apple", apple)
True
>>> user.has_perm("example.can_see_apple", apple)
True
>>> user.remove_perm("example.can_see_apple", apple)
True
>>> user.has_perm("example.can_see_apple", apple)
False
>>> group.has_perm("example.can_see_apple", apple)
False
>>> group.assign_perm("example.can_see_apple", apple)
True
>>> group.has_perm("example.can_see_apple", apple)
True
>>> user.has_perm("example.can_see_apple", apple)
True

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distribution

olp-0.3.0-py2.py3-none-any.whl (6.8 kB view hashes)

Uploaded Python 2 Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page