pryvacy - a set of access control decorators for python

Philosophy

Incur as least overhead as possible when using the access control decorators
Only classes that use the decorators need to be changed, other related classes do not need to -> Can be opted-in easily

Installation

pip install pryvacy

Usage

The package provides 3 access control decorators: @private, @public, @protected that can be used on methods and nested classes (currently only @private can be used on nested classes)

from pryvacy import pryvacy, private, public, protected

@pryvacy
class Foo():
  @public
  def public_method(self):
    pass

  @protected
  def protected_method(self):
    pass

  @private
  def private_method(self):
    pass

Access control rules:

Methods inside Foo are able to access public_method, protected_method, private_method.
Code outside Foo and not inside any Foo's subclass methods can only access public_method.
Methods inside Foo's subclasses (either decorated with @pryvacy or not) can access public_method and protected_method.
Nested classes methods can access public_method, protected_method, private_method.

Pitfalls & Bugs

Disclaimer: The package has not been tested thoroughly! Use with caution! Any contributions are appreciated~

Currently, class-level code cannot access protected_method and private_method. Example:

    class Foo():
        ...
        class Bar(Foo):
            Foo().public_method() # OK!
            Foo().protected_method() # Exception!
            Foo().private_method() # Exception!

The test_back_and_forth test is failing. This scenario is unlikely to happen but I will try to come up with a performant solution.

Current limitations

@private and @protected are not supported on nested classes yet.
No way to enforce access policy on class and instance attributes.

Roadmap

Benchmark decorated classes
Test comprehensively the decorators interaction with the whole ecosystem
Implement @private and @protected on nested classes