landuo (cachedproperty)

just a property but lazier...

Simply put, cachedproperty = property + cached_property.

Version: 1.0.0

Reasons for Creating This

1. In-built property decorator allows you to manage your instances' attributes but at a cost of repeated computation for each read of the attribute.
2. cached_property built by many [1], allows you to cache your instances' attributes but at a cost of losing the ability to implement setter methods of them.

cachedproperty (used interchangably with cached property or its variants) solves this problem by caching of the values of your instances' attributes AND implementing setter methods for each of them.

Objectives

Objective	Description	Status
Backward compatibility	It should be 'easy' (ideally, by adding a one-liner) to existing project to incorporate caching in property.
Caching	Property must be cached. It should allow the use of another 'cache' beyond the instance's __dict__ attribute for instances whose __dict__ attribute is a non-mutable mapping. Read more.	Kinda?
Setter method is available	Cached property should allow for the implementation of .setter(self, value) method.
Deleter method is available	Cached property should allow for the implementation of .deleter(self) method.
Async compatibility	Cached property should be thread-safe. pydanny's cached_property allows for this.	No plans to do that in near future.
Properties should be inherited as it is.	An example would be best here, if self.name returns self._name for a superclass, then for the subclass self.name should still return self._name, UNLESS wilfully overridden, e.g. self.name returns f'Hi, I'm {super().name}, the subclass' managed attribute name prepends 'Hi, I'm '.

Installation

More on Installation

Using pip

With pip installed, you can use the following command to install landuo.

python -m pip install landuo

Check that landuo is installed by importing it in a python console.

Example:

from landuo import property

>>> property
<class 'landuo._lpusedict.BaseMutableCachedProperty'> # Output

How to Use

More on Usage

For Existing Project

If you have an existing project which already uses the python's inbuilt property,

• add landuo to your dependencies,
• add the following declaration into the import declarations in your modules that use the python's inbuilt property.

from landuo import property
#... other import statements

That's it! All your original property should now have caching. By default, the cache used is the instance's attribute __dict__, read more at APIs: cachedproperty.

Glossary

Managed Attribute(s) - The set of attribute(s) of a class which is managed by the cachedproperty / property object. To 'manage' means that the 'read' and/or 'write' capabilities of the attribute is decided by another python object.

Private Variable(s) - Python has no strictly 'private' variable. In this case, the 'private' variable refers to the attribute of a class that is managed by another python object.

Example:

class Person:
    def __init__(self, name):
        self._name = name
        # self._name is referred as the 'private' variable

    @property # `property` manages `self.name` and its underlying private variable is `self._name`
    def name(self):
        return self._name

References

[1] Here is a list of implementations for 'cached_property'.

• functools' cached_property
  
• pydanny's cached_property
  
• penguinolog's cached_property
  
• werkzeug's cached_property
  
• bottles' cached_property
  
• bottles' lazy_attribute
  
• sad2project's LazyProperty
  

Why the Name landuo?

懒惰 (Hàn yǔ Pīn yīn: lănduò) means lazy in Chinese, which describes the cachedproperty class of this package.