What are Python descriptors and how do they work

Short answer

A Python descriptor is any object that implements __get__, __set__, or __delete__. Descriptors control attribute access and power features like @property, classmethod, and staticmethod.

Details

When you access an attribute, Python checks if the retrieved object has a __get__ method. If so, it invokes that method instead of returning the object directly. This lookup protocol is what makes properties behave like attributes while running custom logic. Writing your own descriptor is useful for validation, lazy initialization, and type-checked attributes. Because descriptors live at the class level, they work closely with Python classes and objects and are often combined with Python decorators for cleaner syntax. Advanced use cases include ORM fields and form validators. If you are exploring metaprogramming, descriptors are a stepping stone to understanding Python metaclasses, which control class creation itself.

Tips

• Store instance-specific data in a WeakKeyDictionary to avoid leaks.
• Implement __set_name__ (Python 3.6+) to capture the attribute name automatically.
• Keep descriptor logic simple; complex behavior belongs in the class or a dedicated service.