Skip to content

notes.md

Latest commit

 

History

History
16 lines (13 loc) · 1.73 KB

notes.md

File metadata and controls

16 lines (13 loc) · 1.73 KB
Raw

Chapter 24: Class Metaprogramming — Notes

1. What is a Metaclass?

Just as an object is an instance of a Class, a Class is an instance of a Metaclass. If you type type(7), it returns <class 'int'>. If you type type(int), it returns <class 'type'>. type is the default metaclass that Python uses to build all classes. A custom metaclass is simply a class that inherits from type and overrides __new__ or __init__ to intercept the class-creation process.

2. Evaluation Time vs Execution Time

  • Evaluation Time (Import Time): When Python first imports a module, it reads it top-to-bottom. It executes all top-level code. Crucially, it executes the body of every class definition, builds the class dictionary, and calls the Metaclass/Class Decorators to construct the Class Object in memory.
  • Execution Time (Runtime): This happens when the user actually invokes things. Calling obj = MyClass() triggers __init__. Calling obj.do_work() triggers the method. Metaprogramming logic almost exclusively runs at Evaluation Time.

3. The Three Tiers of Metaprogramming

  1. Class Decorators (@decorator): Evaluated after the class is fully built. Great for attaching new attributes or wrapping the class.
  2. __init_subclass__: A modern (Python 3.6+) feature. Placed in a parent class, it is automatically triggered whenever a subclass is defined. It is the perfect, safe replacement for Metaclasses for 90% of use cases (like automatic plugin registries).
  3. Metaclasses: Evaluated during class creation. They can physically intercept the raw dict of variables and methods before the class object even exists in memory. This is the nuclear option. Use only for massive framework architecture (like Django ORM).