Home  >  Article  >  Backend Development  >  How Does Method Resolution Order (MRO) Differ Between Legacy and New-Style Classes in Python?

How Does Method Resolution Order (MRO) Differ Between Legacy and New-Style Classes in Python?

Barbara Streisand
Barbara StreisandOriginal
2024-10-28 05:49:02178browse

How Does Method Resolution Order (MRO) Differ Between Legacy and New-Style Classes in Python?

Method Resolution Order (MRO) Differences between Legacy and New-Style Classes

New-style classes in Python introduced a significant change in the method resolution order (MRO) compared to legacy classes. While the provided example demonstrates MRO resolution correctly for new-style classes, let's delve into the subtle differences that distinguish the two approaches.

The key difference emerges when an ancestor class appears multiple times in the "naive" depth-first resolution order. Consider the following "diamond inheritance" case:

class A:
    x = 'a'

class B(A):
    pass

class C(A):
    x = 'c'

class D(B, C):
    pass

Using legacy classes, the resolution order would be: D - B - A - C - A. In this order, A is encountered first, resulting in the definition of 'x' being hidden by the subsequent definition in C.

D.x => 'a'

However, new-style classes adopt a revised MRO:

D.__mro__ = (<class '__main__.D'>, <class '__main__.B'>, <class '__main__.C'>, <class '__main__.A'>, <type 'object'>)

This order ensures that ancestor classes appear only once and after all their subclasses. Consequently, overrides like C's modification of 'x' take precedence.

D.x => 'c'

By avoiding multiple appearances of ancestor classes in the resolution order, this approach eliminates ambiguities and ensures intuitive behavior in complex inheritance hierarchies. This is why old-style classes should generally be avoided in favor of new-style classes, especially when dealing with multiple inheritance and diamond-like patterns.

The above is the detailed content of How Does Method Resolution Order (MRO) Differ Between Legacy and New-Style Classes in Python?. For more information, please follow other related articles on the PHP Chinese website!

Statement:
The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn