On Mon, 28 Mar 2022 at 10:41, Steven D'Aprano <steve@pearwood.info> wrote:
On Sun, Mar 27, 2022 at 01:33:05PM -0000, malmiteria wrote:
C3 linéarization might be mathematically proven to be optimal, as in, it's the best way to order a graph. Sure. But why would we order the inheritance tree?
How else are you going to get inheritance without a linear order? The interpreter can only call superclass methods one at a time, in some linear order. You might have a tree that looks like this in part:
# class C(A, B): ...
A B \ / \ / C
but C.method() cannot call A.method() and B.method() *simultaneously*, it must call one followed by the other in *some order*.
"Linearizing" applies to the whole tree. An alternative would be to consider the immediate parents of this class only, not of any other, which could be done by having super return all of the corresponding attributes. This would work nicely for some things, and very very badly for others. It shouldn't be called super, but it absolutely could be called parents or something, and it'd be used in a completely different form of cooperative multiple inheritance - instead of each level calling super to pass control to the next one, each one would do its own thing and put stuff into a shared collection or something. There are lots of ways things can be done, but Python's super is, as the OP says, tied closely to the concept of linearizing the entire tree and walking the MRO. ChrisA