[Python-Dev] Re: Candidate Function Decorator (original) (raw)

Phillip J. Eby pje at telecommunity.com
Tue Apr 13 13:58:08 EDT 2004

At 01:48 PM 4/13/04 -0400, Raymond Hettinger wrote:

> I have the impression that it is proposed that the function f will not > yet be bound to its name (f.funcname) in its environment at the time the > decorator is called.

Hmm, that would be a change. Currently, staticmethod() and classmethod() take fully formed functions as inputs. So, if your impression is correct,

It is. See PEP 318, under "Current Implementation". Also, several proposed decorators such as propget/propset/propdel and "generic" rely on this distinction for their magic.

there would be a small semantic difference between the [classmethod] prefix and f=classmethod(f) postfix decoration.

Yes, and that's the desired semantics, as it makes possible many things that are quite awkward with the existing syntax.

More information about the Python-Dev mailing list