Mutable defaults

[J. Pic]

Ok, maybe:

def foo(x=:[], y=:len(x)):

As a shorthand for:

@default(x=lambda: [], y=lambda x: len(x))
def foo(x=None, y=None):

=: Is the contraction for =lambda:

At the same time, this new syntax avoid breaking def foo(x=lamba: []) which
is completely different as we know.

Le jeu. 11 févr. 2021 à 04:03, Chris Angelico <rosuav at gmail.com> a écrit :

> On Thu, Feb 11, 2021 at 1:55 PM J. Pic <jpic at yourlabs.org> wrote:
> >
> > Adding decorators with some inspect sauce could certainly work with the
> syntax we already have:
> >
> > @default(x=lambda: copy([]), y=lambda x: len(x))
> > def foo(x=None, y=None):
>
> This would work, although the copy is unnecessary here. But you're
> assuming that it magically figures out whether to pass 'x' or not.
> Seems awkward. Also, it doesn't solve the problems of signature
> display, so you're not really much better than:
>
> def foo(x=None, y=None):
>     if x is None: x = []
>     if y is None: y = len(x)
>
> > I think this PoC is doable. But the lambda copy is a boring so, two
> decorators:
> >
> > @default.copy(x=[])
> > @default.call(y=lamba x: len(x))
> > def foo(x=None, y=None):
> >
> > Or with a new syntax:
> >
> > def foo(x:=[], y`=len(x))
>
> Now this is the sort of thing I was talking about: actual syntactic
> support. But this syntax looks pretty terrible, and if you propose it
> like this, I doubt it'll be accepted. Backtick had a very very
> different meaning in Py2, and := is going to be extremely confusing.
>
> This proposal will stand or fall on the syntax. Don't just give
> placeholder syntax. Put some thought into figuring out what would
> actually be good, and propose *that*. The semantics aren't really even
> that hard - it'd be equivalent to an if statement just inside the
> function.
>
> ChrisA
> --
> https://mail.python.org/mailman/listinfo/python-list
>