Greg Ewing wrote:
> Ian Bicking wrote:
>
>> ** can't be changed in this way, either -- it really has to enumerate
>> all the keys in the thing passed to it,
>
> In Py3k this could perhaps be different in the case where
> you're calling a function defined like
>
> def f(*args, **kwds):
> ...
>
> In that case there are no named arguments to look up
> in the dict, so when it's called using
>
> f(**obj)
>
> the obj could be passed straight through to the kwds
> argument.
I thought of that, but that seemed much too clever. For instance, if
you then change the signature to:
f(a, **obj)
Then it doesn't work anymore, because you have to test if a keyword of
'a' was passed in. ** in the signature and ** in the call are really
not very symmetric.
With current substitution this isn't that big a deal, because you have
either positional *or* named markers (or for string.Template positional
are not allowed), and if you see any named marker you assume you were
passed a dictionary-like object. Some of the proposals that mix {1}
with {name} in the same substitution are harder. Also, they have to
look at the name of the marker and see if it is an integer, and thus a
dictionary with a key like '1' cannot be used.
--
Ian Bicking | [EMAIL PROTECTED] | http://blog.ianbicking.org
_______________________________________________
Python-3000 mailing list
Python-3000@python.org
http://mail.python.org/mailman/listinfo/python-3000
Unsubscribe:
http://mail.python.org/mailman/options/python-3000/archive%40mail-archive.com
