Two minor syntactic proposals

[Michael Haggerty]

Greg Ewing <see at my.signature> writes:
> I like this idea too! The way I'd describe it is as allowing a
> general l-value (i.e. anything that can appear on the lhs of an
> assignment) in place of the current identifier-being-bound in def,
> class and import statements.
> 
> (On the downside -- I like the def self.foo() idea too, and this is
> in direct conflict with it! Sigh, can't have everything, I suppose,
> and I think I'd rather have this than def self.foo() if it came down
> to a choice.)

Bengt Richter <bokr at accessone.com> writes:
> I think I prefer just an easier way to type "self."
> How about just the dot? I.e., ".x" means "self.x"

What about combining the ideas: within a class definition or instance
method definition, ".x" means "self.x".  Thus:

class C:
    def .x(arg):
        # defines instance method; "self" would be a magic implicit
        # first parameter (requires special treatment for word "self".

    def C.x(arg):
        # defines class (static) method

    def f(self, arg):
        # Defines instance method (for backwards compatibility)

Michael

-- 
Michael Haggerty
mhagger at alum.mit.edu