Re: Ruby Method Signatures (was Re: Multiton in standard library)

On 3/23/07, TRANS <[e-mail]> wrote:
> On 3/23/07, Rick DeNatale <[e-mail]> wrote:

> > On reflection though, I don't think that this is a good idea anyway.
> > Given that Ruby allows dynamic re-definition what should this do?
> >
> > def foo(x)
> >    end
> >
> >  def bar(*=method(:foo).signature)
> >      p x   # or  p a1
> >   end
> >
> >   bar(1)  #=> 1
> >
> >   def foo(y)
> >   end
> >
> >   bar(1) #=> ?
> >
> > I suppose that the answer is 1, and that the signature is just used as
> > it is when bar is defined, but then I'm not sure that this is really a
> > useful feature, it certainly doesn't make the definition of bar EASIER
> > to read.
>
> Thats' true. The real need though is just so we can define method
> wraps with the same arity. That's the useful part. How else could we
> go it?

Something like this might work, the only glitch is methods like
def foo(d=1)  which have an arity of -1 indicating that they  have a
variable number of arguments, but this doesn't indicate that they have
an upper limit on the number of arguments.

But I think that it gets close:

module PlistGenerator
  def plist
    # the arity is a positive number if all arguments are required
    # otherwise it is -1-number_of_required arguments
    # n = -1 - required
    # required + n = -1
    # required = -1 -n
    n = arity
    req_ct = n >= 0 ? n : -1 - n
    args = (1..req_ct).map {|num| "a#{num}"}
    args << "*opt" if n < 0
    args.join(',')
  end
end

class Method
  include PlistGenerator
end

class UnboundMethod
  include PlistGenerator
end

-- 
Rick DeNatale

My blog on Ruby
http://talklikeaduck.denhaven2.com/