From: Sean O'Halpin <sean.ohalpin@...>
Date: 2007-10-23T05:29:29+09:00
Subject: Re: Conflicts between using respond_to? and extending Ruby

------=_Part_11391_31404378.1193084969494
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

On 10/22/07, ara.t.howard <ara.t.howard@gmail.com> wrote:
>
>
> respond_to? + arity it sometimes better.  other alternatices are to
> mark objects with modules
>
>    module Marker; end
>
>    object.extend Marker
>
>    if Marker === object
>
> so you have a more controlled environment - facets could easily use
> this approach.
>

On 10/22/07, ara.t.howard <ara.t.howard@gmail.com> wrote:
>
>
> respond_to? + arity it sometimes better.  other alternatices are to
> mark objects with modules
>
>    module Marker; end
>
>    object.extend Marker
>
>    if Marker === object
>
> so you have a more controlled environment - facets could easily use
> this approach.
>

I've been using something similar to this approach by using modules to
represent sets of behaviours which a class should have. So you end up with
class definitions like this (bogus example):

  class Duck
    include CanFly
    include CanQuack
    include CanSwim
  end

i.e. composing the class definition from the set of behaviours or protocols
it satisfies.

Checking whether an object will fulfil your expectations then becomes:

  if donald === CanQuack

This is a much stronger guarantee of behaviour than :respond_to? and still
qualifies as duck-typing in my opinion (i.e. not based on inheritance).

Regards,
Sean

------=_Part_11391_31404378.1193084969494--