Issue extending a delegated class

Ruby
1

Hi,

I have a simple class that I’m extending via delegation:

class A
def one ; puts “A::one” ; two ; end
def two ; puts “A::two” ; end
end

class B < SimpleDelegator

def initialize
a = A.new

# do some stuff to a

__setobj__ a
super( a )

end
end

I then instantiate B, and use that, all is well. However, i’m running
into
issues whenever I try to override a method in B:
b = B.new
class << b
def two ; puts “foo” ; end
end

I would have expected b.one to output:
A::one
custom

But instead i get:
A::one
A::two

However, calling the overriden method directly produces the expected
result. I’m guessing the extend is really only adding the function to
the B
instance and not to A? Is there a way to get to A?

Thanks,
-rak-

2

Hi –

On Thu, 27 Jul 2006, rak rok wrote:

I then instantiate B, and use that, all is well. However, i’m running into
issues whenever I try to override a method in B:
b = B.new
class << b

You’re not overriding anything in B here; rather, you’re adding an
instance method to the singleton class of the instance b.

def two ; puts “foo” ; end
end

I would have expected b.one to output:
A::one
custom

(Do you mean “foo” rather than “custom”?)

But instead i get:
A::one
A::two

However, calling the overriden method directly produces the expected
result. I’m guessing the extend is really only adding the function to the B
instance and not to A? Is there a way to get to A?

Your guess is right :slight_smile: When a handles the delegate call to #one,
part of that call is a call to #two – which, as far as a is
concerned, is A#two.

You can always override A#two, but I imagine that’s not what you have
in mind. I’m not sure exactly what you are trying to do, though.

David

3

Hi David,

I would have expected b.one to output:
A::one
custom

(Do you mean “foo” rather than “custom”?)

(Woops, yes i meant “foo”)

Your guess is right :slight_smile: When a handles the delegate call to #one,
part of that call is a call to #two – which, as far as a is
concerned, is A#two.

You can always override A#two, but I imagine that’s not what you have
in mind. I’m not sure exactly what you are trying to do, though.

Makes sense… But then how is it possible to override A#two if all I
have is an instance of B?

Thanks!
-rak-

4

Hi –

On Thu, 27 Jul 2006, rak rok wrote:

Your guess is right :slight_smile: When a handles the delegate call to #one,
part of that call is a call to #two – which, as far as a is
concerned, is A#two.

You can always override A#two, but I imagine that’s not what you have
in mind. I’m not sure exactly what you are trying to do, though.

Makes sense… But then how is it possible to override A#two if all I
have is an instance of B?

You really want b’s delegated object to have a new #two – so:

class << b.getobj
def two
puts “custom”
end
end

or something along those lines.

David

5

You really want b’s delegated object to have a new #two – so:

class << b.getobj
def two
puts “custom”
end
end

Awesome, I didn’t know about getobj, thanks a lot.

-rak-

PS. Thanks for a great book!