I remember in C++, I had to do double-dispatch using the visitor pattern. This is cumbersome, just because each subclass has to have the exact same singly-dispatched code that looks like:

void dispatch(Base& other)
{
   other.dispatch(*this);
}

Is there a nicer way to do this in D, or am I stuck with the same thing?

Sean Eskapp <eatingstaples@gmail.com> wrote:

> I remember in C++, I had to do double-dispatch using the visitor pattern. This
> is cumbersome, just because each subclass has to have the exact same
> singly-dispatched code that looks like:
>
> void dispatch(Base& other)
> {
>    other.dispatch(*this);
> }
>
> Is there a nicer way to do this in D, or am I stuck with the same thing?

Andrei Alexandrescu (yes, the same one) wrote a generic implementation of
multimethods[1] for his book Modern C++ Design[2]. You might want to take
a look at that. If you decide to re-implement it in D, it might also be
worth incorporating in Phobos.

[1]: http://loki-lib.cvs.sourceforge.net/loki-lib/loki/include/loki/MultiMethods.h?view=markup
[2]: http://www.amazon.com/a/dp/0201704315/

-- 
Simen

Sean Eskapp wrote:

> I remember in C++, I had to do double-dispatch using the visitor pattern. This is cumbersome, just because each subclass has to have the exact same singly-dispatched code that looks like:
> 
> void dispatch(Base& other)
> {
>    other.dispatch(*this);
> }
> 
> Is there a nicer way to do this in D, or am I stuck with the same thing?

There isn't really, but you can take out some of the boilerplate with a mixin, which goes a long way:

// The template parameter Visitor is not actually needed, depending on
// what you want to do
mixin template Accept(Visitor)
{
    void accept(Visitor v)
    {
        v.visit(this);
    }
}

class Stuff
{
    mixin Accept!IStuffVisitor;
}



With some ctfe it's also not too hard to take out the boilerplate of creating the visitor interface with a bit of code generation, though I'm not sure if it's worth it.

Sean Eskapp:

> Is there a nicer way to do this in D, or am I stuck with the same thing?

Andrei has recently said no one needs double dispatch (in D) :-) So Andrei will be interested in your use case.

Bye,
bearophile

== Quote from bearophile (bearophileHUGS@lycos.com)'s article
> Sean Eskapp:
> > Is there a nicer way to do this in D, or am I stuck with the same thing?
> Andrei has recently said no one needs double dispatch (in D) :-) So Andrei will
be interested in your use case.
> Bye,
> bearophile

The age-old collision handling problem is how I'm using it.

On 13/02/11 9:32 PM, Sean Eskapp wrote:
> == Quote from bearophile (bearophileHUGS@lycos.com)'s article
>> Sean Eskapp:
>>> Is there a nicer way to do this in D, or am I stuck with the same thing?
>> Andrei has recently said no one needs double dispatch (in D) :-) So Andrei will
> be interested in your use case.
>> Bye,
>> bearophile
>
> The age-old collision handling problem is how I'm using it.

Use a 2D array of function pointers, with the indices coming from unique IDs on the shapes.