Re: [boost] [result_of][fusion] more

From: João Abecasis (jpabecasis_at_[hidden])
Date: 2006-05-30 19:25:34

David Abrahams wrote:
> João Abecasis <jpabecasis_at_[hidden]> writes:
>> 1 - support for function pointers was completely broken. Specifically, I 
>> learned that result_of doesn't handle cv-qualified function pointers.
> 
> Shouldn't we fix result_of?

I think so. I posted about the issues I found with result_of and 
is_pointer separately.

>> 2 - Random Access Sequences were required when Forward Sequences would 
>> have been sufficient. Here I failed to grasp the Fusion Concepts while 
>> porting unpack_args from Fusion v1.
> 
>> 3 - Function references weren't handled at all.
> 
> A result_of problem, or something else?

In this case, something else, in my naive implementation I thought 
everything would be solved with a simple:

     typedef result_of<F(A0, ..., AN)>::type result_type;

But it turns out that if F is a function the compiler will complain that 
we're trying to write a function returning a function. In the end, what 
I did, specifically for this case, was to add a pointer to the function 
type.

Here I also found an issue with cv-qualified functions (if they do 
exist...) and is_function. This one seems to be gcc-4.x specific.

Best regards,

João