Re: [boost] Overloading a function taking a boost::function

From: Dirk Gerrits (dirkg_at_[hidden])
Date: 2002-05-12 17:40:22

----- Original Message -----
From: "David Abrahams" <david.abrahams_at_[hidden]>
To: <boost_at_[hidden]>
Sent: Sunday, 12 May, 2002 22:09
Subject: Re: [boost] Overloading a function taking a boost::function

> before this goes too much further, please see:
>
> boost/python/detail/arg_tuple_size.hpp
> boost/python/preprocessed/arg_tuple_size.hpp
>
> Enjoy,
> Dave

----- Original Message -----
From: "Douglas Gregor" <gregod_at_[hidden]>
To: <boost_at_[hidden]>
Sent: Sunday, 12 May, 2002 23:54
Subject: Re: [boost] Overloading a function taking a boost::function

> It isn't a partial specialization problem, because it's easy to pick out
the
> arity of the function pointer or member function pointer without partial
> specialization. See what Dave Abrahams posted:
>
> ...
>
> Doug

Cool. This arg_tuple_size seems to me as having enough functionality to
create
the discussed Boost.Function constructor. (With some additional
specializations/
overloads for things like boost::function, std::unary_function and
std::binary_function.)

So, is this anything to be expected in the next version of Boost.Function?

----- Original Message -----
From: "Paul Mensonides" <pmenso57_at_[hidden]>
To: <gregod_at_[hidden]>; <boost_at_[hidden]>
Sent: Monday, 13 May, 2002 0:07
Subject: Re: [boost] Overloading a function taking a boost::function

> But that only says how many parameters there are not there types.  Using
the
> sizeof trick won't work for a complete function_traits class.
>
> Paul Mensonides

Very true. But as was said to me earlier in this thread, for some function
objects
not all the parameter types can be determined. (Some results of a
boost::bind for
instance.)

Dirk Gerrits