RE: [boost] Re: new initialization lib, any interest?

From: Yitzhak Sapir (yitzhaks_at_[hidden])
Date: 2002-10-09 06:43:31

On Tue, 8 Oct 2002, Aleksey Gurtovoy wrote:

> >  Also why does list_c allow specifying a type but defaults all
> > parameters to long?
> > Couldn't list_c<class T, T c1, T c2, etc.> have been used?  (Like in
> > integral_c?)
>
> Nope, because due to its variadic interface 'list_c<T,...>' needs to reserve
> some single value for denoting unused arguments, and taking that value out
> of range of possible values of type 'T' is not really a feasible option
> (consider, for instance, T == bool).

This seems to work on VC:

template <class T> struct LargerThan
{
        typedef long type;
        enum { illegal_value = 0xBADBAD };
};

template <> struct LargerThan<bool>
{
        typedef char type;
        enum { illegal_value = -1 };
};

// For pointers, illegal value is 0 (or perhaps, 1)

template <class T, LargerThan<T>::type c1 = LargerThan<T>::illegal_value>
struct list_c
{
...
};

> > And for the pair optimization, I need a "is_pair".  That
> > would have been nice to have in type traits.
> >
>
>     template< typename T >
>     struct is_pair
>         : mpl::logical_and<
>               has_first<T>
>             , has_second<T>
>             >
>     {
>     };
>
>  ;)

What do I include for has_first/has_second?  I couldn't find anything like
it in the type_traits documentation.  Also, if we have an has_first and
has_second, wouldn't it be more complete to have is_pair?