Re: [boost] Named Template Parameters

From: David Abrahams (dave_at_[hidden])
Date: 2006-02-18 14:59:09

Daniel Wallin <dalwan01_at_[hidden]> writes:

> David A. Greene wrote:
>> David Abrahams wrote:
>> 
>>>> BOOST_NTP  (Elemtype);
>>>> BOOST_NTP_C(Bufsize);
>>>> BOOST_NTP_C(Stride);
>>>> BOOST_NTP_C(Iterations);
>>>
>>> I don't know what's under the covers of those macros; we've tried to
>>> make sure that there's a macro-less interface that's at least
>>> reasonably usable, so those would be:
>>>
>>>   template <class T = int>
>>>   struct Elemtype_is
>>>     : parameter::template_keyword<a0_is<>, T>
>>>   {};
>> 
> [...]
>> It's very close to what you specified.  I just find something less
>> wordy easier to comprehend.  I don't know what a0_is<> is, for
>> example.
>
> a0_is<> was a typo, it's supposed to be:
>
>    template <class T = int>
>    struct Elemtype_is
>      : parameter::template_keyword<Elemtype_is<>, T>
>    {};
>
> The point of the T's default value is that we can use Elemtype_is<> as a
> tag-type for the parameter, instead of having a separate tag.

It's a cute trick but I'm not sure we should be doing it.  It permits
the user a usage that I wouldn't want to encourage (i.e. she can leave
out the template argument) and looks a little misleading, as though
we're establishing the default for Elemtype.

>>>     typedef typename parameter::binding<
>>>         params, Elemtype_is<>, int
>>>     >::type ElemtypeP;
>>>
>>>
>>>     typedef typename parameter::binding<
>>>         params, Stride_is<>, mpl::int_<1>
>>>     >::type StrideP;
>> 
>> As I understand it, this both specifies defaults and does the
>> actual mapping between actual arguments and NTP names.
>
> Yes. It's actually quite important to specify the default as late as
> possible, because defaults can depend on the values of other arguments.

A very good point.

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com