Re: [boost] Interest in a small array count facility?

From: Gennaro Prota (gennaro_prota_at_[hidden])
Date: 2006-06-29 08:02:59

On Wed, 28 Jun 2006 23:38:25 -0400, David Abrahams
<dave_at_[hidden]> wrote:

>Gennaro Prota <gennaro_prota_at_[hidden]> writes:
>
>> Nothing.

Actually I implemented the pp-based solution yesterday. It's nice and
absolutely error-proof. But there's a serious drawback. Can you see it
off-hand? (I missed it, in advance, or I would have not written it)

> I didn't say it is wrong, just that I can't (still) see an
>> easier template-only approach for the multidimentional case. It would
>> help if you could post your simpler solution. 
>
>I never solved that problem... until today.  The enclosed works.  It
>could probably be considerably simplified if you untwist the logic a
>bit.  In particular, you could probably find a way to get rid of
>remove_bounds.  The basic idea here is that the only way to force an
>error in some contexts, with vc6, is at the top level, with
>eval0<...>::not_array being a nonexistent type.

I'll see what I'll come up with. Thanks for the hints. I like a lot
the names, by the way: sizeof dimension<n>(a) is great :)

--Gennaro.