Boost logo

Boost :

Subject: Re: [boost] [fusion] repetitive_view
From: Stjepan Rajko (stjepan.rajko_at_[hidden])
Date: 2009-04-12 16:47:57


On Sat, Apr 11, 2009 at 5:53 PM, Joel de Guzman
<joel_at_[hidden]> wrote:
> Stjepan Rajko wrote:
>>
>> Hello,
>>
>> I just noticed `repetitive_view` in boost/fusion/view.  It doesn't
>> appear to be documented, but judging from the test, it appears that
>> repetitive_view<Sequence> is an infinite repetition of the elements of
>> the sequence.  Is that correct?
>>
>
> That is Dan's baby. You should probably ask him. I think a reason
> why it's undocumented is that the basic concepts of a "sequence"
> requires a size. An infinite view does not have a size. We'll
> have to refine the concepts to take in infinite views. I do like
> your lazy_sequence idea. Please coordinate with Dan on how to
> proceed with these infinite views.
>

OK, I'll contact Dan.

I haven't worked any further on the lazy_sequence idea. I took
Steven's suggestion from the end of the thread and just used a
transform from mpl::range_c<int, 0, N> instead. But after using that
a few times, I found that at least adding some syntactic sugar would
be beneficial. So maybe providing lazy_sequence as a thin wrapper
around a transform from mpl::range_c<int, 0, N> would be the way to
go. And maybe the default value for N is the maximum possible, and an
"infinite" sequence is just a finite sequence with maximum possible
size (that can be held by int)?

Best,

Stjepan


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk