Re: Making sort a method of Array<T>, Sparse<T> and octave_value classes

[David Bateman]

John W. Eaton wrote:
> On 31-Jan-2008, David Bateman wrote:
>
> | I see no reason not to do, but there are many ways to do it, and maybe
> | you don't agree with the one I picked.. Since if this patch is accepted
> | it'll be the model for how the others are converted to behave in a
> | similar manner it makes sense to get the way its implemented to be the
> | closest to what you'd like to see.
>
> It seems OK to me.
>
> One thing I don't quite get is the purpose of NO_INSTANTIATE_ARRAY_SORT.
>   
The Array<idx_vector> was giving me problems without it as the
Array<idx_vector>::sort method of Array<T>::sort was instantiated and
that had a default sort based on the > operator that is not defined for
Array<idx_vector>. Maybe there is a better way than this to avoid the
instantiation.

> | Note that I have a new version of this patch that adds tests for cell
> | string arrays and int -> octave_idx_type changes in oct-sort.{cc,h} for
> | 64-bit platforms, though that doesn't change the basic idea of the patch
> | and so I won't send it to the list to avoid mail bombing everyone.
>
> If you think it is complete (or close), then please check it in.
>   
Ok, will do later tonight.

D.

> Thanks,
>
> jwe
>
>   


-- 
David Bateman                                address@hidden
Motorola Labs - Paris                        +33 1 69 35 48 04 (Ph) 
Parc Les Algorithmes, Commune de St Aubin    +33 6 72 01 06 33 (Mob) 
91193 Gif-Sur-Yvette FRANCE                  +33 1 69 35 77 01 (Fax) 

The information contained in this communication has been classified as: 

[x] General Business Information 
[ ] Motorola Internal Use Only 
[ ] Motorola Confidential Proprietary