Re: [Boost-users] testing for multi_index equality

From: francesco biscani (bluescarni_at_[hidden])
Date: 2008-05-09 19:52:18

On Fri, May 9, 2008 at 9:59 PM, JOAQUIN M. LOPEZ MUÑOZ <joaquin_at_[hidden]>
wrote:

>  ------------------------------
> *De:* boost-users-bounces_at_[hidden] [
> boost-users-bounces_at_[hidden]] En nombre de francesco biscani [
> bluescarni_at_[hidden]]
> *Enviado el:* viernes, 09 de mayo de 2008 21:36
> *Para:* boost-users_at_[hidden]
> *Asunto:* Re: [Boost-users] testing for multi_index equality
> >
> >
>  > Hi Joaquin,
> >
> > does this mean that if the zeroth index is unordered or ordered but
> non-unique you could
> > have a false negative in the equality test due to elements ordering being
> undefined?
>  What do you mean by unordered? If the zeroth index is hashed, container
> equality will fail to compile. In all other cases, equality depends on the
> ordering
> of the elements, which is not strictly determined for ordered_non_unique,
> sequenced and random_access. This might not coincide with your required
> notion for equality. I simply mimicked the modus operandi of STL
> containers.
>

Yes, that's what I meant to say :)

Thanks and best regards,

  Francesco.