Re: [boost] [Containers] Should flat_* expose implementation vector?

Subject: Re: [boost] [Containers] Should flat_* expose implementation vector?
From: Scott McMurray (me22.ca+boost_at_[hidden])
Date: 2011-08-05 22:35:36

On Fri, Aug 5, 2011 at 19:05, Michel MORIN <mimomorin_at_[hidden]> wrote:
>
> Or, I would want `uncompared_push_back` (just filling the flat_* container
> without sorting) and `sort` functions.
>

Can we do this with move semantics instead?

I picture putting a bunch of data into a
flat_set::internal_storage_type, then move-constructing a flat_set
from that container, which would do the sorting to fix up the
invariants.

It would also make sense to have a flat_set::release() function that
returns a (moved) temporary of the internal storage.