Subject: Re: [boost] [ boost ] [ Trie ]
From: Kenneth Adam Miller (kennethadammiller_at_[hidden])
Date: 2015-03-06 13:53:35
Yeah, but then you didn't create just one node, you created the entire trie.
On Fri, Mar 6, 2015 at 1:37 PM, Cosmin Boaca <boost.cosmin.boaca_at_[hidden]>
> On 6 March 2015 at 20:32, Kenneth Adam Miller <kennethadammiller_at_[hidden]
> > Well, just conjecture, but I would guess that any one trie_node would not
> > make much sense outside of the context. As in, you have to have a trie
> > sense with respect to it's surroundings. Say you use either the tree
> > representation or a contiguouos array. A copy of a single element would
> > imply that only a few bytes from the contiguous array are removed. You
> > can't do indexing or any of the other magic that a trie is meant to do
> > only one of them.
> I don't get it. You can just create another node by copying it member with
> Unsubscribe & other changes:
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk