RE: [boost] Re: Design for a file system library

From: Jeff Garland (jeff_at_[hidden])
Date: 2002-03-07 17:23:12

> >I think this could be done portably and still no global state.  Of course, no
> >guarantees if someone mounts a new drive or changes the current working
> >directory after iteration has begun....
>
> i think this is a better solution. but it only works for roots, because
> i want to have the value of the current_directory and not its entries.

Yes, that could be a problem since the 'root' does not have a valid 'value' in
what I'm proposing.  BTW, I'd still like to be able to construct the iterator
this way without having me (the client) having to look up the current root or
current working directory directly :-)

Jeff