|
Boost : |
Subject: Re: [boost] [program_options] multitoken arguments
From: Carlo Medas (carlomedas_at_[hidden])
Date: 2009-04-06 04:13:30
Thanks Volodya,
we've been using latest boost 1.38.0 (testing it on Win, Linux, Mac).
I checked SVN HEAD but it seems to me that there hasn't been recent activity
after 1.38.0, or am I missing something?
Thanks in advance.
Kind regards,
\Carlo
On Fri, Apr 3, 2009 at 3:44 PM, Vladimir Prus <vladimir_at_[hidden]>wrote:
> Carlo Medas wrote:
>
> > Dears,
> > we've been using boost::program_options and it works great.
> >
> > We just found a minor thing, which could be an issue, and I'm reporting
> it
> > to you: when specifying and using a multitoken argument, it actually
> works
> > only if put as last parameter of command line; in any other case, it eats
> > all remaining parameters.
> >
> > My specification is as follows:
> >
> > desc.add_options()
> > ("output,o", bpo::value<std::vector<std::string> >()->multitoken(),
> "use
> > output setup for discovery/scan");
> >
> > If then I want to use it like follows, it works:
> >> fing -p1 -p2 -p3 -o 1 2 3
> >
> > But if the user inverts the order, it does not work anymore, meaning that
> p3
> > is eaten by -o option.
> >> fing -p1 -p2 -o 1 2 3 -p3
> >
> > Is this a known limitation?
>
> Yes. Can you try SVN HEAD version -- it should have this fixed.
>
> - Volodya
>
>
> _______________________________________________
> Unsubscribe & other changes:
> http://lists.boost.org/mailman/listinfo.cgi/boost
>
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk