Re: [xml] "double"s and schema validation

[Daniel Veillard <veillard redhat com>]

On Wed, Jul 28, 2010 at 03:52:44PM +0300, Andrew W. Nosenko wrote:
> On Wed, Jul 28, 2010 at 14:50, Csaba Raduly <rcsaba gmail com> wrote:
> > On Wed, Jul 28, 2010 at 11:44 AM, Daniel Veillard  wrote:
> > > On Thu, Jul 22, 2010 at 02:38:16PM +0200, Csaba Raduly wrote:
> > > > Anyway, here's a revised version:
> > > > --- xmlschemastypes2.c  2010-07-21 13:17:12.229467800 +0200
> > > > +++ xmlschemastypes.c   2010-07-22 14:00:05.965759600 +0200
> > (snip)
> > > > +                if (digits_before + digits_after == 0)
> > > > +                    goto return1;
> > >
> > >  Okay that patch looks fine, except for mail mangling and code format,
> > > so I applied it manually, look fine, passes regtests, so pushed to git,
> >
> > I think Andrew Nosenko is right, it would be better to test these
> > separately and not added together:
> >
> > if (digits_before==0 && digits_after==0)
> >  goto return1;
>
> Excuse me, but my point was not about checking counters separately,
> but about there no need counters at all -- single boolean flag is
> fairly enough.

  Honnestly it's not a big deal, but I prefer explicit 0 checks rather
than the addition hack, it's nicer to read. Actually in such code I
think readability is more important than trying to optimize, the
compiler will do it just fine, and that path ain't critical. On the
other hand the maintainability is a serious concern,

  thanks to both of you :-)

Daniel

-- 
Daniel Veillard      | libxml Gnome XML XSLT toolkit  http://xmlsoft.org/
daniel veillard com  | Rpmfind RPM search engine http://rpmfind.net/
http://veillard.com/ | virtualization library  http://libvirt.org/