Re: [xslt] Non-existent variable => error?

[Daniel Veillard <veillard redhat com>]

On Fri, Oct 04, 2002 at 09:10:22PM -0700, Jerome Pesenti wrote:
> 
> --- "Vakoc, Mark" <Mark_Vakoc@jdedwards.com> wrote:
> > 
> > > Now, after taking a second glance, what do we do
> > about it? Confirm that
> > > registering a generic error handler is the only
> > way to find out an error
> > has
> > > occured? Or modify the code to discard everything
> > the library has done
> > > before the error occurred and return NULL (speak:
> > patches are welcome
> > :-))?
> > 
> > My quick vote is to return NULL.  I consider a
> > stylesheet like XML in the
> > sense that something is either 100% valid, or it's
> > not.  Using a
> > unregistered variable is a rather fatal flaw --
> > since libxslt catches this,
> > I don't think it would be wrong to return NULL, and
> > people can rely on the
> > error handler to figure out WHY it happened.
> 
> I vote for NULL too.... especially because the error
> handler for the XSL is currently a global (hopefully
> this will change soon.... as asked by Daniel a while
> ago a posted a bug report on this) which makes the use
> of LibXSLT in a multithreaded application rather
> akward....

  Okay, Mark or someone logs that in bugzilla so I'm sure I keep
track of this and fix it :-)

Daniel

-- 
Daniel Veillard      | Red Hat Network https://rhn.redhat.com/
veillard@redhat.com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/