Re: [relax-ng-comment] conformance test

[James Clark <jjc@jclark.com>]

----- Original Message -----
From: "Gary Stephenson" <garys@ihug.com.au>
To: <relax-ng-comment@lists.oasis-open.org>
Sent: Wednesday, November 14, 2001 12:22 PM
Subject: Re: [relax-ng-comment] conformance test


> Hi all,
>
> Obviously, the same problem applies to many of the other restricitons
also.
>
> Perhaps the definition of "descendant" in section 7.1. should be expanded
to
> explicitly state that the descendants of a ref element are in fact the
> descendants of the corresponding define element.  I assume that is the
> intent - yes?
>
> cheers,
>
> gary

That's NOT the intent. Descendant means child or child of child or ...


>
> ----- Original Message -----
> From: "Gary Stephenson" <garys@ihug.com.au>
> To: <relax-ng-comment@lists.oasis-open.org>
> Sent: Wednesday, November 14, 2001 3:55 PM
> Subject: Re: [relax-ng-comment] conformance test
>
>
> > Hi James,
> >
> > > I think this comment identifies a bug in the spec. Section 4.20 (the
> section
> > > on notAllowed) should say that any definitions that are made
unreachable
> by
> > > the simplification of notAllowed are removed.
> >
> > Great.  On a somewhat similar note, in Section 7.1.1 it says:
> > "The following paths are prohibited:
> >
> >   a.. attribute//element
> >   b.. attribute//attribute"
> > But after simplification, should not the element have been tranformed to
a
> > ref? - due to Section 4.18, whereby
> >
> > "the grammar is transformed so that every element element is the child
of a
> > define element"
> >
> > So should the prohibited path be changed to:
> >
> > attribute//ref
> >
> > or similar?
> >
> > cheers,
> >
> > gary
> >
> >
> >
> >
> > ----------------------------------------------------------------
> > To subscribe or unsubscribe from this elist use the subscription
> > manager: <http://lists.oasis-open.org/ob/adm.pl>
> >
>
>
> ----------------------------------------------------------------
> To subscribe or unsubscribe from this elist use the subscription
> manager: <http://lists.oasis-open.org/ob/adm.pl>
>
>
>