[Home] [By Thread] [By Date] [Recent Entries]



----- Original Message ----- 
From: "Toni Uusitalo" <toni.uusitalo@p...>


> At 14:38 26.1.2004, you wrote:
> >Toni Uusitalo:
> > > At 00:17 26.1.2004, you wrote:
> > > >That is, I agree that this is  a practical approach and have a place 
> > to plug
> > > >it in.
> > >
> > > Another practical approach might be this:
> > >
> > > Defining "XML writer services" API for XML parsers, starting for example
> > > from expat,
> > > API allows to ask parser for qualified names, namespace scopes etc. thus
> >
> >But expat does have this API already.
> 
> I was thinking of something like defining public API for
> int XmlWriterService_CheckName(XMLWriterService XWs, const utf8Byte *s);
> and
> int XmlWriterService_CheckUri(XMLWriterService XWs, const utf8Byte *s);
> that would call checker routines in the parser API thus
> GenX wouldn't have to implement all kinds of XML namechar tables but
> it could utilize already existing ones.
> 
> These kind of functions are not currently part of public expat API.

Correct. Btw, I don't think Expat checks URIs.

In any case, this would be a little difficult to expose,
because these functions are are inlined as macros in 
a way that makes them part of the parsing logic, they
are not callable as separate functions. At a first glance
it is not even easy to see which ones they are.

Karl

Site Map | Privacy Policy | Terms of Use | Trademarks
Free Stylus Studio XML Training:
W3C Member