Re: [xml] Potential wrong usage of xmlIsID() in tree.c

[Kasimier Buchcik <K Buchcik 4commerce de>]

Hi,

On Thu, 2006-02-23 at 07:29 -0500, Daniel Veillard wrote:
> On Thu, Feb 23, 2006 at 01:03:24PM +0100, Kasimier Buchcik wrote:
> > The problem occurs regardless of a validation being performed or not.
> >
> > Example:
> >
> > <!DOCTYPE foo [
> >   <!ELEMENT foo (bar)>
> >   <!ELEMENT bar>
> >   <!ATTLIST bar myId ID #IMPLIED>
> > ]>
> >
> > Assume we parse and validate the following XML:
> > <foo>
> >   <bar myID="1"/>
> > </foo>
> >
> > The API wouldn't prevent us to add a <bar> to the <bar>:
> > <foo>
> >   <bar myID="1">
> >     <bar/>
> >   </bar>
> > </foo>
> >
> > If we now add @myID to the newly created <bar>, then the current
> > ID-autodetection would mark it as of type ID:
> > <foo>
> >   <bar myID="1">
> >     <bar myID="2"/>
> >   </bar>
> > </foo>
> >
> > The problem I see here, is that the second <bar> is not valid according
> > to the DTD, and thus its @myID shouldn't become an ID; that's why
>
>   I think this is wrong. Even if the element raised a validity error, the
> IDNESS property of the attribute would still be there.
>
> Daniel

OK. So my argument against IDness detection based on DTDs is not valid.


Regards & thanks for clarification,

Kasimier