Re: [xml] Questions about xml:id

[Daniel Veillard <veillard redhat com>]

On Tue, Mar 22, 2005 at 08:17:23AM -0500, Rob Richards wrote:
> Daniel Veillard wrote:
>
> > need to be extremely cautious about that part of the code because it
> > affects streaming DTD validation for ID/REF(S) checks, even when the node
> > have been removed. 
> > But in general the code need fixing for correct ID/xml:id support when
> > mutating trees. Not sure I will have time looking at it now, but if you're
> > interested I would take patches of course !
> From what I can discern, SAX wouldn't have an issue as it is either 
> building the tree or just passing name/values. Would there ever be a 
> case where you could have the doc->ids built with no tree? This seem 

  xmlreader, when I said streaming I wasn't thinking of SAX.

> like the only way you could end up with an attribute with no element. As 
> far as the reader goes, it has its own FreeProp function so it should 
> never call xmlFreeProp. In the event you grab an expanded node, can the 

  right but one need to double check there.

> xmlFreeNode functions be used or is it mantatory to use the reader free 
> node functions?

  if you modify a subtree grabbed from the reader you're very likely in
anyway to crash the parser or library, people should not do this !

> As far as trying to fix this area, see attached file for an idea of what 
> I think is a more appropriate direction for this. I believe xmlIsID is 

  using atype might work, yes. I see the idea but the patch looks far from
complete as is (I suppose it's expected).

> used incorrectly when mutating trees when trying to determine if an 
> existing attribute is an id or not. xmlIsID/xmlIsRef should really only 
> be used when determining whether an attribute should be created as an ID 
> or IDRef, but not for testing whether an attribute is really an ID or IDRef.

  but what if people already used that function for that purpose in their code ? 
> Last question on this, as I couldn't find an answer for this anywhere, 
> is when an attribute which is an ID is unlinked from an element, though 
> still is associated to a document, is the attribute still an ID or does 
> it revert to a normal attribute at that point?

  I can't answer either, the safest would be to revert it. From a DTD
perspective it's the (elem name, att name, document) which defines if
it is of type ID, but with xml:id it's just the name which defines it.
So the rule will have to be adapted depending on this.

Daniel

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