Re: [XBL] Conformance

On Sat, 13 Jan 2007, Cameron McCormack wrote:
> > 
> > It doesn't mean they have to expand them, it merely means that they 
> > must be treated as if they were expanded.
> > 
> > Could you give an example of what you think this leaves unclear? I 
> > don't really see a problem with this text. I'm not really sure how 
> > else to phrase it.
> 
> Say you had:
> 
>   <!DOCTYPE blah SYSTEM "blah.dtd">
>   <?xbl href="bindings.xbl"?>
>   <blah>
>     <something>&abc;</something>
>   </blah>
> 
> and bindings.xbl had:
> 
>   <xbl xmlns="…">
>     <binding element="something">
>       <template>
>         <something-else xbl:attr="def=xbl:text"/>
>       </template>
>     </binding>
>   </xbl>
> 
> If it’s not a validating parser, blah.dtd might not be fetched, and 
> the &abc; would stay in the document as an entity reference node.  But 
> the UA “must handle them [the entity references] as if they were fully 
> expanded”, which it can’t do unless it actually does fetch blah.dtd, 
> because it can’t know if &abc; would expand to some text nodes.

Oh, I see what you're saying.

That's covered by DOM3 Core, which says "When an EntityReference node 
represents a reference to an unknown entity, the node has no children and 
its replacement value, when used by Attr.value for example, is empty".

See:

   http://www.w3.org/TR/DOM-Level-3-Core/core.html#ID-11C98490

HTH,
-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'