[tei-council] Conference call Nov. 29

[Syd Bauman]

> oh, sorry, I didn't realize you meant you needed a solution to handle
> multiple targets.

No apologies -- you were right not to realize it, it wasn't really
part of the original request. :-)


> but thats the same as it [handling IDREFS with exslt or waiting for
> XSLT2] is now, isn't it?

I think so, but I'm not sure. (Which is why I'm asking.)


> a quick "normalize-space()" around the @target should clear that up

Ah, OK. Thanks.


> <xsl:key name="IDS" match="*" use="@xml:id"/>

OK, thanks.


> no; if no mode is defined, it uses default rules (which isn't
> what you want). so you have a fallback
>   <xsl:template match="*" mode="xref">

OK, thanks.


> apart from the substring-after thing, this is exactly what I'd
> have done for P4 too.

Roger, that's the information I was looking for. Thanks.


> but wasn't this all discussed in SO? when you agreed on the
> syntax, you must have had some ideas of the algorithms for
> processing it?

Yes and no. Let's discuss it over a beer (orange Fanta :-) someday. 
 

> by the way, remember that the Relax NG processor is orthogonal. The
> fact that the schema says "a b c " is a "list" has no impact on
> XSLT processing. To the XML parser "a b c" is "a b c", no more no
> less. A processor might read the schema to assist in the
> processing, but the XML parser won't be offering help.

Right, good point. Thanks for reminding me.