David,

thanks for the pointer to xml:id.

While it works perfectly in xsltproc it is not a big help.
The intended use for the duplicate node elimination is
providing a XPath evaluator for browsers.

The support for xml:id in browsers is not that big currently:
http://en.wikipedia.org/wiki/Comparison_of_layout_engines_(XML)#Core_technologies


So these ideas are a dead end, xml:id not supported wide enough, and
DOCTYPE cannot work by your explanation. Therefore id() function
cannot help, too.

I will have to go back to the working solution with intersection and
ghost nodes.


Mit besten Gruessen / Best wishes,

Hermann Stamm-Wilbrandt
Developer, XML Compiler, L3
WebSphere DataPower SOA Appliances
----------------------------------------------------------------------
IBM Deutschland Research & Development GmbH
Vorsitzender des Aufsichtsrats: Martin Jetter
Geschaeftsfuehrung: Dirk Wittkopp
Sitz der Gesellschaft: Boeblingen
Registergericht: Amtsgericht Stuttgart, HRB 243294



From:       David Carlisle <davidc@xxxxxxxxx>
To:         xsl-list@xxxxxxxxxxxxxxxxxxxxxx
Cc:         Hermann Stamm-Wilbrandt/Germany/IBM@IBMDE
Date:       08/24/2010 09:23 PM
Subject:    Re: Fw:  Question on duplicate node elimination



On 24/08/2010 19:52, Hermann Stamm-Wilbrandt wrote:
> Is DOCTYPE supported by exslt:node-set()?

it can't be as node-set takes a rtf and converts it to a node, but
neither rtf nor node data types model DTD.  You may be better off with
xml:id attributes.

David