RE: [xsl] Sorting RTF

Subject: RE: [xsl] Sorting RTF
From: "Michael Kay" <mike@xxxxxxxxxxxx>
Date: Fri, 19 Nov 2004 15:11:40 -0000
If you're using Saxon 7.8 then

(a) you should be using Saxon 8.1.1

(b) you are using XSLT 2.0 and therefore don't need the xx:node-set()
function

(c) the Saxon namespace should be http://saxon.sf.net/

The most likely explanation for the "casting empty sequence" error is that
there is no @dc:date attribute - certainly, you haven't given us any
information about it, nor did you mention the dc namespace.

Michael Kay
http://www.saxonica.com/ 

> -----Original Message-----
> From: Paul Hebble [mailto:maceo@xxxxxxxx] 
> Sent: 18 November 2004 14:29
> To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx
> Subject: [xsl] Sorting RTF
> 
> Hi all,
> 
> Apologies in advance if this is an FAQ (and I don't see how 
> it couldn't
> be), but several days of searching the web, usenet, and mailing lists
> have not turned up a solution.
> 
> I'm using Saxon 7.8 with these namespaces:
> 	xmlns:exsl="http://exslt.org/common";
> 	xmlns:saxon="http://icl.com/saxon";
> 	extension-element-prefixes="saxon exsl"
> 
> Here's what I want to do:
> 
> <xsl:variable name="items">
> 	<xsl:apply-templates select="*"/>
> </xsl:variable>
> <xsl:for-each select="exsl:node-set($items)/*">
> 	<xsl:sort select="xs:dateTime(@dc:date)" order="descending"/>
> 	<xsl:copy-of select="."/>
> </xsl:for-each>
> 
> Saxon responds to this with this error on the for-each line:
> "An empty sequence is not allowed as the value in 'cast as' 
> expression".
> 
> I know $items isn't empty because if I replace the for-each with a
> copy-of, I get the expected nodes.  I have tried setting my 
> stylesheet's
> version to 1.0, 1.1, and 2.0, and I have also tried these:
> 	select="$items"
> 	select="$items/*"
> 	select="exsl:node-set($items)"
> 	select="saxon:node-set($items)"
> 	select="saxon:node-set($items)/*"
> 
> The last two give me this error:
> "No function found matching saxon:node-set with one argument"
> Which makes me think that maybe xmlns:saxon is wrong, but I 
> haven't seen
> any others in my searching.
> 
> Any help would be appreciated.
> 
> -- 
> 	Paul

Current Thread