RE: Selecting a node set

I was hoping to avoid any variable reference, namely because I cannot
imagine what variable I would pass into it.  The resultant node set is
entirely conditional on the contents of the node immediatly following it (or
I can get essentially the same results using the preceding node)

You didn't say what additional information was needed, but I can tell you
the following:

The structure is as follows:

<stream>
	<endTime>
		<interval value="123" channelId="ch1"
channelDetailId="chd2">
		...
		<interval>
	</endTime>
	...
	<endTime>
	...
	</endTime>
</stream>

There are N endTime elements per stream
There are N intervals per endTime
Within an endTime, channelDetailId should uniquely identify an interval

Essentially I want the set of all endTimes whose following sibling endTime
does not have the same number of intervals with channelDetailIds matching
the current endTime's set of intervals.

The xpath: "endTime[interval/@channelDetailId !=
following-sibling::endTime[1]/interval/@channelDetailId ]" does not give the
proper results - it yields all but the very last endTime



-----Original Message----- 

The information you're giving us is insufficient. In general I think the
answer
should be that you cannot have a single XPath expression for the above
without also
including in it some variable references.

In any specific case the answer may be different.

For example, if it is the case that a "channelDetailId" attribute value
uniquely
identifies an "interval" element, then the following XPath expression will
yield the
desired result:

endTime[interval/@channelDetailId 
     !=  
       following-sibling::endTime[1]/interval/@channelDetailId
       ]


This will be even simpler if we knew that an "endTime" has only one
"interval"

Cheers,
Dimitre Novatchev.


__________________________________________________
Do You Yahoo!?
Get personalized email addresses from Yahoo! Mail
http://personal.mail.yahoo.com/

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list