RE: [xsl] Creating XPath2 functions:passing a set of nodes, what is the signature construction?

["Michael Kay" <mike@xxxxxxxxxxxx>]

I would recommend that you define the argument as a sequence of doubles

as="xs:double*"

rather than as a sequence of nodes. If you call it supplying a sequence of
nodes, they will then automatically be atomized to yield the contained
doubles. If you declared it to require a sequence of nodes

as="node()*"

then you would be unnecessarily constrained in how you call the function.

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

> -----Original Message-----
> From: cknell@xxxxxxxxxx [mailto:cknell@xxxxxxxxxx] 
> Sent: 25 January 2006 21:26
> To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx
> Subject: [xsl] Creating XPath2 functions:passing a set of 
> nodes, what is the signature construction?
> 
> I am learning XPath2 using Saxon 8.6. I'd like to create a 
> funtion to compute the standard deviation of the value of a 
> set of nodes. Is there some special way to construct the 
> signature of the function when the argument is a set of nodes 
> as opposed to a single node?
> 
> I have written a funtion that takes a single node as an 
> argument of type xs:double. Is there some special way to 
> state that the function will take a set of nodes of type 
> xs:double, or do I simply state that the argument will be of 
> type xs:double?
> 
> -- 
> Charles Knell
> cknell@xxxxxxxxxx - email