great Fez wrote:


> I'm an xpath newbie and I'm trying to figure out how to return all
> sibling nodes or child nodes that do not have children (so only go to
> depth = 0 or 1).
> 
> I have the following:
> 
> <top>
>   <one></one>
>   <two></two>
>   <three></three>
>   <alpha1>
>      <letter1></letter1>
>      <letter2></letter2>
>      <beta>
>         <lettera></lettera>
>         <letterb></letterb>
>      </beta>
>   </alpha1>
>   <alpha2>
>      <letter1></letter1>
>      <letter2></letter2>
>      <beta>
>         <lettera></lettera>
>         <letterb></letterb>
>      </beta>
>   </alpha2>
> <top>
> 
> And I'm trying to retrieve only nodes that have no children:
> <one></one>
> <two></two>
> <three></three>

   /top/*[not(node())]
selects all child elements of the top root element that do not have any 
child nodes/contents. If the top name is just an example you can of 
course also use the more general
   /*/*[not(node())]

> and for a later call node that only have children:
>   <alpha1>
>   </alpha1>
>   <alpha2>
>   <alpha2>

Then
   /top/*[node()]
or more general
   /*/*[node()]

Note that node() selects all kind of nodes (element nodes, text nodes, 
comment nodes, processing instruction nodes), if you are only interested 
if element nodes have element node children then e.g.
   /*/*[*]
is more appropriate.
-- 

	Martin Honnen --- MVP XML
	http://JavaScript.FAQTs.com/