I have the following XML (which is actually HTML):
<html>
<h4>something</h4>
<p>a</p>
<p>b</p>
<h4>otherthing</h4>
<p>c</p>
</html>
Can a XPath selects the "p" nodes that are following siblings of the first "h4" node but not following siblings of second "h4" node (selecting "p" node a & b only)?
Assuming no <p> preceding the first <h4>,
//h4[2]/preceding-sibling::p
My take
//p[preceding-sibling::h4[1] and not(preceding-sibling::h4[position() > 1])]
finds all p elements which are siblings of the first h4 but not siblings of any other h4 on the same axis
Alternative
//h4[1]/following-sibling::p[count(preceding-sibling::h4) = 1]
finds all following p element of the first h4 element that do have exactly one preceding h4 element
Use:
/*/h4[1]/following-sibling::p
[not(count(preceding-sibling::* | /*/h4[2])
=
count(preceding-sibling::*)
)
]
More generally, the intersection of two nodesets $ns1 and $ns2 is selected by:
$ns1[count(.|$ns2) = count($ns2)]
The fact that a node $n is not in a node-set $ns1 is expressed by:
not(count($n | $ns1) = count($ns1))
This is fundamental set theory and usage of the standard XPath | (union operator and not() function.