Question

Let’s say I have the following xml:

<one>
  <two> 
    <three>
      <three>
        foo
      </three>
      <four />
      <three>
        bar
        <four />
      </three>
    </three>
  </two>
</one>

What XPath query can I write that will return only these nodes:

<three>
  foo
</three>

and

<three>
  bar
  <four />
</three>

but not this one:

<three>
  <three>
    foo
  </three>
  <four />
  <three>
    bar
    <four />
  </three>
</three>

I’ve tried this:

//three/descendant-or-self::*/text()/ancestor::three

but it returns the one I don’t want. I also can’t assume that this is the exact structure of the xml, so I don’t think that the .. operator will work. Essentially, I want the youngest ancestor of the text nodes that’s of type three.

Any suggestions?

Answer 1

I think //three[text() and not (descendant::three)] will do what you want.