35

I'm looking for the right xpath syntax to get a specific parent of an element. Example:

root
   |- div
   |     |
   |     |----??? ---|
   |     |           |-a [class=1]
   |     |                      |- text[ A TEXT I DON'T WANT]
   |     |
   |     |
   |     |
   |     |-text[THE TEXT]
   |
   |-div 
   |    |-text[THE TEXT I DON'T WANT]
   |
   |-div 
   |    |-text[THE TEXT I DON'T WANT]

I want to get the text "THE TEXT" but the one that contains a [class=1] inside the same div. Something like this:

//div//a[@class=1]/text[contains(.,'A TEXT')]/parent::*/parent::*.... <till div element>  /text
lucid_dreamer
  • 362
  • 4
  • 9
Nir Bentzy
  • 351
  • 1
  • 3
  • 3

2 Answers2

69

Given the XML

<?xml version="1.0"?>
<root>
    <foo id="id1">
        <foo id="i2">
            <baz/>
        </foo>
    </foo>
</root>

You can find the nearest ancestor foo element from baz using the XPath expression:

//baz/ancestor::foo[1]

Which will select the foo element node of id "i2".

So in your example (if I understand right) once you have got the "a" element you want, you can get "back up" the tree to the nearest ancestor div by appending "/ancestor::div[1]" to your expression.

alexbrn
  • 2,050
  • 13
  • 11
10

Use:

/root/div[.//a[@class='1']]/text()

This selects any text node that is a child of any a element that has a class attribute with value '1' and that (the a element) is a descendent of any div element that is a child of the top element named root.

Dimitre Novatchev
  • 240,661
  • 26
  • 293
  • 431