XPath to find text node that is a sibling of other nodes

Question

Given the following fragment of html:

<fieldset>
  <legend>My Legend</legend>
  <p>Some text</p>
  Text to capture
</fieldset>

Is there an xpath expression that will return only the 'Text to capture' text node?

Trying

/fieldset/text()

yields three nodes, not just the one I need.

Answer 1

Assuming what you want is the text node containing non whitespace text :

//fieldset/text()[normalize-space(.)]

If what you want is the last text node, then:

//fieldset/text()[last()]

Answer 2

I recommend you accept Steven D. Majewski's answer, but here is the explanation (text nodes highlighted with square brackets):

<fieldset>[
  ]<legend>My Legend</legend>[
  ]<p>Some text</p>[
  Text to capture
]</fieldset>

so /fieldset/text() returns

• "\n "
• "\n "
• "\n Text to capture\n"

And this is why you want /fieldset/text()[normalize-space()], and you want the result trimmed before use.

Also note that the above is short for /fieldset/text()[normalize-space(.) != '']. When normalize-space() returns a non-empty string, the predicate evaluates to true, while the empty string evaluates to false.