Get a specific element with xmllint

Consider the following HTML code as in file example.html

<!DOCTYPE html>
<head>
 <meta charset="utf-8">

<body>
 <div class="container">
  <h1>
   <a href="/">
    <img src="/images/image.png"/>
   </a>
  </h1>
   <ul>
    <li><a href="/" >link1<br /><span>content</span></a></li>
    <li><a href="/folder" >link2<br /><span>more contend</span></a></li>
   </ul>
 </div><!-- .container -->
 <div class="container">
  <ul class="disc">
   <li><strong>no interest 1</strong></li>
   <li><strong>no interest 2</strong></li>
   <li><strong>no interest 3</strong></li>
   <li>keyword1: <strong>unkown_content</strong></li>
   <li>keyword2: <strong>unkown_content</strong></li>
  </ul>
 </div><!-- .container -->
 <div class="container">
  <ul class="disc">
   <li><a href="/folder1">not interested</a></li>
   <li><a href="/folder2">not interested</a></li>
   <li><a href="/folder3">not interested</a></li>
   <li>keyword1: <strong>unkown_content</strong></li>
   <li>keyword2: <strong>unkown_content</strong></li>
  </ul>
 </div><!-- .container -->
</body>
</html>

I would like to get the unkown_content after the second keyword1. I have tried

xmllint --html --xpath "//li[text()='keyword1: ']" example.html 2>/dev/null
<li>keyword1: <strong>unkown_content</strong></li><li>keyword1: <strong>unkown_content</strong></li>

As you can see this returns both nodes with keyword1. How do I get the content of the last occurence?

Solution

Try xmllint --html --xpath "(//li[text()='keyword1: '])[last()]" example.html 2>/dev/null.

Note the addition of [last()] - it points to the last occurrence of //li[text()='keyword1: ']. The braces around //li[text()='keyword1: '] are necessary, too, as [last()] is applied to a nodeset.