CODEWITHSUNDEEP
phpxmlsimplexml
reactor
reactor

Reputation: 1931

xpath with simplexml not able to access nested element

I have an xml like so:

  <?xml version="1.0" encoding="UTF-8"?>
  <foo SeqNum="1">
    <bar>1234</bar>
  </foo>
  <foo SeqNum="20">
    <bar>6789</bar>
  </foo>

and I'm trying to get the value 6789 with this query:

$xml = "<?xml version="1.0" encoding="UTF-8"?>
  <foo SeqNum="1">
    <bar>1234</bar>
  </foo>
  <foo SeqNum="20">
    <bar>6789</bar>
  </foo>";
$simple = new SimpleXMLElement($xml);
$result = $simple->xpath('//*[@SeqNum="20"]/bar/'); // result gives me nothing

So I tried to just get the parent like so

$result = $simple->xpath('//*[@SeqNum="20"]')[0]->asXML();

which gives me:

  <foo SeqNum="20">
    <bar>6789</bar>
  </foo>

So I'm almost there but am really stuck about what I'm not understanding. Thank you!

Upvotes: 1

Views: 214

Answers (1)

ThW
ThW

Reputation: 19502

Here are several mistakes in the question. The XML needs a root element and the trailing / breaks the expression. The literal quotes need to be changed to single quotes (or all the inner double quotes need to be escaped.)

Fixed example:

$xml = '<?xml version="1.0" encoding="UTF-8"?>
<foo>
  <foo SeqNum="1">
    <bar>1234</bar>
  </foo>
  <foo SeqNum="20">
    <bar>6789</bar>
  </foo>
</foo>';

$simple = new SimpleXMLElement($xml);
$result = $simple->xpath('//*[@SeqNum="20"]/bar');
var_dump((string)$result[0]);

Output:

string(4) "6789"

With Namespaces

If your XML is using namespaces you will have to define an alias/prefix for this namespace URI and use that in the Xpath expression.

$xml = <<<'XML'
<?xml version="1.0" encoding="UTF-8" ?> 
<p:foo xmlns:p="http://www.example.com">   
   <p:foo SeqNum="1">     
     <p:bar>1234</p:bar>   
   </p:foo>   
   <p:foo SeqNum="20">     
     <p:bar>6789</p:bar>   
   </p:foo> 
</p:foo> 
XML;
 
$simple = new SimpleXMLElement($xml);
$simple->registerXpathNamespace('e', 'http://www.example.com');
$result = $simple->xpath('//*[@SeqNum="20"]/e:bar');
var_dump((string)$result[0]);

The example uses a different alias for the expression to show that the document and the expression are separate - only the namespace URI has to match.

Namespaces have to be unique so they are defined with an URI (a superset of URL). Because that would get messy aliases are used in node names. The following 3 elements all can be read as {http://www.example.com}bar.

  • <p:bar xmlns:p="http://www.example.com"/>
  • <e:bar xmlns:e="http://www.example.com"/>
  • <bar xmlns="http://www.example.com"/>

Upvotes: 3

Related Questions