如何使用xpath [PHP]批量解析HTML？

Question

I tried all sorts of things but couldn't find a solution. I want to retrieve elements from html code using xpath in php.

Ex:

<div class='student'>
 <div class='name'>Michael</div>
 <div class='age'>26</div>
</div>
<div class='student'>
 <div class='name'>Joseph</div>
 <div class='age'>27</div>
</div>

I want to retrieve the information and put them in an array as follows:

$student[0][name] = Michael;
$student[0][age] = 26;
$student[1][name] = Joseph;
$student[1][age] = 27;`

In other words i want the matching ages to stay with the names.

I tried the following:

$dom = new DOMDocument();
$dom->loadHTML($html);
$xpathDom = new DomXPath($dom);
$homepostcontentNodes = $xpathDom->query("//*[contains(@class, 'student')]//*[contains(@class, 'name')]");`

However, this is only grabbing me the nodes 'names' How can i get the matching age nodes?

Answer 1

Of course it is only grabbing the nodes name - you are telling it to!

What you will need to do is in two steps:

1. Pick out all the student nodes
2. For each student node, pick out the columns

This is a pretty standard step in linearization of data, and the XPath queries are simple:

Step 1

You pretty much have it:

 $studentNodes = $xpathDom->query("//div[contains(@class, 'student')]");

This will return all your student nodes.

Step 2

This is where the magic happens. We have our nodes, we can loop through them (DOMNodeList implements Iterator, so we can foreach-loop through them). What we need to figure out is how to find its children...

...Oh wait. DOMNode implements a method called getNodePath which returns the full, direct XPath path to the node. This allows us to then simply append /div to get all the div direct descendents to the node!

Another quick foreach, and we get this code:

$studentNodes = $xpathDom->query("//div[contains(@class, 'student')]");
$result = array();
foreach ($studentNodes as $v) {
// Child nodes: student
$r = array();
$columns = $xpathDom->query($v->getNodePath()."/div");
foreach ($columns as $v2) {
           // Attributes allows me to get the 'class' property of the node. Bit clunky, but there's no alternative
    $r[$v2->attributes->getNamedItem("class")->textContent] = $v2->textContent;
}
$result[] = $r;
}
var_dump($result);

Full fiddle: http://codepad.viper-7.com/t868Wh