DOMDocument remove script tags from HTML source

I used @Alex's approach here to remove script tags from a HTML document using the built in DOMDocument. The problem is if I have a script tag with Javascript content and then another script tag that links to an external Javascript source file, not all script tags are removed from the HTML.

$result = '
<!doctype html>
<html>
    <head>
        <meta charset="utf-8">
        <title>
            hey
        </title>
        <script type="text/javascript" src="http://ajax.googleapis.com/ajax/libs/jquery/1.9.1/jquery.min.js"></script>
        <script>
            alert("hello");
        </script>
    </head>
    <body>hey</body>
</html>
';

$dom = new DOMDocument();
if($dom->loadHTML($result))
{
    $script_tags = $dom->getElementsByTagName('script');

    $length = $script_tags->length;

    for ($i = 0; $i < $length; $i++) {
        if(is_object($script_tags->item($i)->parentNode)) {
            $script_tags->item($i)->parentNode->removeChild($script_tags->item($i));
        }
    }

    echo $dom->saveHTML();
}

The above code outputs:

<html>
    <head>
        <meta charset="utf-8">
        <title>hey</title>
        <script>
        alert("hello");
        </script>
    </head>
    <body>
        hey
    </body>
</html>

As you can see from the output, only the external script tag was removed. Is there anything I can do to ensure all script tags are removed?

Your error is actually trivial. A DOMNode object (and all its descendants - DOMElement, DOMNodeList and a few others!) is automatically updated when its parent element changes, most notably when its number of children change. This is written on a couple of lines in the PHP doc, but is mostly swept under the carpet.

If you loop using ($k instanceof DOMNode)->length, and subsequently remove elements from the nodes, you'll notice that the length property actually changes! I had to write my own library to counteract this and a few other quirks.

The solution:

if($dom->loadHTML($result))
{
    while (($r = $dom->getElementsByTagName("script")) && $r->length) {
            $r->item(0)->parentNode->removeChild($r->item(0));
    }
echo $dom->saveHTML();

I'm not actually looping - just popping the first element one at a time. The result: http://sebrenauld.co.uk/domremovescript.php

To avoid that you get the surprises of a live node list -- that gets shorter as you delete nodes -- you could work with a copy into an array using iterator_to_array:

foreach(iterator_to_array($dom->getElementsByTagName($tag)) as $node) {
    $node->parentNode->removeChild($node);
};

来源：https://stackoverflow.com/questions/15925961/domdocument-remove-script-tags-from-html-source

标签

php

html-parsing

xss

domdocument

script-tag