Modify HTML Using PHP

readme.md

Modify HTML Using PHP

Instead of relying on unsafe regular expressions and string manipulation, we can utilize PHP's built-in DOM extension for modifying HTML.

The manipulate_html() function from this snippet allows you to pass some HTML code, traverse & modify each of its DOM nodes in a callback and will return to you the modified HTML code.

The following example modifies all images in an HTML snippet to use lazy loading:

manipulate_html('<img src="foo.jpg">', function (DOMNode $node) {
    if ($node->nodeName === 'img') {
        $node->setAttribute('loading', 'lazy');
    }
});

// Returns '<img src="foo.jpg" loading="lazy">'

This is just a single element, but you can test this snippet with any website:

manipulate_html(
    file_get_contents('https://www.php.net/'),
    function (DOMNode $node) {
        // ...
    }
);

manipulate_html.php

<?php

function walk_dom(DOMNode $domNode, callable $callback): void
{
    foreach ($domNode->childNodes as $node) {
        $callback($node);

        if ($node->hasChildNodes()) {
            walk_dom($node, $callback);
        }
    }
}

function manipulate_html(string $html, callable $callback): string
{
    $dom = new DOMDocument();

    // Don't spread warnings when encountering malformed HTML
    $previousXmlErrorBehavior = libxml_use_internal_errors(true);

    // Use XML processing instruction to properly interpret document as UTF-8
    @$dom->loadHTML(
        '<?xml encoding="utf-8" ?>' . $html,
        LIBXML_HTML_NOIMPLIED | LIBXML_HTML_NODEFDTD
    );
    foreach ($dom->childNodes as $item) {
        if ($item->nodeType === XML_PI_NODE) {
            $dom->removeChild($item);
        }
    }
    $dom->encoding = 'UTF-8';

    walk_dom($dom, $callback);

    // Turn DOM back into HTML and remove leading/trailing whitespace
    $result = trim($dom->saveHTML());

    // Restore previous XML error behavior
    libxml_use_internal_errors($previousXmlErrorBehavior);

    return $result;
}