Question

I have html content stored in a database table. in that html content I want to replace “SOME WORDS” with a link tag. But if “SOME WORDS” is already inside a link tag i should omit them..

e.g.
The content

<p>Lorem ipsum dolor SOME WORDS, consectetur adipiscing elit. <a href="http://example.com">SOME WORDS</a> elementum pharetra velit at cursus. Quisque blandit, nibh at eleifend ullamcorper</p>

The output should be

<p>Lorem ipsum dolor <a href="http://someurl">SOME WORDS</a>, consectetur adipiscing elit. <a href="http://example.com">SOME WORDS</a> elementum pharetra velit at cursus. Quisque blandit, nibh at eleifend ullamcorper</p>

as you can see, it should exclude existing link texts when replacing.

Some guidance to get in to the right track is very much appreciated.

Answer 1

This is how you could solve it using DOMDocument instead of regular expressions:

$contents = <<<EOS
<p>Lorem ipsum dolor SOME WORDS, consectetur adipiscing elit. <a href="http://example.com">SOME WORDS</a> elementum pharetra velit at cursus. Quisque blandit, nibh at eleifend ullamcorper</p>
EOS;

$doc = new DOMDocument;
libxml_use_internal_errors(true);
$doc->loadHTML($contents);
libxml_clear_errors();

$xp = new DOMXPath($doc);

// find all text nodes
foreach ($xp->query('//text()') as $node) {
        // make sure it's not inside an anchor
        if ($node->parentNode->nodeName !== 'a') {
                $node->nodeValue = str_replace(
                    'SOME WORDS', 
                    'SOME OTHER WORDS', 
                    $node->nodeValue
                );
        }
}
// DOMDocument creates a full document and puts your fragment inside a body tag
// So we enumerate the children and save their HTML representation
$body = $doc->getElementsByTagName('body')->item(0);
foreach ($body->childNodes as $node) {
        echo $doc->saveHTML($node);
}