Question

I wrote a simple function which takes a block of text, extracts urls from it, and replaces all urls with an <a href> tag around them.

E.g http://site.com should become <a href="http://site.com">http://site.com</a>

Code:

function parseUrls( $string )
{
    $string = trim($string);
    $pattern = '%\bhttp[s]?://[A-z0-9/\.\-_]+%i';
    $replacement = '<a href="$1">$1</a>';

    $string = preg_replace($pattern, $replacement, $string);

    return $string;
}

However if I pass the following string as input:

hello https://google.com test http://test.com/something.html abc http://site.com

The output I get is:

hello <a href=""></a> test <a href=""></a> abc <a href=""></a> 

I.e so the urls are being matched, but $replacement isn’t being applied correctly. May be my usage of $1 is wrong somehow?

What am I doing wrong?

Answer 1

You don’t have a capturing group that $1 would refer to.

Use $replacement = '<a href="$0" target="_BLANK">$0</a>'; instead.

Also, don’t use A-z in your character class (it matches more than you think: there are some non-letter characters between ASCII Z and a). A-Z is enough since you’ve made it case-insensitive anyway.