Regex match any url's in string with and without www and create a clickable URL

Question

There are many similar questions, but I still have not found a solution to what I try to achieve in php. I preg_match_all a string which can contain URLs written in various ways, but also contains text which should not match. What I need to match is:

www.something.com 
https://something.com
http://something.com
https://www.something.com
http://www.something.com

And any /..../.... after the URL, but not:

www.something.com</p> // this should match everything until the '</p>'
www.something.com. // this should match everything until the '.'

So far I got so far is

/((http|https)\:\/\/)?[a-zA-Z0-9\.\/\?\:@\-_=#]+\.([a-zA-Z0-9\&\.\/\?\:@\-_=#])*/

and the function

if(preg_match_all("/((http|https)\:\/\/)?[a-zA-Z0-9\.\/\?\:@\-_=#]+\.([a-zA-Z0-9\&\.\/\?\:@\-_=#])*/",$text,$urls)){
    foreach($urls[0]as $url ){
        $text = str_replace($url,'<a href="'.$url.'">'.$url.'</a>',$text);
    }
}

but this gives a problem with http://www.... (the http:// won't be inlcuded in the displayed text), and with a URL without http or https the created link is relative to the domain I show the page on. Suggestions?

Here's a live Demo

Edit: my best regex so for any URL with http or https is /(http|https)\:\/\/[a-zA-Z0-9\-\.]+(\.[a-zA-Z]{2,3})?(\/[A-Za-z0-9-._~!$&()*+,;=:]*)*/. Now I just need a way to regex the URLs with only www.something... and transform that into http://www.something... in the href.

Here's another live demo with different examples.

Edit 2: the answer from this question is quite good. The only problem with this that I still encounter is with </p> after the URL and if there are words before and after a dot (this.for example).

$url = '@(http)?(s)?(://)?(([a-zA-Z])([-\w]+\.)+([^\s\.]+[^\s]*)+[^,.\s])@';
$string = preg_replace($url, '<a href="http$2://$4" target="_blank" title="$0">$0</a>', $string);
echo $string;

Please create a live demo to demonstrate the issue. – revo Dec 06 '18 at 21:31 — revo, Dec 06 '18 at 21:31
@revo, did it. See https://3v4l.org/VRXuf – Dirk J. Faber Dec 06 '18 at 21:35 — Dirk J. Faber, Dec 06 '18 at 21:35

Victor · Accepted Answer · 2018-12-06T23:03:23.383

2

Maybe this one fits your needs:

$text = preg_replace_callback('~(https?://|www)[a-z\d.-]+[\w/.?=&%:#]*\w~i', function($m) {
    $prefix = stripos($m[0], 'www') === 0 ? 'http://' : '';
    return "<a href='{$prefix}{$m[0]}'>{$m[0]}</a>";
}, $text);

edited Dec 06 '18 at 23:03

answered Dec 06 '18 at 22:26

Victor

5,493
1
27
28

Are you sure? If so, can you publish full text used to test it? – Victor Dec 06 '18 at 22:46
I also wonder why, I am trying to find out what's wrong, but this is the result: – Dirk J. Faber Dec 06 '18 at 22:47
http://www.something.com'>www.something.com http://http://www.something.com'>www.something.com https://http://www.something.com'>www.something.com – Dirk J. Faber Dec 06 '18 at 22:48
I understand what is wrong now, the `$matches` are both the URLs and the prefixes. – Dirk J. Faber Dec 06 '18 at 22:54
@DirkJ.Faber This is because it replaces same URLs multiple times. – Victor Dec 06 '18 at 22:55

score 1 · Answer 2 · answered Dec 06 '18 at 23:03

$text =  "<p>Some string www.test.com with urls http://test.com in it http://www.test.com. </p>";
$text = preg_replace_callback("@(http)?(s)?(://)?(([a-zA-Z])([-\w]+\.)+([^\s\.]+[^\s]*)+[^,.\s])@", 'replace_callback', $text);

function replace_callback($matches){
    return '<a href="' . $matches[0] . '" target="_blank">' . $matches[0] . '</a>';
}

score 0 · Answer 3 · answered Dec 06 '18 at 21:21

0

You regex was almost correct!

You we're matching a literal dot \. followed by 0 or more group of characters including the dot.

So i changed it to matching a literal dot followed by 1 or more characters excluding the dot which seems to be what you want, here is the final regex:

((http|https)\:\/\/)?[a-zA-Z0-9\.\/\?\:@\-_=#]+\.([a-zA-Z0-9\&\/\?\:@\-_=#])+

See it in action: https://regex101.com/r/h5pUvC/3/

answered Dec 06 '18 at 21:21

Dmitri Chebotarev

2,429
1
11
7

I used your regex, but I still have some problems. I updated the question to make things more clear. – Dirk J. Faber Dec 06 '18 at 21:29
See also https://regex101.com/r/h5pUvC/4 for an example of another problem I encounter. – Dirk J. Faber Dec 06 '18 at 22:05

Regex match any url's in string with and without www and create a clickable URL

3 Answers3