[RESOLVED] If in REGEX

Shawazi

Part 3 of the script I'm slowly writing up 🙂

This converts all http://www.whatever.com/ to a link. It cuts off things at 60 characters.

$pattern = '/([^"\'])(http[s]?:\/\/[[a-zA-Z0-9_-]+\.]?[a-zA-Z0-9_\/?&%-]+\.[a-zA-Z0-9_-]+\/?[[a-zA-Z0-9+_:;=\/?&%\.-]+]?)/e';
$text = preg_replace($pattern,  "'\\1<a href=\"\\2\">'.substr('\\2', 0, 60).'</a>'" , $text);

I tried doing some type of if statement like

$replacement =  "'\\1<a href=\"\\2\">'.(($blah) ? substr(\"\\2\", 0, 5) : \"No blah!\").'</a>'"

But it just blanks out $text. I need to make it so things over 60 characters get a ... on the end, otherwise don't truncate. I know to use strlen instead of $blah after but I can't get this simple test to work.

Kudose

Shouldn't you be using double quotes to start concatenating again?

$replacement =  "'\\1<a href=\"\\2\">".(($blah) ? substr("\\2", 0, 5) : "No blah!")."</a>'";

I don't even know if the above will run right, but I do know your regex isn't taking multiple sub-domains into account.

I've seen web sites with URL's like http://tvad64.promos.website.com

Weedpacket

Have you considered using [man]preg_replace_callback[/man]? Then you can replace the matched string with any transformation of it that you can describe in PHP.

Shawazi

Kudose:
I had http://manage.hosted.filefront.com/live.php?page=fpad&redirect=http://www.filefront.com/
as one of the links I tested with and it changed it fine.

Shawazi

PHPBuilder just did to that link what I am trying to do now ha

Shawazi

I tried:

function urlConcat($matches) {
	return $matches[2];
}
$pattern = '/([^"\'])(http[s]?:\/\/[[a-zA-Z0-9_-]+\.]?[a-zA-Z0-9_\/?&%-]+\.[a-zA-Z0-9_-]+\/?[[a-zA-Z0-9+_:;=\/?&%\.-]+]?)/e';
$text = preg_replace_callback($pattern, 'urlConcat', $text);
echo $text;

But now it just gets rid of my links all together.

Weedpacket

Shawazi wrote:
But now it just gets rid of my links all together.

Well yeah; you're still trying to do everything in the regular expression. What you wrote is no different to

$pattern = '/([^"\'])(http[s]?:\/\/[[a-zA-Z0-9_-]+\.]?[a-zA-Z0-9_\/?&%-]+\.[a-zA-Z0-9_-]+\/?[[a-zA-Z0-9+_:;=\/?&%\.-]+]?)/e';
$text = preg_replace($pattern, '\\2', $text);
echo $text;

And because you've still got that /e modifier, you're trying to execute that second part of that match.

Why not just look for stuff that looks something like a URL with a simpler expression), pass that to the callback function, which can then take a closer look. If it turns out not to be a URL after all you can always just return $matches[0].

Shawazi

Agh forgot the e modifier. it's good now thanks