Pseudo tag Autolink Regex Tweak Needed

[BizLab]

Hey there,

 

I have a pseudo link tag [link][/link] and an auto hyperlink preg_replace. The auto hyperlink works just fine, but i want it to ignore any links that are inside my pseudo link element.

 

here is the hyperlink regex (which works great for any of you guys that need one - copy & paste BUT change the [link] to your standard <a> element)

 

$str = preg_replace('|(?<!href=")(https?://[A-Za-z0-9+\-=._/*(),@\'$:;&!?%]+)|i','[link href="$1" target="_blank"]$1[/link]', $str);
// Takes : http://www.mycoolsite.com/
// Outputs: [link href="http://www.mycoolsite.com/" target="_blank"]http://www.mycoolsite.com/[/link]  - WHICH IS GOOD, BUT read further

 

The problem comes when editing content that has already been hyperlinked. The output from that then becomes

 

// [link href="[link href="http://www.mycoolsite.com/" target="_blank"]http://www.mycoolsite.com/[/link]" target="_blank"]http://www.mycoolsite.com/[/link]

 

And so on. The only mod i need to make is to check for the [link] tag, and if it exists, do not re-hyperlink it.

 

Thanks to anyone with suggestions!

 

[BizLab]

Anyone have ideas?? I'm very confident that the change is isolated to the first section, and will only need to identify the [link][/link] pseudo element instead of the standard <a></a> element. Everything i have tried has filed though..

 

$str = preg_replace('|(?<!href=") // changes here??

[.josh]

Your posted regex can't make what you are saying it's making. 

 

You have: (?<!href=")(https?://[A-Za-z0-9+\-=._/*(),@\'$:;&!?%]+)

 

This pattern says to match (https?://[A-Za-z0-9+\-=._/*(),@\'$:;&!?%]+) if it is not preceded by href=". 

 

So if you have...

 

[link href="http://www.mycoolsite.com/" target="_blank"]http://www.mycoolsite.com/[/link]

 

...the http://www.mycoolsite.com/ in your [link] tag won't match, because it is preceded by href="

 

The only way you could be getting the output you say you are getting is if you are doing something more than this preg_replace(), and have not posted everything you are actually doing here. 

 

However, I do see a flaw in what you have posted:

 

test:

$str = "http://www.mycoolsite.com/";

$str = preg_replace('|(?<!href=")(https?://[A-Za-z0-9+\-=._/*(),@\'$:;&!?%]+)|i','[link href="$1" target="_blank"]$1[/link]', $str);
echo "round 1 : " . $str . "<br/>";

$str = preg_replace('|(?<!href=")(https?://[A-Za-z0-9+\-=._/*(),@\'$:;&!?%]+)|i','[link href="$1" target="_blank"]$1[/link]', $str);
echo "round 2 : " . $str . "<br/>";

$str = preg_replace('|(?<!href=")(https?://[A-Za-z0-9+\-=._/*(),@\'$:;&!?%]+)|i','[link href="$1" target="_blank"]$1[/link]', $str);
echo "round 3 : " . $str . "<br/>";

 

output:

round 1 : [link href="http://www.mycoolsite.com/" target="_blank"]http://www.mycoolsite.com/[/link]
round 2 : [link href="http://www.mycoolsite.com/" target="_blank"][link href="http://www.mycoolsite.com/" target="_blank"]http://www.mycoolsite.com/[/link][/link]
round 3 : [link href="http://www.mycoolsite.com/" target="_blank"][link href="http://www.mycoolsite.com/" target="_blank"][link href="http://www.mycoolsite.com/" target="_blank"]http://www.mycoolsite.com/[/link][/link][/link]

 

So I do see a flaw..but not what you are saying.  I see matching and replacing the original link inside the tag, not inside your [link..] tag. 

 

This should solve that...

$str = preg_replace('~(?<!link href="|\])(https?://[A-Za-z0-9+\-=._/*(),@\'$:;&!?%]+)~i','[link href="$1" target="_blank"]$1[/link]', $str);

 

p.s. - I changed your pattern delimiter from | to ~  you should not use | as the pattern delimiter because it has special meaning in regex (it is the "or" operator).  While it is possible to use | as the delimiter, it makes it very difficult to then use it as the "or" operator in the pattern.

 

 

[BizLab]

Dude. Awesomeness is the KEY to REGEX  i didn't realize the exclusion on the front end of this expression - simply perfect!

 

Thanks a million once again CV!

 

SOLVED!

[BizLab]

One more thing just to clarify the question of how my regex was behaving initially - it turns out that i was escaping the string before running this regex, so that once the string hit the regex for evaluation, the string didn't match, and was RE-converted

 

Thanks again CV