HELP: Adjust my link-finder function?

[MungoBBQ]

Hi,

I've got the following function for finding URL:s and e-mail addresses in a string:

function strLinks($text) {
   $text = preg_replace("/((http(s?):\/\/)|(www\.))([\w\.:!?~=&\-_\/]+)/i","<a href=\"http$3://$4$5\" target=\"_blank\"><u>$2$4$5</u></a>",$text); 
   $text = preg_replace("/([w.-_]+)(@)([w.-_]+)/i", "<a href=\"mailto:$0\"><u>$0</u></a>",$text);
   return $text;
    }

It works great, except it has two annoying "bugs":

1. If the string begins with an URL or e-mail address, it won't find it for some reason.

2. If the string contains HTML hyperlinks, it screws up on these.

So, I would like to have this function adjusted to find hyperlink items even if they are in the beginning, and to:

1. Ignore URLs and e-mail addresses inside HTML tags
   and
2. Ignore URLs and e-mail addresses nested in <a> and </a> tags.

Is there anybody out there who can help me with this? I'd be eternally grateful!

//Mans

[xblue]

Bug 1 I couldn't reproduce, could you post again if it still occurs using my example?

Bug 2: It can be done, though it's fairly complicated. I use urls beginning with http:// as example.

1)
Put together three subpatterns that match the unwanted cases as well as the wanted one.

Some simple examples:
URL that is already enclosed in <a ...> </a> tags:
<a\s[^{>]+>http://[\s]+</a>}

URL within a tag:
<[^{>]+http://[>]+>}

URL:
http://[^\s]+
(modify them according to your needs)

EDITED: It may be useful to test each of them separately to make sure they match what you expect them to match.

2)
Combine them in an alternation. Make sure the unwanted cases go first, and enclose them in parenthesis:
pattern:
%(<a\s[^{>]+>http://[\s]+</a>)|(<[>]+http://[>]+>)|http://[\s]+%i}

3)
Now you can make use of the e-modifier that allows to evaluate the replacement argument: if the complete match equals the first or second subpattern, replace it with itself, else add the link tag.

$text = "http://www.domain.com this was the first url, let's start a new line to celebrate this event.\n
However there is more to come <img scr=\"http://www.domain.net/pic.gif\"> 
Image path: http://www.domain.net/pic.gif 
And here comes one enclosed in tags: <a href=\"http://www.domain.org\">http://www.domain.org</a>";

echo preg_replace('%(<a\\s[^>]+>http://[^\\s]+</a>)|(<[^>]+http://[^>]+>)|http://[^\\s]+%ie',
                  '"$0"=="$1" || "$0"=="$2" ? "$0" : "<a href=\"$0\">$0</a>"', $text);

[MungoBBQ]

Thanks xblue,

your code works really good in respect to ignoring URLs inside <a>-</a> tag blocks!

However, now URLs which do not begin with "http://" are no longer hyperlinked, neither are e-mail addresses.

Is there a simple way for me to add in those functions again?

//Mans

[xblue]

Hi, that's why I tried to explain it in detail: for you to be able to do it yourself 😉

For email address, follow the 3 steps I outlined, and if you encounter problems with any of them, people in this forum will most probably be willing to help. To catch urls that don't begin with [url]http://,[/url] it should be enough to modify the subpatterns accordingly.

[MungoBBQ]

Thanks,

Yes, I managed to get the e-mail addresses linked OK, using the pattern I had before. I also managed to get the "www."-addresses linked, but I can not figure out how to add the "http://" to the beginning of the hyperlink only if it is not already present...

Here is the function as it currently stands:

function strLinks($text) {
	$text = preg_replace('%(<a\s[^>]+>[url]http://[/url][^\s]+</a>)|(<[^>]+[url]http://[/url][^>]+>)|([url]www.[/url]|[url]http://[/url])[^\s]+%ie',
                  '"$0"=="$1" || "$0"=="$2" ? "$0" : "<a target=\"_blank\" href=\"$0\">$0</a><b>$1</b>"', $text);
	$text = preg_replace("/([w.-_]+)(@)([w.-_]+)(\.)([w.-_]+)/i", "<a href=\"mailto:$0\">$0</a>",$text);
	return $text;
	}

Now, the problem is that the only data I get out is $0, which always contains the complete matched URL - so how do I insert "http://" there if the URL doesn't already have it? Thanks again for your swift and exceptional help!

[xblue]

Hi,

I didn't look at your code closely since it's messed up but the
forum - please remember to disable "Automatically parse URLs"
next time 😉

This probably would be possible somehow, however I think it
would be quite an effort to do this properly.

So let's look at a simpler way: assume that www.something
occurs - you'd probably only want to change it if it comes after a
space (or new line) or is at the beginning of the string. Thus, you
could set up an additional preg_replace statement like:

$text = preg_replace("%(^|\\s)(www.[^\\s]+)%i", "$1<a href=\\"http://$2\\">$2</a>", $text);

Actually, you could use this simpler approach for urls beginning
with http:// as well and should get rid of most unwanted cases
(is within tags or enclosed in tags already)
😃 😃 😃

[MungoBBQ]

Thanks, you are so smart... 😃

I adjusted things a little more, and I think I finally got it working just the way I wanted!

The following example didn't work at first:

<a href="http://www.domain.com">www.domain.com</a>

That piece of any string was detected and the result was:

<a href="http://www.domain.com"><a href="http://www.domain.com">www.domain.com</a></a>

I.E. double hyperlinks. So I copied and adjusted your preg:

$text = preg_replace('%(<a\s[^>]+>http://[^\s]+</a>)|(<[^>]+http://[^>]+>)|(http://)[^\s]+%ie',
                  '"$0"=="$1" || "$0"=="$2" ? "$0" : "<a target=\"_blank\" href=\"$0\">$0</a>"', $text);

...to:

$text = preg_replace('%(<a\s[^>]+>http://|www.[^\s]+</a>)|(<[^>]+www.[^>]+>)|(www.)[^\s]+%ie',
                  '"$0"=="$1" || "$0"=="$2" ? "$0" : "<a target=\"_blank\" href=\"http://$0\">$0</a>"', $text);	

Running both of these together now seems to work. I'm not sure why though. 🙂 I added the "http://|www." to that second regexp, as I guessed that the "|" would make it work for both http:// AND www. ... and I guess it works.

Thanks a million times for your help!