Ok, the URL matching and replacing regexps here don't work as well as they should.  I tried fixing them last night but nothing I did seems to work right.  The raw code is:

$value = preg_replace('/(http:\/\/[A-Za-z0-9_-]+(?:[.][A-Za-z0-9_-]+)+\S*)/xims', '<a href="$1">$1</a>', $value);
$value = preg_replace('/(?<!http:\/\/)(www[.][A-Za-z0-9_-]+(?:[.][A-Za-z0-9_-]+)+\S*)/xims', '<a href="http://$1">$1</a>', $value);
$value = preg_replace('/([A-Za-z0-9_.-]+@[A-Za-z0-9_-]+(?:[.][A-Za-z0-9_-]+)+)/xims', '<a href="mailto:$1">$1</a>', $value);

However, we still have the problem of:

aaahttp://www.example.com

Anytime I try and fix that, the regexp stops working entirely on what I want it to do.

The raw PERL-compatible regexps:

/(http:\/\/[A-Za-z0-9_-]+(?:[.][A-Za-z0-9_-]+)+\S*)/
/(?<!http:\/\/)(www[.][A-Za-z0-9_-]+(?:[.][A-Za-z0-9_-]+)+\S*)/
/([A-Za-z0-9_.-]+@[A-Za-z0-9_-]+(?:[.][A-Za-z0-9_-]+)+)/

In addition to just working like JOS, I'd like the URL matching to work a little better: I don't want them to match to a trailing period, comma, or trailing brackets or braces -- I didn't have much luck with that, either.

+100 bonus points for the best link-matching regexp.

The above example you gave seems fine.

It's not supposed to match.

why not?

Anyway that shit's too convoluted for me to play with today.  I'm burned out on regexps  :)

No space before the http.

None of your regexps is LOOKING for a space before the http.

You need some \b's in there.

Ummm.. tried that, from the OP:

"Anytime I try and fix that, the regexp stops working entirely on what I want it to do."

That makes no sense.

No kidding!!  Very frustrating.

...or I simply was too tired last night -- I think I was using the wrong backlashed character (maybe \w).  Ok, one problem solved.  Now how would I stop the last period, comma, bracket, or brace from matching?

Why not let RegEx match URL with Comma, bracket, brace and with the current match, check if the last character(s) are in your black list and remove them?

I know jack about regular expressions, but some searching around turned up this:

http://gnosis.cx/publish/programming/regular_expressions.html

The very end of the page may be useful to those that understand what he's explaining.

AHA,

Could you please respond to my emial with request for SVN access?

Thanks,

Have you compared your one with those on offer at www.regexlib.com to see what is different?

I dont know whether I'm mad or not, but I think I have a wuunderful idea - check for a 2xx return status for the URL. That will avoid all that reg what.

In order to test the return from the URL you have to FIND the URL.

Moron.

And why do you need to find the URL ?

To test it.
Moron

How are you going to programmatically test the URL if you don't know what the URL is?  Explain it to us, magic man.

Testing a sliding window of text to see if its maybe a URL by checking the error return from HTTP is akin to expecting Hamlet from a barrel load of monkeys without giving them a typewriter ribbon.

Psshhh... if it's reasonable to expect the literature it's reasonable to expect the engineering of an ink ribbon.

Silly man.

:)