Parse a url form a text string

fizzwizz

I would like to parse the url out of a random text string

$string="See http://www.somedomain.com/randompage.php for more details";

I would like
$link="http://www.somedomain.com/randompage.php";

Many thanks .. Fizz

pyro

Here's a regex that will do it for you:

<?PHP

$string="See http://www.somedomain.com/randompage.php for more details";

preg_match_all("/((?:http|ftp)s?:\\/\\/[-a-z0-9]+(?:\\.[-a-z0-9]+)+(?:[-\\w:@&?=+,.!\\/~*'%#$]*(?<![.,?!]))?)/i", $string, $matches);

$links = $matches[0]; # $links is an array of all the links
echo $links[0];
?>

drawmack

http://us3.php.net/ereg_replace - it's covered here.

fizzwizz

Hi Drawmack

I've looked at ereg_replace and don't see how it can help.
The closest fuction I could find was preg_match but how to manipulate the arguements is beyond my expertise.

Could you be more explicit with how to do this? (Don't forget we are in the beginners section 🙂 )

Many thanks .... Fizz

fizzwizz

Hi Pyro

Sorry, ... I missed your answer.

Thanks.... that works just as I wanted it to!

The arguements used in this fnction

preg_match_all("/((?:http|ftp)s?:\/\/[-a-z0-9]+(?:.[-a-z0-9]+)+(?:[-\w:@&?=+,.!\/~'%#$](?<![.,?!]))?)/i", $string, $matches);

are confusing to me .... namely /((?:http|ftp)s?:\/\/[-a-z0-9]+(?:.[-a-z0-9]+)+(?:[-\w:@&?=+,.!\/~'%#$](?<![.,?!]))?)/i

Can you help explain this or point me to somewhere I can read up on it. I don't need it for the problem I have posted as you have solved it nicely for me .... but its obvious this is the next place I need to go with my php education 🙂

Best wishes .... Fizz

drawmack

from the ereg replace page

<?php
$text = ereg_replace("[[:alpha:]]+://[^<>[:space:]]+[[:alnum:]/]",
                     "<a href=\"\\0\">\\0</a>", $text);
?>

pyro

Here's the regex broken down.

color=darkblue[/color] - match http or ftp (the ?: specifies that it does not need to remember the match
s? - with an optional s (for secure sites)
[-a-z0-9]+ - one or more -, a-z, or 0-9
color=darkblue+[/color] - one or more set of one or more -, a-z, or0-9
color=darkblue?[/color] the characters inside this character set are basically all read literallly, except the \w, which is a-z,A-Z0-9_ the end part (?<![.,?!]) tells it not to match a . , ? or ! at the end, as that is probably going to be punctuation, and not part of the URL.
/i - make the entire regex case-insensitive

As far as where to learn, take a look at the PCRE section of the manual. Also, Jeffrey Friedl's book "Mastering Regular Expressions" is quite good.