regex: replace links captions by url

[gdfhghjdfghgfhf]

Hello,

 

i have a string that looks like this (a lot of different links with different attributes)

 

<a href="http://www.mediafire.com/something">http://www.mediafire.com/something</a>
text text text text text text text text text text text 
<a href="http://www.somelink.com">
text text text text text text text 
<a href="http://megaupload.com/something/blabla">Some link without WWW</a>
text text text text text text text text text text text text text text 
<a title="somename" href="http://www.rapidshare.com/download" style="color:#000000" target="_blank">some confusing link</a>
...........more text here..........
<a href="http://www.microsoft.com">
....more text...........
<a href="http://www.4shared.com/download.php?file=myfile"><img src="a_link_with_an_image.gif"></a>

 

I'm trying to replace the links and change the captions with the URL of the link.... BUT only for links pointing to these sites:

mediafire.com, megaupload.com, rapidshare.com and 4shared.com

 

Additionally, i want to add [DL][/DL] tags between the replaced links

 

the links can come with or without WWW and with or without additional attributes (like Style, target, title, etc...)

 

Here is the result i would want with the above example:

[DL]<a href="http://www.mediafire.com/something">http://www.mediafire.com/something</a>[/DL]
text text text text text text text text text text text 
<a href="http://www.somelink.com">link</a>
text text text text text text text 
[DL]<a href="http://megaupload.com/something/blabla">http://megaupload.com/something/blabla</a>[/DL]
text text text text text text text text text text text text text text 
[DL]<a title="somename" href="http://www.rapidshare.com/download" style="color:#000000" target="_blank">http://www.rapidshare.com/download</a>[/DL]
...........more text here..........
<a href="http://www.microsoft.com">
....more text...........
[DL]<a href="http://www.4shared.com/download.php?file=myfile">http://www.4shared.com/download.php?file=myfile</a>[/DL]

 

Thanks a LOT to whoever can help me with this! I'm newbie to regex and i gave up after a lot of searchs

[cags]

Hmm... difficult to say, is that input and output definately correct or is it an example you quickly knocked up? The reason I ask is line 3 of the input is not a valid anchor link, not is the microsoft one further down. With regex this can make a big difference since the obvious way to match your pattern will require searching for the '</a>' part.

[sasa]

<?php
$test = '<a href="http://www.mediafire.com/something">http://www.mediafire.com/something</a>
text text text text text text text text text text text 
<a href="http://www.somelink.com">
text text text text text text text 
<a href="http://megaupload.com/something/blabla">Some link without WWW</a>
text text text text text text text text text text text text text text 
<a title="somename" href="http://www.rapidshare.com/download" style="color:#000000" target="_blank">some confusing link</a>
...........more text here..........
<a href="http://www.microsoft.com">
....more text...........
<a href="http://www.4shared.com/download.php?file=myfile"><img src="a_link_with_an_image.gif"></a>';
$patern ='#(<a\s[^>]*href\="([^"]*(mediafire\.com|rapidshare\.com|megaupload\.com|4shared\.com)[^"]*)"[^>]*>).*?</a>#';
echo preg_replace($patern, '[DL]$1$2</a>[/DL]', $test);
?>

[gdfhghjdfghgfhf]

Yes, cags is right, i did an error in my example (sorry!)

 

But sasa's code did exactly what i wanted with a little tweak!

 

thanks a lot!!!

[gdfhghjdfghgfhf]

i have one more problem

 

this code will not replace links with single quote, like this:

<a href='http://www.mediafire.com/file'>link</a>

[cags]

Untested, but something liket his should suffice..

 

$pattern ='#(<a\s[^>]*href\=("|\')([^\1]*(mediafire\.com|rapidshare\.com|megaupload\.com|4shared\.com)[^\1]*)\1[^>]*>).*?</a>#';

[gdfhghjdfghgfhf]

unless i am doing something wrong, this is not working

[cags]

Sorry, my bad, \1 assumed ("|\') was pattern 1, I didn't see the opening bracket at the start. It should have been \2. Which means you will also need to change the $2 in the replace pattern to $3 as it's been moved up one. Looking at the pattern though I'm not convinced it will work as their are no lasy quantifiers I think it may grab multiple urls in one.

[gdfhghjdfghgfhf]

though I'm not convinced it will work as their are no lasy quantifiers I think it may grab multiple urls in one.

I'm not sure what you mean.... how can i add that lazy quantifier ?

[thebadbad]

I would probably parse each URL with parse_url(), for more reliable results:

 

<?php
$str = '<a href="http://www.mediafire.com/something">http://www.mediafire.com/something</a>
text text text text text text text text text text text
<a href="http://www.somelink.com"></a>
text text text text text text text
<a href="http://megaupload.com/something/blabla">Some link without WWW</a>
text text text text text text text text text text text text text text
<a title="somename" href="http://www.rapidshare.com/download" style="color:#000000" target="_blank">some confusing link</a>
...........more text here..........
<a href="http://www.microsoft.com"></a>
....more text...........
<a href="http://www.4shared.com/download.php?file=myfile"><img src="a_link_with_an_image.gif"></a>';
function _callback($matches) {
$domains = array('mediafire.com', 'megaupload.com', 'rapidshare.com', '4shared.com');
$domain = parse_url($matches[2], PHP_URL_HOST);
//remove any sub domains
$parts = array_reverse(explode('.', $domain));
$domain = "{$parts[1]}.{$parts[0]}";
if (in_array($domain, $domains)) {
	$matches[0] = "[DL]{$matches[0]}[/DL]";
}
return $matches[0];
}
$str = preg_replace_callback('~<a\b[^>]+\bhref\s?=\s?([\'"])(.+?)\1[^>]*>.*?</a>~is', '_callback', $str);
echo $str;
?>

 

If you add a domain with a double TLD (e.g. .co.uk) to the $domains array, you would have to rewrite the code.