Need help with ereg_replace on a URL

[yktan]

Hi guys,

I've been searching the net for hours on what I'm trying to do and I suddenly saw that this site has a lot of resources!

Anyway, I'm not very strong in regular expressions, so I hope someone can help me with this. Basically, I want to search through a string and replace all occurances of a URL and reconstruct it to another URL.

For example, I want to use ereg_replace to transform this URL:
http://www.mydomain.com/index.php?option=com_forum&Itemid=37&page=viewtopic&t=194&highlight=test
to this:
http://www.mydomain.com/forum/viewtopic.php?t=194

As you can see, the variable in bold is the one that I want to retain.

Thank you in advance for your help.

Regards,
yktan

[dalar]

Hi,

quick question, do all your links contain &highlight=test at the end of the url? or is it different for some links?

[yktan]

Thanks for your quick response dalar. Not all links contain the &highlight=test after the URL.

[yktan]

Actually to simplify matters, all I need to do is for ereg_replace to detect all strings that starts with http:// that contains viewtopic&t=x (where x is a number of unknown digits) and replace the whole string with a http://whatever.com/viewtopic?t=x. The replace string http://whatever.com/viewtopic?t= can be static but I do not know how to get the number from the string that is found.

[dalar]

I was going to mention ereg_replace too, but things are a little more complicated since the ending bit is different for all links.

Going with your way is the best, once we remove the main link http://whatever.com/viewtopic?t= we know that the rest of the line contains numbers and then after that it contains other characters which are not needed.

We also can;t assume that every number are a certain length in digits. We could use the explode() function to then go through each number until it hits a character - that way we know when we have the full topic id number.

Of course this whole process is much simpler using an actual regular expression, but I would need to look that up somewhere in a book.

Are your bunch of links stored in an array?

[yktan]

Actually, what's happening is I'm detaching phpbb from Mambo to a brand new phpbb installed on another directory. I just need to use ereg_replace on all URLs posted in the old phpbb to use the new URL...

[dalar]

ok, i managed to come up with this:

<?php

// arrays containg the URL data
$URLs = array("http://whatever.com/viewtopic?t=321&blah=laba", 
"http://whatever.com/viewtopic?t=262&blah=laba", 
"http://whatever.com/viewtopic?t=1732&blah=laba", 
"http://whatever.com/viewtopic?t=32&blah=laba",
"http://whatever.com/viewtopic?t=6322&blah=laba");

// print out the links first with the unwanted data
	foreach ($URLs as $value)
	{
		echo "$value <br />";
	}

echo "<br/><br/>";

//go through each array

foreach ($URLs as $value)
{

//cut out the main body of the link, so were left with the topic ID and unwanted data

$choppedURL = ereg_replace("http://whatever.com/viewtopic\?t=","",$value);

// print out the main URL body, we are going to echo out the topic ID numbers to this string.

echo "http://whatever.com/viewtopic?t=";


$i=1;
$j=0;

// go through each letter/number and verify if its a number, if it is echo it.
// if we get to a non-number then close the while loop and move to the next

					while ($i == 1) {

						$digit = substr($choppedURL, $j, 1);

							if( is_numeric( $digit ) ) {

									echo "$digit";
							}

					else
							{
								$i=0;
							}

							$j++;

					}


echo"<br />";

}

?>

give it a shot and run it, it uses the array to get the data initally.

[Drakla]

No need for all this ereg faffing. All the variables passed via the url are available in the $_GET array, so just do a search, or key search on that (look at the array functions on php.net), pull the value out and tag it onto your destination url.

That's presuming these urls aren't just stored in an array - if they are then you can use [man]parse_url[/man] and the [man]parse_str[/man] on the query to get all the variables out.

[yktan]

Thanks for taking the time to help me dalar, but I think I didn't explain it clear enough and confused you.

What I need is to transform a string like this:
http://www.mydomain.com/index.php?o...page=viewtopic&t=194&highlight=test
to
http://www.mydomain.com/forum/viewtopic.php?t=194

Anyway, I found that I can live with the highlight=test part in the result string, so the following super simple regex managed to do the job

$url = "abc def ghi http://www.mydomain.com/index.php?option=com_forum&Itemid=37&page=viewtopic&t=194&highlight=test xxx www qqq http://www.mydomain.com/index.php?option=com_forum&Itemid=37&page=viewtopic&t=194";

$url = ereg_replace("http://www.mydomain.com/index.php\?option=com_forum&Itemid=[0-9]*&page=viewtopic&", "http://www.mydomain.com/viewtopic?", $url);

The result is:
abc def ghi http://www.phplinkdirectory.com/viewtopic?t=194&highlight=test xxx www qqq http://www.phplinkdirectory.com/viewtopic?t=194

What I am curious is, why does this code:

$url = "abc def ghi http://www.mydomain.com/index.php?option=com_forum&Itemid=37&page=viewtopic&t=194&highlight=test xxx www qqq http://www.mydomain.com/index.php?option=com_forum&Itemid=37&page=viewtopic&t=194";

$url = ereg_replace("http://www.mydomain.com/index.php\?.*&page=viewtopic&", "http://www.mydomain.com/viewtopic?", $url);

Gives the result of:
abc def ghi http://www.phplinkdirectory.com/viewtopic?t=194

Why does it replace the first instance only and stops?

Thanks.