Extracting Root Domain

[-Karl-]

Hello,

 

I'm having difficulty extracting the root domain from an URL, whilst excluding a subdomain if one is there.

 

Basically:

 

				 $txt = 'http://i.imgur.com';
				 $parse = parse_url($txt);
				 $url = $parse['host'];

 

Returns i.imgur.com, but I only need imgur.com. However, the reason it's complicated is because the url could also be www.imgur.com, so you can't simply explode by ".", then there's also the problem with other tld's.

 

For example, m465.imageshack.us, all I need it to return is imageshack.us.

 

What would be the best way to achieve this?

Edited October 22, 2012 by -Karl-

[cyberRobot]

Couldn't you explode the value as mentioned to get the TLD. Then just test if the first part contains "www"?

[-Karl-]

Couldn't you explode the value as mentioned to get the TLD. Then just test if the first part contains "www"?

Not really, since 456.imageshack.us/image.png contains three periods, whereas tinypic.com/image.png contains two.

[cyberRobot]

That would be another test. If the array created by explode contains two items, you have what you need. If it has more, check if the first one is "www".

[shaddowman]

I think the best possible solution is the use of built in PHP function parse_url.

You can extract the domain name and might get the exact domain name by the help of string manipulator functions susch as substr() and strpos()

 

 

<?php
$url = parse_url("http://123.456.imageshack.us/image.png");
print_r($url);
?>

[-Karl-]

I think the best possible solution is the use of built in PHP function parse_url.

You can extract the domain name and might get the exact domain name by the help of string manipulator functions susch as substr() and strpos()

 

 

<?php
$url = parse_url("http://123.456.imageshack.us/image.png");
print_r($url);
?>

If you read my first post, parse_url() is exactly what I am using...

 

I then explained the reason why it's not suffice.

[-Karl-]

Ended up doing an explode:

				 $parse = parse_url($image);
				 $url = $parse['host'];
				 $explode = explode('.',$url);
				 if(count($explode) > 2) {
					 $url = $explode[1].'.'.$explode[2];
				 }

 

Just need to do some testing for all possible variables now.

Edited October 22, 2012 by -Karl-

[Psycho]

    $parsed = parse_url($image);
   $host_parts = explode('.', $parsed['host']);
   $domain = implode('.', array_slice($host_parts, count($host_parts)-2));

Edited October 22, 2012 by Psycho

[ignace]

Based off Psycho's solution. You can pass a PHP_URL_HOST constant to parse_url also you can pass a negative value as offset to array_slice.

 

$image = 'http://123.456.imageshack.us/image.png';
$domain = implode('.', array_slice(explode('.', parse_url($image, PHP_URL_HOST)), -2)); // imageshack.us

Edited October 22, 2012 by ignace

[kicken]

Note that some extensions have multiple periods (eg bbc.co.uk). If you're concerned about domains like that you'll need to create a list of all the possible TLDs to test against.

[Psycho]

. . . also you can pass a negative value as offset to array_slice

 

I should have known that.