How? parse HTTP_REFERER to grab domain

Anon

I need to parse the HTTP_REFERER variable and grab the refering domain name. There seems to be many ways to go about this however, I was hoping to hear suggestions on what might be the best and most efficient method.

The reason I ask is that I have a web site which has multiple domains names pointed to it. I would like to be able to set a session varaible to check against to define domain specific content.

I understand I can use regular expressions to parse the refering URL, but have heard those are slow, I have also read elsewhere that I can use certain string functions from PHP to do this. In either case I am lost in how to write the expression. The problem is that not only are there differeing domain names pointed to the site, they are aliased on the server so...

http://www.domain1.com
http://domain1.com
http://www.domain2.com
http://domain2.com

all point to the sane site.

Any suggestions would be ever so greatly appreciated. Thanks in Advance.

Alan T. Miller

Anon

Okay, I have done a little bit of digging and now know that I can use $HTTP_HOST to get the actual host name the person typed in. That is great and works like a charm.

What I need to do now though is write an expression that will parse the $HTTP_HOST variable and give me the actual domain name. As it stands, people can reach the hosts by either tying in:

http://www.domain1.com or
http://domain1.com

the $HTTP_HOST variable will return the string for each of the above...

www.domain1.com
domain1.com

how can I strip out the "www" if it exists and just grab the actual domain name?

So in the end I would have a script that went something like this:

// Grab host name and define
// hostname variable

$hostname = "$HTTP_HOST";

if ($hostname == "domain1) {
start session and
load domain1 specific info
}

if ($hostname == "domain2) {
start session and
locad domain2 specific info
}

Again, any help is always more than appreciated.

studot

Try this:

if (ereg ("www.", substr ($domain_name, 4))):
/ Routine to www. prefix /
else:
/ Routine to handle domain without prefix /
endif;

Hope it helps,

StuG