My mail server has an anti spam feature that enables it to scan the body content of any incoming mails for certain words. And, when it founds them the email is either rejected or deleted.

Now, I began to realize that it may be the best approach to combat spam email because I can create filters that would reject emails which contains certain hyperlink names, like the spammers’ domain name, email address or even his porn image file name.

I harvest most spammer’s domain name, email addresses, image file name and source (img src="http) from looking at their html source code. If they use html number code, I can always use them on my filter, as well. I also translate them at http://javascript.internet.com/equiv...-revealer.html just to know what they mean in plane English.

I collect them most from links like:
Sorry if this email caused you inconvenience.
to stop me sending you more please go here
To stop this go here
remove me
Click here to Claim Your prize!
no more? push here
Visit us at:
not for me
CLICK HERE FOR MORE INFO
Since, RBL and firewall ACL can only block already known offending hosts and IP addresses – you can still get the spammer email again and again from newly found open relay and open proxies. But, if you block it through mail body content filter, you will never see the spammer email again unless he/she edits it with a new domain name and so on.
The good thing about this approach is that the spammer may never know - why his/her email is being rejected. The other good thing is that the spammer will have to pay for a new domain names and host service, if the spammer email has to have value like an email reply from the victim or a web visit from the victim.
What do you guys think about this approach? And, why it’s not a common feature on most email server software?