I've noticed more and more that whenever a fully qualified domain ends in a two letter combo that is actually an english word, this create a problem because often people will start a new sentence with one of these short words and forget to put a space after the last period and before the new "sentence.Like" this! (is that example, if "like" were a TLD, then "sentence.Like" might get picked up by a spam filter for SURBL checking.
The problem here is that MANY domains, for example, that end in ".to" are listed in SURBL.
I don't want my solution to be merely to only check domains within e-mails which start with "http://" because this would miss many spams.
But the only solution I can think of at the moment is ignore any potential domains within the message that end in these particular TLDs unless the "http://" is really there.
Any suggestions?
Rob McEwen PowerView Systems rob@PowerViewSystems.com
On Wednesday, August 9, 2006, 11:53:48 AM, Rob Systems) wrote:
I don't want my solution to be merely to only check domains within e-mails which start with "http://" because this would miss many spams.
But the only solution I can think of at the moment is ignore any potential domains within the message that end in these particular TLDs unless the "http://" is really there.
SpamAssassin takes one valid approach: look at how common mail clients render messages and check the strings they would treat as URIs. That tends to be things like http:// www. something, etc.
There's probably no universal or perfect answer.
Jeff C. -- Don't harm innocent bystanders.