July 07, 2004
Comment spam alert

I just deleted 10 comment spams from a weight loss site. If you're using MT Blacklist, save yourself some grief and add the following regexp to your ban list:

weigh[\w\d\-_]*less[\w\d\-_]*rx[\w\d\-_]*\.[a -z]{2,}

Take out everything between the brackets and put a ".com" at the end if you want to kill the specific site instead of variations on it.

Posted by Charles Kuffner on July 07, 2004 to Administrivia | TrackBack

I just added MTCloseComments plugin (using the lovely Mt-plugins manager that lets me add plugins via a browser window from the office). It should help the whiterose blog empire cope with the spammers

Posted by: Michael on July 7, 2004 11:49 AM

Thanks. I'll admit, all those algebra-looking statements freak me out. Just like regular algebra!

Posted by: Mac Thomason on July 7, 2004 12:12 PM

off topic

from political wire

DeLay Retains Legal Team
House Majority Leader Tom DeLay (R-TX) has "retained lawyers to defend him in both a Congressional ethics probe and an ongoing investigation into Texas’ 2002 legislative races," Roll Call reports.


Posted by: bz on July 7, 2004 12:28 PM

For the record, Mac:

In a regular expression, any item inside square brackets can be matched. What's inside these square brackets is:

\w = any letter
\d = any number
\- = the "-" character
_ = the "_" character

The asterisk immediately after the square bracket means "any number of occurrances, including zero". Put together, the expression


means "any combination of letters, numbers, dashes, and underscores, whether they actually appear or not".

The expression "\.[a-z]{2,}" at the end means "match any letter combination that's at least two letters long and which follows a period", which takes care of all domain suffixes.

Thus, the whole thing not only matches the offending domain (which I can't name here as it'll be blocked by MT-Blacklist), it also matches any other mutatation of that domain name you can think of.

Posted by: Charles Kuffner on July 7, 2004 12:29 PM

BZ - Thanks, it's on my to-be-blogged list for the day (which is growing as we speak).

Posted by: Charles Kuffner on July 7, 2004 12:29 PM

I feel totally geeky that I was able to figure out the meaning of that code all on my own, from first principles. :)

Posted by: blurker gone bad on July 7, 2004 4:27 PM

I'll add my endorsement of MT-Close Comments; I set it at the default of 21 days old and older, and my spam count has (knock wood) fallen to nearly non-existent. It's in conjunction with MT-Blacklist, of course.

Posted by: Linkmeister on July 7, 2004 4:32 PM


Posted by: momo on July 21, 2004 1:01 PM