matching devious non ascii spam strings

[Dan Jacobson]

To match the name of a famous spam drug for spamassassin, I use
/v[il1\xA0-\xFF]agra/i
as they now are using all the accented versions of "i".
I suppose I will have to do the same for the a's etc. too.
I have just used the entire range I see on "man iso_8859_1" and more,
rather than whittle it down. I don't suppose there are much better ways.

 

[Chris Mattern]

To match the name of a famous spam drug for spamassassin, I use
/v[il1\xA0-\xFF]agra/i
as they now are using all the accented versions of "i".
I suppose I will have to do the same for the a's etc. too.
I have just used the entire range I see on "man iso_8859_1" and more,
rather than whittle it down. I don't suppose there are much better ways.

The Bayesian filters worry about that kind of crud automatically.
Let them do the work.

--
Christopher Mattern

"Which one you figure tracked us?"
"The ugly one, sir."
"...Could you be more specific?"