invalid html warfare

[David Relson]

At 07:29 PM 5/27/03, Peter Jaques wrote:
>On 27 May 03,  3:13PM, David Relson wrote:
> > Works fine on my machine.  FWIW, I'm running bogofilter-0.13.2.1 :-)  I'll
> > bet you're running an older version :-(
> >
> > Download the newer version.  It worked for my simple test case
> > "vi<jccmkcol>a<bmcc>gr<junk>a".  It should work for your message.
>
>when you say it works, what do you mean? I ran this post through
>"bogofilter -R", & it didn't have a token for "vi", much less "viagra",
>which is what i'd hoped for;) (though "jccmkcol" did show up).
>
>i'm running 0.13.2.1.

Hi Peter,

It's good that you have the latest version.  I wonder why you're getting a 
different result than I am.  Must be a significant detail about which I 
don't know.

The quick parser check is to use "bogolexer -p -C" which prints the tokens 
(using bogofilter's default settings).  Is the mail labeled text/html?  If 
not, html parsing rules don't apply.  If you still can't figure what's 
going on, gzip the message and send it to me so I can take a look at it.

David