SPAM with lots of random words and good words

[David Relson]

On Tue, 20 Jan 2004 10:30:52 -0000
pgb at adelard.com wrote:

> On 19 Jan 2004 at 14:28, David Relson wrote:
> 
> > Currently bogofilter recognizes the Content-Disposition header lines
> > and keeps track of whether inline or attachment was specified. 
> > However, the information is not begin used.
> > 
> > Bogofilter also keeps track of Content-Type and ignores tokens from
> > mime parts labeled as application, image, and message.
> > 
> So I infer that an attached spam email would be classed as a mime 
> message part and be ignored.
> i.e. it is safe to attach a spam without zipping it first - right??

Try forwarding a message to yourself and then run "bogofilter -vvv <msg"
to see what happens.  Alternatively, gunzip the attached file and run it
through bogofilter.

The RFC822 attachment is a mime-part, bogofilter knows about them, and
scores many of them.  Binary mime-parts such as images, spreadsheets,
word processor documents, etc are not scored.

Looking at my earlier message, I realize the statement about
Content-Disposition header lines is ambiguous.  The lexer recognizes the
lines and sets internal mime flags.  However bogofilter doesn't use the
flag while processing the message.

Hope this helps :-)

David
-------------- next part --------------
A non-text attachment was scrubbed...
Name: msg.rfc822.txt.gz
Type: application/octet-stream
Size: 607 bytes
Desc: not available
URL: <http://www.bogofilter.org/pipermail/bogofilter/attachments/20040120/170c6db1/attachment.obj>