procmail, mutt and bogofilter problems

[David Relson]

At 01:11 PM 11/14/02, jerry wrote:

>Hey list:

... [snip] ...

>:0HB:
>* ? bogofilter
>IN-Mail/spam
>
>:0HB:
>* ? bogofilter
>IN-spam
>
>If, however these two rules are commented out, mail is correctly
>sorted and procmail logging occurs.

I'm going to skip commeenting on procmail because I'm not that familiar 
with it.  Hopefully someone else will chip in with the info.

>I am also having difficulties understanding how to have bogofilter read
>and update the goodlist/spamlist, I've tried to pass my self sorted spam
>list thru bogofilter:
>
>bogofilter -p /home/jerry/Mail/inbox/spam -s -v -v

What version of bogofilter are you using?  The current version 0.8.0 would 
tell respond to the above command line with:

Error:  Invalid combination of options.
     Options '-s', '-n', '-S', and '-N' are used when registering words.
     Options '-p', '-u', '-e', and '-R' are used when classifying messages.
     The two sets of options may not be used together.

Basically '-s' and '-n' are used when you know you have spam (or ham) and 
are training bogofilter.  '-p' is used when you want bogofilter to classify 
a message and make a spam/ham determination.  As a last detail, there 
should be a "<" in front of the file name since bogofilter reads the 
message from stdin.

>bogofilter -p -s -u -e -v -v -v -l <
>/home/jerry/Mail/inbox/spam/spam3.txt >bogofilter2.txt
>bogofilter2.txt has a list of words and associated rankings, on that I
>am guessing. the final line has:
>86  P_Q_S_invsum_logsum    0.22808   0.98986  0.187268   -22.263
>-394.837
>Is the 0.98986, from above the overall score this email recieves?
>I apologize for the long message and would gladly accept any help.

You have generated the Rtable, which is a data frame that can be used for 
input to the R math package.  Use fewer '-v' flags to get more easily 
understood output.

I'd also strongly recommend reading the man page to get a feel for how 
bogofilter operates and how its command line options are used.

David