Integration with training-based bayesian filters

[Sebastian Spaeth]

> The easy way is to expect the user enter the spam thread, which contains
> exactly one message (never seen longer spam threads, still wondering
> why...), and then press some key and confirm the destination, station
> purple hell.  Then you exit the thread.  To enter another one...

I just have a saved view where all possible junk mails are shown. Most
of the time they really all are spam as I can see from the subjects and
I simply press "*" and "+spam" to tag all those mails as spam.

A cron script occassionally passes all spam-tagged mails to a bayesian filter.

Something like this does not require me to fudge a lot with elisp (which
I am not good with).

> So, the crux is, to pipe the file to the classifier you need the filename,
> and the filename appears to be easily available only in the 'show' mode.

An within elisp solution would be elegant for sure. I've found that
tagging messages and then later processing those messages outside of
emacs works nicely for me. Here is the 5 line python script that shows
all filenames for messages matching a certain query:
http://notmuchmail.org/howto/

Sebastian