[Spambayes] Training spambayes on a large set of mailbox files
edwardam at interlix.com
Wed Feb 19 19:43:00 EST 2003
Comments inline ...
BTW: on another topic ... Does spambayes support any SQL servers?
If not I'm tempted to look at adding support....
On Wed, 2003-02-19 at 18:56, Neale Pickett wrote:
> Edward Muller <edwardam at interlix.com> writes:
> > Is there an easy way to train spambayes on a large set of mailbox files?
> > It would be a read pain in the but to have to use multiple options on
> > the command line for each mailbox file....
> Sure. Do it from python:
> >>> import mboxtrain
> >>> h = hammie.open('/path/to/your/database', True, 'c')
> >>> for m in ('hambox1', 'hambox2', 'hambox3'):
> ... mboxtrain.train(h, m, False, False)
> (I haven't tested this, so back up your mailboxes before you do it.)
hehe. I'll probably do something like this ...
> If you're in unix, just do this:
> $ for i in hambox1 hambox2 hambox3; do
> > mboxtrain.py -g $i
> > done
That would work if they were all in the same directory. I have a fairly
deep directory structure of mailboxes (IMAP server)...
I'll probably just construct a loop and a function or two in python to
> Although maybe this should be yet another option to mboxtrain.
Interlix - President
Web Hosting - PC Service & Support
Custom Programming - Network Service & Support
More information about the Spambayes