Newbie Q

[David Relson]

At 08:45 PM 10/14/02, Clint Adams wrote:
> > Are you sure?  I just took a quick look (via "grep -i from") of the 0.3 
> and
> > 0.5 source code.  lexer.l recognizes a FROM token, which is used only in
> > register_words() to increment a message count.  The same code is in the
> > current version 0 0.7.4.
>
>No, I'm not sure.  I have mail from ESR saying that the problem should
>be fixed in 0.7; however it was 0.7 I was writing him about.

Clint,

I guess he never got to the task.  I can see classifying a batch of 
messages in a single bogofilter run as being a useful test 
tool.  Certainly, it'd go faster than "formail <mailbox -s bogofilter -s" 
which does lots of program runs with attendant wordlist 
opens/reads/write/closes, etc.  If we posit that bogofilter usually runs 
with procmail (or similar tool that'll feed one message at a time to it), 
the code is fine as it is.

David


For summay digest subscription: bogofilter-digest-subscribe at aotto.com