| |
In lugnet.off-topic.pun, Tom McDonald writes:
> In lugnet.off-topic.fun, Todd Lehman writes:
> > In lugnet.off-topic.fun, Robert Munafo writes:
> > > That algorithm won't find the following [...] that didn't have "when
> > > replying". You could special case the third week of April and use the
> > > "when replying" test for everything since then.
True -- I knowingly didn't take that into consideration. Yet.
This should be fairly easy to fix -- just buffer all text after the
signature /\bTom McD\b/ but before a ( blank line or a quoted line or a header
line ) and grep that for "spam".
I'll probably get around to doing that Mondayish if no one else has by then.
> > Since there are only about 350 of 'em, and editing will probably be • necessary
> > on most of 'em anyway, I'll bet it's quicker just to pop the whole thing • into
> > an editor and delete stuff by hand. If you could do one every 10 seconds,
> > then you'd have it done in an hour, which is certainly less time that it • would
> > take to write a program to do same correctly in all cases. :)
True, but I got a program which get ~95% of them within five minutes. Is this
a case of fuzzy logic justification? :-,
> And yes, there was a spampuck sans spamcake. Also, not all of the directives
> start with "when replying" either. I haven't counted, but I managed to slurp
> dozens so far from a Lugnet search. I didn't expect to have a ready-for-
> primetime list with 24 hours anyway.
I took the cakeless spams into consideration. Check out my latest at:
http://www.lugnet.com/off-topic/geek/?n=252
Cheers,
- jsproat
|
|
Message is in Reply To:
71 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
