Date: Sat, 3 Feb 1996 02:53:03 -0700 (MST) From: Dave Andersen <angio@aros.net> To: giles@nemeton.com.au (Giles Lean) Cc: hackers@freebsd.org, jhk@time.cdrom.com Subject: Re: Any perl maniacs out there have a desire to improve our mail robot? Message-ID: <199602030953.CAA19117@terra.aros.net> In-Reply-To: <199602030721.SAA18649@nemeton.com.au> from "Giles Lean" at Feb 3, 96 06:21:14 pm
next in thread | previous in thread | raw e-mail | index | archive | help
Lo and behold, Giles Lean once said:
> On Fri, 02 Feb 1996 15:31:46 -0800 "Jordan K. Hubbard" wrote:
>
> > For those many folks still stuck with email as their only recourse, it
> > seems like we could try a little harder to make this robot more useful!
> >
> > Any takers?
>
> "Bags I!"
Here's another taker for it. :-)
> > First on the list would be hot links to current docs, whereever those
> > might be (I don't think we have ascii excerpts from the handbook on
> > freefall - suggestions?), then perhaps some ability to search the
> > archives in /usr/local/mail/archive and return messages matched (up to
> > some max threshold) by date/subject/pattern/mailing list.
>
> Links are easy. Searching is easy too, although too much data will
> beat up the machine unless we pre-index. Should this hook into the
> search engine on the Web server?
Complete agreement on both counts. The pre-indexing isn't my area of
specialty, though. We have a similar FAQ searching thing here, but we
manually index the entries.
> > If someone wanted to get *really* fancy, they could even try to write
> > an automated question parser which matches queries up with specially
> > prepared FAQ entries.. Should be, what, 3 or 4 lines of PERL? :-)
>
> Yeah, right! And we'll get Randal Schwartz to re-write the 3 or 4
> lines to 1 line and make it a sendmail alias. :-)
>
> A FAQ lookup by keyword is probably a good idea.
Doing this in a general way isn't hard at all. If you have your
indices, then all you do is strip the non-keywords out of the query, and
then return a matching list as though they'd done a keyword search. As
scary as it sounds, this actually isn't too bad to do in a couple of
lines. I'd be more than pleased to do the english-language query parsing
code. I actually have a neat little keyword-weighing idea floating
around in my head that's half-coded already.
> P.S. I also propose moving to perl5 -- any objections?
*shrug* :) No opinion either way.
-Dave Andersen
--
angio@aros.net Complete virtual hosting and business-oriented
system administration Internet services. (WWW, FTP, email)
http://www.aros.net/ http://www.aros.net/about/virtual/
"There are only two industries that refer to thier customers as 'users'."
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199602030953.CAA19117>