|
From: Joe R. J. <jj...@cl...> - 2003-11-15 04:09:04
|
On Fri, 14 Nov 2003, Gilles Detillieux wrote:
> Date: Fri, 14 Nov 2003 12:08:00 -0600 (CST)
> From: Gilles Detillieux <gr...@sc...>
> To: Joe R. Jah <jj...@cl...>
> Cc: "ht://Dig developers list" <htd...@li...>
> Subject: Re: [htdig-dev] 3.2.0b5 Testing
>
> According to Joe R. Jah:
> > On Thu, 13 Nov 2003, Gilles Detillieux wrote:
> > > According to Joe R. Jah:
> > > > htfuzzy metaphone dumps core, but it works fine with endings, etc.
> > >
> > > Just a hunch, but I'd guess that you had a metaphone database left over
> > > from 3.1.6, and that the newer DB code in 3.2.0b5 doesn't like it. We
> > > had put some tests for this in some of the other programs, but maybe not
> > > this one. Try deleting the databases and running htfuzzy metaphone again.
> > > The same would probably go for soundex and accents databases, if you use
> > > either of those.
> >
> > Spot on! I had an older db.metaphone.db left over; none of the other
> > supported algorithms had any problem:
> > soundex
> > accents
> > endings
> > synonyms
>
> endings and synonyms are generated a bit differently, and most importantly
> are built in a temporary spot then moved into place, so they don't collide
> with any existing databases until the new one is complete. soundex and
> accents use the same code as metaphone for writing the database from the
> generated word list, so they'd all potentially have the same problem.
> The scrambling of data may not have been bad enough to cause a segfault
> (yet) but may have led to corrupt databases. If you had existing
> databases for accents and/or soundex in place, built by 3.1.x, before
> regenerating them for 3.2.0b5, you should remove them and try again,
> just as for metaphone.
Thanks; I see. It'd be nice if it could be made to quit, spitting a
warning, instead of dumping core.
Regards,
Joe
--
_/ _/_/_/ _/ ____________ __o
_/ _/ _/ _/ ______________ _-\<,_
_/ _/ _/_/_/ _/ _/ ......(_)/ (_)
_/_/ oe _/ _/. _/_/ ah jj...@cl...
|