Date: Wed, 18 Jun 2003 07:22:45 -0400
To: lojban@yahoogroups.com
Subject: Re: [lojban] Re: mumym problem
Message-ID: <20030618112240.GD2095@ccil.org>
References: <20030617165623.104468ff.rizen@surreality.us> <LPBBLNNHBOGBGAINBIEFMEEPDHAA.ragnarok@pobox.com> <20030617172152.40049388.rizen@surreality.us>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20030617172152.40049388.rizen@surreality.us>
User-Agent: Mutt/1.3.28i
From: John Cowan <cowan@ccil.org>
Content-Length: 1063
Lines: 26

Theodore Reed scripsit:

> self.gismu_validator    =
> re.compile("^[bcdfgjklmnprstvxz]([bcdfgjklmnprstvxz][aeiou][bcdfgjklmnp
> rstvxz]|[aeiou][bcdfgjklmnprstvxz][bcdfgjklmnprstvxz])[aeiou]$")
> 
> All guesses are checked against that regexp. So you *can* guess fake
> words, but they must be valid gismu. 

Validity requires a good bit more than that.  "fgoto" for example will
pass but is invalid.  A CCVCV gismu must begin with one of the 48 valid
initial clusters, and even a CVCCV can't have a bogus cluster in the
middle, like "fapzu".

I am *not* saying that you should enforce these constraints, just that
that's what is required for a valid gismu form.  Personally, I would let
people guess any old @#$*.  Why not, it's their guess they are wasting.
It's the programmer's choice.

-- 
Values of beeta will give rise to dom!          John Cowan
(5th/6th edition 'mv' said this if you tried    http://www.ccil.org/~cowan
to rename '.' or '..' entries; see              jcowan@reutershealth.com
http://cm.bell-labs.com/cm/cs/who/dmr/odd.html)