From lojbab@xxxxxx.xxxx Mon Dec 21 00:07:24 1998
X-Digest-Num: 32
Message-ID: <44114.32.114.959273824@eGroups.com>
Date: Mon, 21 Dec 1998 03:07:24 -0500
From: "Bob LeChevalier (lojbab)" <lojbab@xxxxxx.xxxx
Subject: Re: lujvo making and the 'tosmabru failure' test

At 07:37 AM 12/21/98 +0000, you wrote:
>From: Richard Curnow <richard@curnow.demon.co.uk>
>...
> Bob> If a lujvo will fail the "tosmabru" test, then that is the
> Bob> location at which it will break. The test is therefore to
> Bob> determine whether the remainder of the word, after taking off
> Bob> the first CV, is a valid lujvo of its own.
>
>Agreed, so does the following algorithm achieve the desired effect:
>
>A. Work through steps 1..4 of the algorithm as in the ref. grammar.
>
>B. If the candidate lujvo starts with at least one CVC rafsi, consider
>the form after removing the initial CV. Working from the left, try to
>partition this into valid rafsi (not necessarily at the points where
>rafsi were originally joined earlier), possibly with hyphens between
>them; if this can be done then the test fails and the lujvo has to
>start CVCyC..., otherwise it passes and the extra 'y' is not needed.

I think this has the same effect and the book algorithm, but I am anything
but a master of algorithm equivalencing.