Message-Id: <199812210810.DAA06672@stmpy.cais.net>
Date: Mon, 21 Dec 1998 03:07:24 -0500
To: Richard Curnow <richard@curnow.demon.co.uk>
From: "Bob LeChevalier (lojbab)" <lojbab@lojban.org>
Cc: lojban@onelist.com
In-Reply-To: <m2u2yqaqtz.fsf@curnow.demon.co.uk>
References: <"Bob LeChevalier's message of Mon, 21 Dec 1998 02:00:26 -0500> <199812210704.CAA06313@stmpy.cais.net>
Mailing-List: list lojban@onelist.com; contact lojban-owner@onelist.com
Precedence: bulk
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Subject: [lojban] Re: lujvo making and the 'tosmabru failure' test
Content-Length: 1530
Lines: 34


From: "Bob LeChevalier (lojbab)" <lojbab@lojban.org>

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.


------------------------------------------------------------------------
To unsubscribe from this mailing list, or to change your subscription
to digest, go to the ONElist web site, at http://www.onelist.com and
select the User Center link from the menu bar on the left.
------------------------------------------------------------------------
To unsubscribe, send mail to lojban-unsubscribe@onelist.com