From phma@webjockey.net Fri Dec 13 15:18:52 2002 Return-Path: X-Sender: phma@ixazon.dynip.com X-Apparently-To: lojban@yahoogroups.com Received: (EGP: mail-8_2_3_0); 13 Dec 2002 23:18:52 -0000 Received: (qmail 25419 invoked from network); 13 Dec 2002 23:18:52 -0000 Received: from unknown (66.218.66.218) by m10.grp.scd.yahoo.com with QMQP; 13 Dec 2002 23:18:52 -0000 Received: from unknown (HELO neofelis.ixazon.lan) (208.150.110.21) by mta3.grp.scd.yahoo.com with SMTP; 13 Dec 2002 23:18:51 -0000 Received: by neofelis.ixazon.lan (Postfix, from userid 500) id E4A723C5DF; Fri, 13 Dec 2002 18:18:50 -0500 (EST) Content-Type: text/plain; charset="iso-8859-1" To: lojban@yahoogroups.com Subject: Re: [lojban] Word resolution algorithm so far Date: Fri, 13 Dec 2002 18:18:47 -0500 X-Mailer: KMail [version 1.2] References: In-Reply-To: X-Spamtrap: fesmri@ixazon.dynip.com MIME-Version: 1.0 Message-Id: <0212131818470G.03697@neofelis> Content-Transfer-Encoding: 8bit Sender: phma@ixazon.dynip.com From: Pierre Abbat X-Yahoo-Group-Post: member; u=92712300 On Friday 13 December 2002 18:01, Jorge Llambias wrote: > My preferrence would be to declare "y" next to a vowel > impermissible always, just as "'" next to a consonant. > > Also, I would prefer to restrict the appearance of "y" next > to "'" to the word ".y'y" and nowhere else. > > If so, all of the above strings should give errors. Actually, since I check for a string containing nothing but cmavo before I check for medial 'y', all these are handled by the cmavo splitter. It currently calls "da'yba" a single cmavo, though, which is wrong. According to the Book, "a'y" through "u'y" and "y'a" through "y'u" are reserved, so they should be called valid by the lexer and called errors by the parser when it can't find their selma'o. "doy" and "byo" will be caught later by a step that examines all the resolved words for permissible adjacent pairs. phma