From rlpowell@digitalkingdom.org Wed May 12 23:49:51 2004 Received: with ECARTIS (v1.0.0; list lojban-list); Wed, 12 May 2004 23:49:51 -0700 (PDT) Received: from rlpowell by chain.digitalkingdom.org with local (Exim 4.32) id 1BOA2I-000084-Rn; Wed, 12 May 2004 23:49:42 -0700 Date: Wed, 12 May 2004 23:49:42 -0700 To: lojban-list@lojban.org Subject: [lojban] Re: "zo da bu" should not be valid (was Re: Re: My parser, SI, SA, and ZOI) Message-ID: <20040513064942.GV4461@digitalkingdom.org> Mail-Followup-To: lojban-list@lojban.org References: <20040510191837.GK5570@digitalkingdom.org> <20040510193718.13449.qmail@web41902.mail.yahoo.com> <20040513063029.GU4461@digitalkingdom.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20040513063029.GU4461@digitalkingdom.org> User-Agent: Mutt/1.5.5.1+cvs20040105i From: Robin Lee Powell X-archive-position: 7787 X-ecartis-version: Ecartis v1.0.0 Sender: lojban-list-bounce@lojban.org Errors-to: lojban-list-bounce@lojban.org X-original-sender: rlpowell@digitalkingdom.org Precedence: bulk Reply-to: lojban-list@lojban.org X-list: lojban-list On Wed, May 12, 2004 at 11:30:29PM -0700, Robin Lee Powell wrote: > I've spent a good portion of my free time since you posted this > thinking about this issue. It turns out that the answer is "Yes". > > There's a simple reason for this: it's the only solution that fits the > current cmavo definitions. > > zo takes a single Lojban word. bu takes a single Lobjan word. si > takes as single word, or an arbitrary string of non-Lojban text. > > This is how they are defined. There's something else that's worth mentioning: I've made the general meta-decision that a string of si keeps eating text; it doesn't get re-evaluated until the string of si is done. This is enshrined in canon in the form of the 4-si-for-zoi rule. That was probably incomprehensible. Here's an example: zoi gy broda gy si si gy -- what happens to the second si? There are two possible answers: "It gets re-evaluated", so we are left with: zoi gy broda si gy because si has no meaning inside of a zoi clause, and by erasing the zoi closer we opened up a zoi clause, which then eats the si. The other answer is: "It keeps eating text", so we are left with: zoi gy gy I've taken the latter position. The reason is simple: the former position is untenable. Completely. It makes it totally impossible to use si to erase zo, zoi, and possibly lo'u. Try it, you'll see. "zo da si si si si ...", for example: if interpreted in the former fashion, every odd si erases the word before it (which except for the first means it's erasing a si that got caught by zo) and every even si gets caught by zo. I'm mentioning all this partly on general principles and partly because I just had to add a special case for "ZO SI SI" into the SI handling rules in my PEG grammar to thwart the grammar's default behaviour given how SI and ZO are defined which is (you guessed it) to re-evaluate in place, leading to the infinite repitition shown above. This amused me.[1] There were already such rules in places for the five cases of ZOI and SI; the lack of such a rule for ZO goes back to me mis-understanding ZO and having it and its argument turn into a single word from the POV of the rest of the grammar. -Robin [1]: The fact that my grammar trying to do silly things with infinite, non-escapable repitition was amusing enough to me to cause me to write this giant e-mail is almost without question a sign of deep mental pathology, or something. -- http://www.digitalkingdom.org/~rlpowell/ *** I'm a *male* Robin. "Many philosophical problems are caused by such things as the simple inability to shut up." -- David Stove, liberally paraphrased. http://www.lojban.org/ *** loi pimlu na srana .i ti rokci morsi