Re: [lojban] Re: Initial impression

[Jorge Llambías <jjllambias@gmail.com>]

On Mon, Jan 4, 2010 at 1:34 PM, Adam D. Lopresto <adam@pubcrawler.org> wrote:
>
> Actually, I think ZOhOI ({la'oi} and {zo'oi}) can be parsed just fine
> (though
> it requires pauses/stops/those hated periods before and after the non-lojban
> word).  It just requires a change to the grammar.  I have yet to see any
> actual problems.

It can be done, but it requires some rethinking of the morphology
algorithm. What the PEG algorithm currently does is first break a
string of phonemes into words, so for example "la'oi lopresto" will be
read as three words, "la'oi", "lo" and "presto", before doing any
syntactic parsing of the words. It doesn't care whether there is a
pause/space between "lo" and "presto" or not, or whether "la'oi" is a
defined cmavo or not. Selmaho ZOhOI would require the algorithm to pay
attention to pause/spaces at the syntactic level, which it currently
does not.

mu'o mi'e xorxes