From enkefalos@xxxxxxx.xxxx Mon May 24 23:24:58 1999
X-Digest-Num: 147
Message-ID: <44114.147.862.959273824@eGroups.com>
Date: Mon, 24 May 1999 23:24:58 PDT
From: "Arnt Richard Johansen" <enkefalos@xxxxxxx.xxxx
Subject: Parser/glosser suggestion

I have now been hammering at the beta glosser for about three weeks now,
and I must say that it has worked just fine. The only real objection I
have to it, is the speed and the inelegance of it (but I guess there is
little to do with the latter, since that probably would require someone
with good enough time to recode the thing in C).

However, a significant improvement of it, which I don't think would
require too much efforth, is to use different keywords on the gismu
according to conversion with selma'o SE.

For instance, take the sentence ".o'asai do nelci le lojbo xe fanva".
With the current version of the glosser, this is the result:

o'a sai do nelci le
(Pride!) (-strong emotion!) {fond one}: you fond {thing liked}: the
UI CAI *nelci-1* KOhA GISMU *nelci-2* LE

lojbo xe fanva
Lojbanic 5th place of translate
GISMU SE GISMU

Someone who isn't familiar with the place structure of "fanva", would
have to check the gismu list to find out what "5th place of translate"
refers to. But what I would like to see in an eventual final release of
the glosser, is the following:

o'a sai do nelci le
(Pride!) (-strong emotion!) {fond one}: you fond {thing liked}: the
UI CAI *nelci-1* KOhA GISMU *nelci-2* LE

lojbo xe fanva
Lojbanic translation
GISMU SE+GISMU

Is it possible to use data from oblique.key to achieve this?

co'o mi'e tsali