Difference between revisions of "camxes"

From Lojban
Jump to: navigation, search
m
 
m
Line 1: Line 1:
  
See xorxes' definitions at
+
A Lojban parser in Java by [[jbocre: camgusmis|camgusmis]] and [[User:xorxes|xorxes]].
  
[http://jbovlaste.lojban.org/dict/mabla]
+
* [http://www.digitalkingdom.org/~rlpowell/hobbies/lojban/grammar/index.html eb page.]
 +
* [http://www.digitalkingdom.org/~rlpowell/hobbies/lojban/grammar/lojban.peg.txt ojban PEG Grammar]
  
and
+
* [http://www.digitalkingdom.org/~rlpowell/hobbies/lojban/grammar/lojban_morphology.peg.txt ojban PEG Morphology]
  
[http://jbovlaste.lojban.org/dict/zabna] for the actual
+
== Bug reports / requests ==
  
definiton proposals.
+
* "ke ko'a ce ko'e ke'e ce ko'i" does not parse. —[[jbocre: kpreid|kpreid]]
 +
* The parse tree for "ko'a ce ke ko'e ce ko'i ke'e" is unnecessarily hard to process, as it contains the ke/ke'e brackets in the same parent node (sumti1) as the elements of the outer list; the (somewhat postprocessed) output in TermL syntax is sumti1(sumti2, joikEk, KE, sumti, KEhE); it would be preferable for the last three elements to have a subnode to themselves. —[[jbocre: kpreid|kpreid]]
  
We should distinguish the way ''mabla'' is used from the way it is
+
* Non-Lojban words in ZOI quotes yield invalid TermL: "zoi zoi $ zoi" contains innerWord( $ ) when it should be innerWord( "$" ). 'zoi zoi " zoi' is probably another useful test case. —[[jbocre: kpreid|kpreid]]
 
 
officially defined. ''mabla'' cannot mean "this selbri is used in a
 
 
 
derogatory fashion", because components of lujvo are not used to
 
 
 
make comments on the lujvo they form, much less on its use. The
 
 
 
actual usage definition of ''mabla'' is something like
 
 
 
"x1 is
 
 
 
deplorable/wretched/shitty/awful/rotten/miserable/contemptible/
 
 
 
/crappy/inferior/low-quality in property x2 by standard x3; x1
 
 
 
stinks/sucks in aspect x2 according to x3"
 
 
 
The official definition is something that could be used to talk
 
 
 
'''about''' language, but it is not what corresponds to the actual
 
 
 
usage of the word. ''mabla'', in usage if not officially, '''is''' a
 
 
 
derogatory word, it doesn't '''mean''' "is a derogatory word".
 
 
 
The same applies to ''zabna''.
 
 
 
From lojban-beginners:
 
 
 
<pre>
 
 
 
> > I read the following short conversation on the main Lojban list:
 
 
 
> > (Person A) lo lijda prenu cu je'a carmi mabla
 
 
 
[[jbocre: ...|...]]
 
 
 
> > (Person A) Religious people are indeed extremely derogatory!
 
 
 
[[jbocre: ...|...]]
 
 
 
> Yeah, basically. Person A presumably meant to say {se mabla} rather than
 
 
 
> {mabla}, which makes the first statement more sensible. Replacing {mabla}
 
 
 
> with {se mabla}:
 
 
 
> Person A: "(I) do indeed intensely deride religion people."
 
 
 
</pre>
 
 
 
The gi'uste definition of {mabla} is hoplessly confused:
 
 
 
mabla [[jbocre: mal|mal]] derogative
 
 
 
x1 is a derogative connotation/sense of x2 used by x3;
 
 
 
x3 derogates/'curses at' x2 in form x1 [jbocre: bloody (British sense),
 
 
 
fucking, shit];
 
 
 
{mabla} is thus defined in three inconsistent ways. According to the
 
 
 
first definition, it is a relationship between a meaning and an
 
 
 
expression, like {smuni}. Obviously neither x1 nor x2 of that first
 
 
 
definition makes sense for {lo lijda prenu} because people are not
 
 
 
expressions nor connotations/senses of expressions.
 
 
 
The second definition (which is inconsistent with the first) would
 
 
 
allow {lo lijda prenu} in the x2, it is possible to curse at people.
 
 
 
But I doubt that's what the original poster had in mind. He wasn't
 
 
 
informing us that he is in the habit of insulting religious people,
 
 
 
or that he insulted them, or that he will insult them, nor even that
 
 
 
he was in the process of insulting them. Even if all that is true,
 
 
 
he did not give the impression to me that that is what he was trying
 
 
 
to tell us.
 
 
 
He was using the third definition of {mabla} (inconsistent with the
 
 
 
two previous ones) to insult religious people (not to tell us that
 
 
 
he was insulting them). He basically meant to say something like
 
 
 
"religious people are shit".
 
 
 
Even though this third definition of {mabla} is the least explicit
 
 
 
one in the gi'uste, only appearing in brackets and with no explicit
 
 
 
place structure, I do believe it is the correct one. {mabla} was
 
 
 
meant and has mostly been used _as_ a derogatory word, _as_ a curse
 
 
 
word, not as a word that _means_ "x1 is derogatory sense of
 
 
 
(word/expression) x2", nor as a word that means "x3 curses at
 
 
 
(person/object) x2".
 
 
 
So even though I don't approve of the content of what the poster
 
 
 
said, I have to admit that he used the word in the way I consider
 
 
 
should be used to say what he meant.
 

Revision as of 16:45, 4 November 2013

A Lojban parser in Java by camgusmis and xorxes.

Bug reports / requests

  • "ke ko'a ce ko'e ke'e ce ko'i" does not parse. —kpreid
  • The parse tree for "ko'a ce ke ko'e ce ko'i ke'e" is unnecessarily hard to process, as it contains the ke/ke'e brackets in the same parent node (sumti1) as the elements of the outer list; the (somewhat postprocessed) output in TermL syntax is sumti1(sumti2, joikEk, KE, sumti, KEhE); it would be preferable for the last three elements to have a subnode to themselves. —kpreid
  • Non-Lojban words in ZOI quotes yield invalid TermL: "zoi zoi $ zoi" contains innerWord( $ ) when it should be innerWord( "$" ). 'zoi zoi " zoi' is probably another useful test case. —kpreid