Re: [bpfk] Proposal: sumti must always be tagged with "tag" even if it's elidible.

[Ilmen <ilmen.pokebip@gmail.com>]

On 08/04/2015 01:16, Jorge Llambías wrote:

On Tue, Apr 7, 2015 at 10:46 AM, Ilmen <ilmen.pokebip@gmail.com> wrote:

abs_term = "absorbable term", "term that becomes a selbritcita when followed by a selbri".

I implemented this absorbable/non-absorbable term grammar split in order to implement the standard behavior of selbritcita as per Jbofihe; that is, {lo nu brodo ba brode} parses as {lo nu brodo ba CU brode} in the standard version of Camxes; however, {mi ba brode gi'e pu brodo} parses as {mi CU ba brode gi'e pu brodo}, so the behavior of "ba" varies depending on its environment: if "ba" is within a bridi-tail, it is non-absorbable (it cannot become a selbritcita), but if it is in a bridi-head, it is absorbable as a selbritcita.

Does "mi ba brode gi'e pu brodo" parse as "mi CU ba brode gi'e pu brodo" and not as "mi ba (ku) CU brode gi'e pu brodo" in standard camxes?  I think it's the latter. "tag selbri" _should_ have priority over "tag (ku) selbri", but I don't think it does in current camxes.

Standard Camxes.js: ([mi CU] [{ba brode} VAU] [gi'e {<pu brodo> VAU} VAU])

Does Jbofihe use camxes? I thought it didn't, but this page doesn't say: http://mw.lojban.org/papri/jbofi%27e

No, Jbofihe is a C program, it is not based on a PEG grammar, if I recall correctly.

I know this tag grammar split is rather ugly, but at this time I hadn't been able to find a better solution. Any alternative implementation is welcomed.

Which sentence currently uses "term" but couldn't use "abs_term" instead? Hmm, I just looked, and it seems that gek_termset, LAhE term, and BE term could have a (slight) problem, in the sense that KU would not be elidible in "ge ca (ku) gi ba (ku) broda", "la'e ba (ku) broda", "broda be ba (ku) brode".

Standard Camxes.js:
ge ca ku gi ba ku broda  -->  ([{ge <ca ku> gi <ba ku>} CU] [broda VAU])
ge ca gi ba broda  -->  SyntaxError: Expected [,] but "b" found.
la'e ba ku broda  -->  SyntaxError: Expected [,] but "b" found.
la'e ba broda  -->  SyntaxError: Expected [,] but "b" found.
broda be ba ku brode  -->  ([{broda <be (¹ba ku¹) BEhO>} brode] VAU)
broda be ba brode  -->  ([{broda <be (¹ba KU¹) BEhO>} brode] VAU)

Experimental Camxes.js:
ge ca ku gi ba ku broda  -->  ([ge {ca ku} gi {ba ku}] [CU {broda VAU}])
ge ca gi ba broda  -->  SyntaxError: Expected [,] but "b" found.
la'e ba ku broda  -->  ([FA {la'e <ba ku> LUhU}] [CU {broda VAU}])
la'e ba broda  -->  ([FA {la'e <ba KU> LUhU}] [CU {broda VAU}])
broda be ba ku brode  -->  (CU [{broda <be (¹ba ku¹) BEhO>} brode] VAU)
broda be ba brode  -->  (CU [{broda <be (¹ba KU¹) BEhO>} brode] VAU)

So yeah, {ge ca gi ba broda} is broken for both parsers.
{la'e ba (ku) broda} is an experimental grammar if I recall correctly, so it's normal that the standard Camxes rejects it.

(BTW, "ge ca (ku) gi ba (ku) klama" seems to be failing if we elide the first "ku".)

Maybe it's better to just make the KU of "tag KU" not elidible in those three special contexts (are there more?).

Yeah, I've noticed this flaw a few days ago. I'm not sure whether this can be fixed.

mi'e la .ilmen. mu'o

--
You received this message because you are subscribed to the Google Groups "BPFK" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bpfk-list+unsubscribe@googlegroups.com.
To post to this group, send email to bpfk-list@googlegroups.com.
Visit this group at http://groups.google.com/group/bpfk-list.
For more options, visit https://groups.google.com/d/optout.