[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [edict-jmdict] Re: いい/よい POS matters

Le 2014-07-29 07:36, Stuart McGraw smcg4191@frii.com [edict-jmdict] a écrit : 
On 07/26/2014 03:14 AM, Jean-Luc Léger reiga@dspnet.fr.eu.org
[edict-jmdict] wrote:

Hi,

For one word PoS, I think it would be easier to name them with the
concerned word.
Something like adj-i-ii, v1-kureru, cop-da, cop-desu and so on.
Because the more weird code we had, the more chance someone is gonna
mistake one for another.
Since it's for only one entry, the overhead is not going to be important 
:)

JL


I would agree if assigning all PoS keywords from scratch.

But given that there are already many short, obscure keywords
(88 are 7 characters or less, only 4 longer than that) and
that code that uses them may take advantage of their shortness
(for example they align in a column properly when included
between \t characters, except for the 4 miscreants), I would
prefer to continue with short obscure tags.
Well that's exactly because we already have many obscure keywords that I am against adding new ones when we can help it. I am at work so I can't check but I don't think we have many one-word keywords yet, so their obscure naming can be easily explained. 



My preference would be something like
  adj-ii    adjective - ii special class
  v1-ik     ichidan verb - kureru special class
  cop-da    copula da

Is there any need for a cop-desu tag?  If denoting です as a
copula is desirable, why not just have a cop tag to apply to
anything one wishes to note as a copula?  My desire for cop-da
is to single out the だ word so it can be properly conjugated
but that does not apply to です which is already conjugated.
(C.f. すむ[v5m] versus すみません[int] or 御座る[v4r] versus
御座います[exp])



I'll answer to that in another post.

JL