At 19:39 01/08/16 +0100, Simon Cozens wrote:
On Fri, Aug 17, 2001 at 02:57:31AM +0900, SADAHIRO Tomoyuki wrote:
> Considering the particularity of Hangul Syllables,
> would it be better to gather those functions into one module?
I think so, yes; if I was dealing with Hangul data, I don't think
I'd be surprised to have to download a CPAN module to get the naming
algorithms and so on.
But you may be surprised that you need it just to do normalization,
I guess. Maybe one solution is to separate (de)composition and
naming algorithms for Hangul into two separate modules. But I
guess I would prefer to have (de)composition in the normalization
module; it's just two rather simple functions. The functions are
actually so simple that it would not really hurt if there were
copies in both the lingua...hangul module (for hangul linguistic
processing) and the normalization module (for general normalization).
Regards, Martin.
