[Uim] Determine the semantic of reset

TOKUNAGA Hiroyuki tkng at xem.jp
Tue Jun 7 11:17:47 EEST 2005


On Mon, 06 Jun 2005 15:34:39 +0800
James Su <suzhe at tsinghua.org.cn> wrote:

> Hi,
>   The commit event should be emitted by input method itself. In
> another word, whether a preedit string is committable should be
> decided by input method rather than frontend (client app) or user. So

I agree half of your opinion, but I think if it's configurable, that
would be better.

> I don't think it's necessary to export a "committable" property from
> input method side to client application. If you really want such

Applications don't handle such property of course. That will handle
within input method library layer. API for applications should be very
simple.


> property, it should be a private property within individual input
> method. So having a public API for it should be useless and not
> practical. And if you have a look at the api of SCIM, you may find
> that scim provides a separated string element called "aux_string"
> besides preedit string, which is mainly used for displaying hintings
> and messages. Preedit string is always only for composing.

Simple aux_string is not enough for part of Japanese input method. 'New
word learing-mode' in uim-skk needs mixing of committable strings and
none-committable strings.

Regards,

-- 
TOKUNAGA Hiroyuki
tkng at xem.jp



More information about the uim mailing list