[Uim] Replacing some preedit relevant codes with 'ustr'

[YamaKen]

At Wed, 20 Oct 2004 06:13:24 +0900,
tkng at xem.jp wrote:
> 
> On Mon, 18 Oct 2004 14:59:54 +0900
> YamaKen <yamaken at bp.iij4u.or.jp> wrote:
> 
> > I'm going to replace some complicated preedit relevant codes
> > with 'ustr' I wrote recently. Although I believe that there is
> > no problem, stop me immediately if you feel some problem with
> > its design or the meaning of the name.

> > Since ustr is not dedicated to implement preedit, I named it
> > generic one rather than 'preedit'.
> 
> Please let me know the other uses of ustr. I do not conceive of any uses
> other than preedit.

For example, its cursor operations are useful for parsing
key-str such as "<Control><Alt>delete".

In addition, even if ustr is used to implement preedit, it is
not always represented by single ustr. Suppose following cases.

- Ordinary Japanese multi-segment IMs such as Anthy or Canna may
  use an ustr to implement string of segments. It will ease
  implementing some complex operations. Each segments will also
  ustr in this case

- If we have to implement bidi text someday, ustr can be used to
  represent string of (nested) strings

- uim-skk has the nested editing areas for recursive learning
  which forms a single preedit

In those cases, ustr is only an element of preedit. So I didn't
name itself 'preedit'.

> I'm OK about rewriting current implemantation with this ustr based
> implementation. That will save our maintenance cost.

Sure. The simplification is important to prepare future
extensions.

-------------------------------
YamaKen  yamaken at bp.iij4u.or.jp