Re: Entering Unicode characters

[handa <handa@gnu.org>]

In article <831t90ztgm.fsf@gnu.org>, Eli Zaretskii <eliz@gnu.org> writes:

> Perhaps Handa-san could comment on the reasons for this design in
> 'quail-get-translations': why does it show the list of possible
> translations only sometimes?  Is it just an optimization?

The deference is that an input method is determinitic or
non-deterministic.

Most input methods are deterministic, i.e. the defined key sequence
reaches to a target character (or a string), but Chinese input methods
are not deterministic, i.e. the defined key sequence reaches only to a
group of candidate characters.  A user has to select one of them by arrow
keys or by an index number.  So, for such an input method, showing
candidate characters is mandatory.

And, chinese input methods do not show possible translations as this:

> a["'/^_`~] 1.ä 2.á 3.å 4.â 5.ª 6.à 7.ã

The above means that when you type ", you get ä, when you type ', you
get á, ...

But chinese input methods show candidates as this (when you type "ni"
with chinese-py):

ni[aenu]  (01/??) 1.你 2.呢 3.尼 4.泥 5.逆 6.倪 7.匿 8.拟 9.腻 0.妮

This means that when you type 1, you get 你, when you type 2, you get
呢, ..., and when you type a (or e, n, u), the different candidates are
shown.

---
K. Handa
handa@gnu.org