CJKV Information Processing, 2nd Edition

Input Techniques

|

309

the hexadecimal ISO-2022-JP value 5178 or the Unicode scalar value 528 for the

Japanese kanji  (ken or tsurugi), but imagine doing this for thousands or tens of thou-

sands of ideographs. While input by reading may yield more than one candidate character

from which to choose, it is—in a seeming paradox—the most productive and most widely

used method yet invented. Figure 5-1 illustrates the four possible stages of input: it shows

how the ow of input information travels and how dierent input methods and keyboards

interface at each stage.





   



















Input stages and input method interaction—JapaneseFigure 5-1.

The Input Method

CJKV input soware is usually referred to as an input method (or FEP, which is an abbre-

viation for front-end processor). e input method is aptly named because it captures the

keyboard input, processes it to become the intended text, and then sends it to the applica-

tion or other soware. Typically, the input method runs as a separate process in its own

input window, largely independent of the underlying application. Now, thanks to OS-level

APIs, the use of input methods is much more seamless than in the past.

e input methods that were available at the time when the rst edition of this book

was published