CJKV Information Processing, 2nd Edition

598

|

Chapter 9: Information Processing Techniques

characters are handled properly. For example, the standard (that is, unlocalized) version

of Microso Word (for Mac OS) is one of the most popular word-processing applications

ever, but at one time failed to handle two-byte characters properly.

Character Deletion

It is quite likely that you will encounter text-processing soware that deletes only one byte

of a two-byte character. ose that have been properly adapted to CJKV locales are able

to detect whether the character in front of the insertion point is represented by two bytes,

and subsequently deletes both bytes. is problem can be avoided if you remember to

press the delete key twice when dealing with two-byte characters. If you are not careful,

loss or corruption of data may result.

Let’s take a closer look at this problem. Table 9-11 provides a sample Shi-JIS–encoded

Japanese text string. e rst process that will be applied is the deletion of the last char-

acter. e rst example deletes the last character (consisting of two bytes), and the second

deletes the last byte (more precisely, the last byte of the last character). Finally, we add

another character, , at the insertion point. Note how the undeleted rst byte le over

from the second example aects the interpretation of the added character (the encoded

value of this added character is highlighted).

Character deletio ...