CJKV Information Processing, 2nd Edition

Byte Versus Character Handling

|

599

Table 9-12 illustrates what happens with the same character string, but when EUC-JP–

encoded.

Character deletion example—EUC-JPTable 9-12.

Text representation EUC-JP representation

Original string



   

4 1      1

Correct

Delete



  

4 1    

Add character



   

4 1     5 5

Incorrect

Delete



  

4 1     

Add character



   

4 1      C5 B5

is problem is xed by keeping track of the characters at the insertion point—whether

they are represented by one or more bytes. If a byte happens to be the second byte of a

two-byte character, both bytes must be deleted with a single keystroke. In the case of

three-byte characters (for example, characters from EUC-JP code set 3—JIS X 0212-1990

characters), three bytes must be deleted. Extreme examples include EUC-TW and GB

18030 encodings, both of which have a four-byte representation. at’s a lot of bytes,

meaning that a lot can go wrong if you’re not careful. Also, anything outside of the BMP,

when dealing with the UTF-8 encoding form, is four bytes.

Character Insertion

Inserting characters is problematic only when the insertion point—that is, the cursor—is

between the two bytes that represent a two-byte character. is then splits the two-byte

character and results in data loss. is section, as you may have expected, relates ...