CJKV Information Processing, 2nd Edition

Code Conversion Algorithms

|

579

However, when dealing with conversion between character sets of dierent locales, such

as between GB 2312-80 and Big Five, the relationship is not always one-to-one, thus

round-trip conversion is not always possible. For this reason, and given the extent to

which Unicode is supported in today’s OSes and applications, it makes sense to keep your

data in Unicode, unless you have a very good reason not to do so.

e code conversion techniques in this section cover Unicode, along with two CJKV-

specic encoding methods: ISO-2022 and EUC. e Japanese-specic Shi-JIS encoding

method is also covered, as is conversion to and from Row-Cell notation. ese techniques

can be easily applied to any CJKV locale as long as their character sets are based on these

encoding methods or notations.

It is best to treat the vendor encoding methods, as described in Appendix F, as exceptional

cases. It is also best to avoid using such encoding methods and character sets if your so-

ware requires the maximum amount of exibility and information interchange—this is a

portability issue.

e following sections contain more detailed information about dealing with the conver-

sion of these and other encoding methods. Two of the conversion algorithms require the

use of somewhat complex functions for maximum eciency (at least, when writing code

in a language other ...