Ideally, SBCS/DBCS mixed strings should be supported wherever character strings can be used in your program. The following breaks usages to give you a general idea on their DBCS support priorities:
Usually SBCS characters are used for numeric data, although each DBCS code page has DBCS numeric characters in it. You do not need to support DBCS for numeric data. However, country-unique formatting of numeric data should be considered.
Usually SBCS characters are used for coded data such as employee code and department code, although there are DBCS alphabetic characters as well as DBCS numeric characters. Therefore, you need not to support DBCS for coded data.
Examples of text data are:
Messages and panels are called PII. PII should be DBCS-enabled and translated to national languages, such as Japanese and Korean.
Fields and elements for text data, product description, for example, should support DBCS.
Examples of this category are:
Examples are:
DBCS input operation, however, requires more keystrokes than SBCS input. Because of this, the user may not utilize DBCS name support. Selecting from a list with DBCS items are preferred instead of inputting DBCS items repeatedly.
Therefore, if you provide some mechanism which minimizes DBCS input, DBCS name support will be fully appreciated, as for example, the case of DBCS variable name support in program languages. If you also provide a syntax sensitive editor for the languages, it might be a good idea to list up the already defined variable names for selection at the appropriate places.
See Programming Language Compiler and Full Screen General Editor for examples of this category.