Unicode and Multibyte Character Set (MBCS) Support
Updated: July 2009
Some languages, for example, Japanese and Chinese, have large character sets. To support programming for these markets, the Microsoft Foundation Class Library (MFC) is enabled for two different approaches to handling large character sets:
The entire class library is conditionally enabled for Unicode characters and strings. In particular, class CString is Unicode-enabled.
(xx represents the version number of the file; for example, '80' means version 8.0.)
CString is based on the TCHAR data type. If the symbol _UNICODE is defined for a build of your program, TCHAR is defined as type wchar_t, a 16-bit character encoding type. Otherwise, TCHAR is defined as char, the normal 8-bit character encoding. Therefore, under Unicode, a CString is composed of 16-bit characters. Without Unicode, it is composed of characters of type char.
To complete Unicode programming of your application, you must also:
Use the _T macro to conditionally code literal strings to be portable to Unicode.
When you pass strings, pay attention to whether function arguments require a length in characters or a length in bytes. The difference is important if you are using Unicode strings.
Use portable versions of the C run-time string-handling functions.
Use the following data types for characters and character pointers:
TCHAR Where you would use char.
LPTSTR Where you would use char*.
LPCTSTR Where you would use const char*. CString provides the operator LPCTSTR to convert between CString and LPCTSTR.
CString also supplies Unicode-aware constructors, assignment operators, and comparison operators.
For related information about Unicode programming, see Unicode and MBCS and Unicode in MFC. The Run-Time Library Reference defines portable versions of all its string-handling functions. See the category Internationalization.
The class library is also enabled for multibyte character sets, specifically for double-byte character sets (DBCS).
Under this scheme, a character can be one or two bytes wide. If it is two bytes wide, its first byte is a special "lead byte" that is chosen from a particular range, depending on which code page is in use. Taken together, the lead and "trail bytes" specify a unique character encoding.
If the symbol _MBCS is defined for a build of your program, type TCHAR, on which CString is based, maps to char. It is up to you to determine which bytes in a CString are lead bytes and which are trail bytes. The C run-time library supplies functions to help you determine this. For more information, see General MBCS Programming Advice.
Under DBCS, a given string can contain all single-byte ANSI characters, all double-byte characters, or a combination of the two. These possibilities require special care in parsing strings. This includes CString objects.
Unicode string serialization in MFC can read both Unicode and MBCS strings regardless of which version of the application that you are running. Your data files are portable between Unicode and MBCS versions of your program.
CString member functions use special "generic text" versions of the C run-time functions they call, or they use Unicode-aware functions. Therefore, for example, if a CString function would typically call strcmp, it calls the corresponding generic-text function _tcscmp instead. Depending on how the symbols _MBCS and _UNICODE are defined, _tcscmp maps as follows:
Neither symbol defined
The symbols _MBCS and _UNICODE are mutually exclusive.
Similarly, CString methods are implemented by using "generic" data type mappings. To enable both MBCS and Unicode, MFC uses TCHAR for char, LPTSTR for char*, and LPCTSTR for const char*. These ensure the correct mappings for either MBCS or Unicode.