Question

For example, MessageBox function has LPCTSTR typed argument for text and caption, which is a pointer to char or pointer to wchar when _UNICODE or _MBCS are defined, respectively.

How does the MessageBox function interpret those stings? As which encoding?

Only explanation I managed to find is this:

http://msdn.microsoft.com/en-us/library/cwe8bzh0(VS.90).aspx

But it doesn’t say anything about encoding? Just that in case of _MBCS one character takes up one wchar (which is 16-bit on Windows) and that in case of _UNICODE one or two char’s (8-bit).

So are those some Microsoft’s versions of UTF-8 and UTF-16 that ignore anything that has to be encoded in 3 or four bytes in case of UTF-8 and anything that has to be encoded in 4 bytes in case of UTF-16? And is there a way to show anything outside of basic multilingual plane of Unicode with MessageBox?

Answer 1

There are normally two different implementations of each function:

• MessageBoxA, which accepts ANSI strings
• MessageBoxW, which accepts Unicode strings

Here, ‘ANSI’ means the multi-byte code page currently assigned to the process. This varies according to the user’s preferences and locale setting, although Win32 API functions such as WideCharToMultiByte can be counted on to do the right conversion, and the GetACP function will tell you the code page in use. MSDN explains the ANSI code page and how it interacts with Unicode.

‘Unicode’ generally means UCS-2; that is, support for characters above 0xFFFF isn’t consistent. I haven’t tried this, but UI functions such as MessageBox in recent versions (> Windows 2000) should support characters outside the BMP.