How/Where to use _T() a.k.a. TEXT() macros?

[cpjust]

Hi,
I've read that the _T("...") or TEXT("...") macros basically expand to L"..." when UNICODE is defined and "..." when it isn't.
For literal strings like that, it makes sense, but what about pointers? Is there a way to do the same kind of thing with variables?
If I do this:

char szData[256];
strcpy( szData, "Hello World" );
std::wstring wStr( _T( szData ) );  // C2065: 'LszData' : undeclared identifier

I get an error since it just puts an L infront of the variable name.

 

[xwb]

Use TCHARs. The problem is that there isn't an equivalent for std::string so you'll have to make up your own. Have a look at tchar.h for more info

#include <tchar.h>
#include <string>
#ifdef _UNICODE
#define tstring wstring
#else
#define tstring string
#endif
TCHAR szData[256];
_tcscpy( szData, _T("Hello World") );
std::tstring wStr( _T( szData ) );

 

[cpjust]

#ifdef _UNICODE
#define tstring wstring
#else
#define tstring string
#endif

Yeah, that's basically the only thing I could think of to put a bandage on this problem...
It's a hell of a mess though, since I'm trying to create an interface between two badly designed components; one which uses UNICODE and passes CString objects (across a DLL boundary none the less) and another that is only non-UNICODE and not const-correct.

 

[uolj]

I don't think that solves your problem, though. It sounds like you still have to convert from non-Unicode to Unicode. There are functions to do that, but I don't remember the name(s).

 

[xwb]

wcstombs and mbstowcs

 

[cpjust]

Yup, those are the ones.
I used them in my functions which basically look like this:

Tstring ASCII_ToWide( const char* );
Tstring ASCII_ToWide( const CString& );
Tstring WideToASCII( const wchar_t* );
Tstring WideToASCII( const CString& );