Unicode wxWidgets in a Non-Unicode World

wxWidgets can be compiled to pure Unicode mode by setting wxUSE_UNICODE to 1 in the header file $(WX_WIDGETS_ROOT)\include\wx\msw\setup.h

But what if the rest of your code that interfaces with Unicode wxWidgets is still not Unicode, but uses char and STL std::string? The changes needed at the interface are:

• Wrap all characters and string literals passed to wxWidgets in wxT(). It converts them to Unicode. Ex: wxT("Hell World!")

• wxWidget functions do not deal with STL std::string but with wxString. Do the conversion using wxConvUTF8, which is an instance of the wxMBConv class. MBConv stands for Multi-Byte Conversion, it helps to convert between Multi-Byte formats and Unicode. wxConvUTF8 as you can guess converts multi-byte to the UTF-8 Unicode encoding. Ex:

  std::string stdStr(“Hell World!”); wxString wxStr(stdStr.c_str(), wxConvUTF8);

• When wxWidgets functions throw out a wxString, convert them back to char string using wxString::mb_str(). This returns the multi-byte representation of the internal Unicode text in wxString.

If you are curious about this fishy Multi-Byte business, check out Windows: Multi-Byte.