Unicode strings require the use of WIDE characters defined using 16 bits rather than the usual 8 bits. That is what "wchar_t " defines in your C++ code.
So each character is coded using two bytes rather than the usual one. You just need to use the MultiByteToWideChar API function to translate from an ANSI string to the UNICODE representation of the same string. What I think you end up with is the original set of bytes interleaved with "00" and terminated with a double "00"

On current ifort on Windows use

USE IFWINTY

INTEGER(WCHAR) :: wide_character_string(number_of_characters)

Quote:

michaelgreen wrote:

Thanks for the reply.

Unfortunately I can't get the program to compile when using MultiByteToWideChar because its 5th argument must be a wide character string, and that's just what I don't seem to have access to in Fortran. The question therefore boils down to: how does one declare a wide character string in Fortran?

Many thanks
Mike

If your character string MYSTRING is, say, 20 characters long, then define an INTEGER(2) buffer to receive the wide string and make it more than 2*(20+1) bytes long:

CHARACTER*20 MYSTRING
INTEGER(2) MYWIDESTRING(50)

Then use the location of MYWIDESTRING as the buffer to receive the wide character string. Here is an example program that compiles and works to produce an array of bytes containing the wide-character version of an input character string.

!****************************************************************************
!
! PROGRAM: multibytecharacters
!
! PURPOSE: Convert a character string to an array of wide characters
! represented as 2-byte integers.
!
!****************************************************************************

program multibytecharacters
use dfwin
implicit none

character*40 mystring
integer(2), allocatable :: mywidestring(:)
integer iret, isize, i

mystring="this is my null-terminated string"//char(0)
! Find out how big the receiving buffer should be, in wide characters...
! by setting the buffer length isize initially to zero
isize=0
iret=MultiByteToWideChar(CP_ACP,MB_PRECOMPOSED, mystring, len_trim(mystring),mywidestring,isize )
! iret contains the required buffer length in 2-byte characters, so allocate the required number
allocate(mywidestring(iret))
print *,'required buffer size,iret =',iret
if(size(mywidestring).ge.iret) then
! Do the conversion...
isize=iret
iret=MultiByteToWideChar(CP_ACP,MB_PRECOMPOSED, mystring, len_trim(mystring),mywidestring,isize )

print *,'Number of multibyte characters written,isize=',isize
print *, mystring(1:len_trim(mystring))
print *, (mywidestring(i),i=1,isize)
else
print *,'Buffer length required = ', iret, ', buffer size =',size(mywidestring)
endif
pause
deallocate(mywidestring)
end program multibytecharacters