HOWTO: Pass a Double-Byte Character to a Win32 API Function (244046)

MORE INFORMATION

Some functions in the Win32 API are character-oriented. The interface of these functions accepts one- or two-character code parameters through the UINT data type. However, a DBCS character is defined by two bytes, a lead byte and trail byte, which form a 16-bit character code. The combination of a lead byte and a trail byte to form a 16-bit character code is called a character pair.

Character pairs can be passed to character-oriented functions by packing the lead and trail bytes into the UINT parameter of the function. To pack a DBCS character code into the UINT parameter, place the trail byte in the low-order byte of the UINT and place the lead byte into the next highest order byte of the UINT.

The following PackDBCS sample function demonstrates this technique:

UINT    PackDBCS(BYTE Lead, BYTE Trail)
{
    UINT nChar = 0;

    nChar = Lead;           // Lead byte first.
    nChar = nChar << 8;     // Shift to next highest order byte.
    nChar = nChar | Trail;  // Then trail byte.

    return nChar;
}
				

The PackDBCS function can be used in situations similar to that shown in the following code sample, where the pStr variable is assumed to be properly incremented along a DBCS string by using a function similar to CharNext. The CharNext function is a DBCS processing function that is provided by the operating system and documented in the Platform SDK.

Sample Code

    UINT    Char;

    if (IsDBCSLeadByte(*pStr))
    {
        Char = PackDBCS(pStr[0], pStr[1]);
    }
    else
    {
        Char = pStr[0];
    }
    
    GetCharWidth32(hDC, Char, Char, pWidth);
				

The following partial list of functions are examples where this packing technique can be used to specify a DBCS character code.

• GetCharABCWidths