How does one convert from CString to unsigned char* with Unicode?

Question

Hello everyone, I'm trying to do something simple here. When I execute the following code in Visual Studio 2008 using the unicode character set, xmlString is correct.

Unfortunately I need to convert the CString to a unsigned char*. Using the code below, ucStr becomes "<" (i.e. the first character of xmlString).

How should I convert the CString to an unsigned char* and retain all the information?

     CString xmlString;
  xmlString.Format( _T("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?><gateway><config-read><%s /></config-read></gateway>"), keyName);

  unsigned char * ucStr = reinterpret_cast<unsigned char *> (xmlString.GetBuffer());
  pgIRequest->SendXmlData( "dgv/gateway.xml", ucStr, xmlString.GetLength() + 1) ;

Answer 1

I assume "SendXmlData" wants the number of bytes and not the number of characters.

If so, you want to change "GetLength() + 1" to "(GetLength() + 1)*sizeof(xmlString[0])".

Answer 2

I think you want wcstombs, or, rather, its more secure counterpart wcstombs_s.

Answer 3

I find the simplest is to use the CStringA constructor, like this:

 CString xmlString;
 xmlString.Format( _T("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?><gateway><config-read><%s /></config-read></gateway>"), "test");

 printf("%s\n",xmlString );  // fails "<"

 //unsigned char * ucStr = reinterpret_cast<unsigned char *> (xmlString.GetBuffer());

 CStringA ucStr( xmlString );

 printf("%s\n",ucStr );   // works!

Answer 4

Here's the code that finally worked:

CString xmlString;
  xmlString.Format( _T("<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?><gateway><config-read><%s /></config-read></gateway>"), keyName);
  CStringA ucStr( xmlString );
  unsigned char * ucStr2 = reinterpret_cast<unsigned char *> (ucStr.GetBuffer());
  pgIRequest->SendXmlData( "dgv/gateway.xml", ucStr2, xmlString.GetLength() + 1) ;// target on gateway to download

Answer 5

This codes snippet may not work when your 'keyName' variable starts to contain characters that are not representable in the ISO-8859-1 encoding. For this I recommend creating a string with UTF-8 as encoding value, convert to UTF-8 byte stream using WideCharToMultiByte using the CP_UTF8 codepage and send out resulting utf8 byte stream.