Unicode CString conversion to int

[ajbharani]

Hello, if I try to convert a CString object to int in Multibyte character set mode, I face no problem.

CString strNumber = _T("1234");
int nNumber = atoi(strNumber.GetBuffer(strNumber.GetLength()));
cout << nNumber <<endl;

But if I go for unicode character set, I get the following error.

error C2664: 'atoi' : cannot convert parameter 1 from 'unsigned short *' to 'const char *'

How can I succeed with the conversion? Thanks.

[Alex F]

Every string API has TCHAR version which compiles both for Multibyte and Unicode. You can find this in MSDN topic for the function. For atoi, this is _tstoi:

int nNumber = _tstoi(strNumber.GetBuffer(strNumber.GetLength()));

[ajbharani]

Thanks. Let me look into it.

[Shadowrun]

Hi! Im newbie in VisualC++ and this if my fisrt message...

What's that "_t" we have to add? Is it only in VS 2008 ??


I've been googling around but it doesn't manage "_t" properly.

[_Superman_]

That is a macro that get expanded to different things depending on whether UNICODE is defined.

Take the example of atoi in this example.
atoi is for the non-unicode version and _wtoi is for the unicode version.

As Alex F mentioned, if you use _tstoi it would get expanded to either atoi or _wtoi depending on whether the build you are taking is a unicode build or not.

The _T("string") macro is used for strings which gets expanded to L"string" for unicode and simply "string" for non-unicode builds.

[Shadowrun]

Thanks superman.

When I read a .txt file (made by NOtepad) with CFile class and then print on screen what I find on file, there are usually bad characters at the end. It doesn't matter if I read 5, 10 or 20 bytes.

        CFile f;
        f.Open(m_file, CFile::modeRead);
	CHAR s[80];
	int bytesread=f.Read(s,10);
	f.Close();
	m_out+=s;                 //m_out is "Cstring" linked to an EDIT CONTROL box)
	UpdateData(FALSE);

(Im making an MFC app).

Should I declare "s" as unicode ? How is it ?

Or should I avoid CFile commands?

[_Superman_]

Declare as
WCHAR s[80];

If you want it to work for both unicode and non-unicode without a code change, declare as
TCHAR s[80];

[krmed]

You're seeing "bad" characters because your character array is not NULL terminated. Just add the following to your code and you should be fine:

        CFile f;
        f.Open(m_file, CFile::modeRead);
	CHAR s[80];
	int bytesread=f.Read(s,10);
	f.Close();
	s[10] = 0;    // NULL terminator - (the 10 is just the number of bytes you read).
	m_out+=s;                 //m_out is "Cstring" linked to an EDIT CONTROL box)
	UpdateData(FALSE);

Hope that helps.