I'm attempting to read text into a CString using an LPSTREAM, but it doesn't seem to be working correctly, here is the code that I'm calling:
static HRESULT UTL_ReadStreamTxt(MyStorage* pSrcStg, const char* pszStream, CString* myCStr, int size)
{
HRESULT hrRet = STG_E_INVALIDPARAMETER;
LPSTREAM lpSrc = NULL;
ULONG ul;
TRY
{
USES_CONVERSION;
HRESULT hrSrc = pSrcStg->GetStg()->OpenStream(CT2COLE(strSrc),
NULL,
STGM_READ | STGM_SHARE_EXCLUSIVE,
0,
&lpSrc);
if (hrSrc != NOERROR)
{
hrRet = hrSrc;
}
else
{
hrRet = lpSrc->Read(&myCStr, size, NULL); // Read into CString
}
}
CATCH_ALL(e)
{
hrRet = STG_E_UNKNOWN;
}
END_CATCH_ALL
_AfxRelease((LPUNKNOWN*)&lpSrc);
return hrRet;
}
When it reads into the string, Visual Studio says that the data in the CString is corrupted.
The compound storage's stream contents are the following:
abcdefghijklmnopqrstuvwxyzabcdefghijklmnopqrstuvwxyz
I'm not entirely sure I'm using Read() correctly, How do I fix this issue?
The main problem is that you are passing a bad pointer to Read()
. You are passing the memory address of the myCStr
parameter itself, not the address of the CString
being pointed at, or more accurately, the memory address of a character buffer that the CString
owns. The code compiles only because Read()
takes a simple void*
pointer to a buffer, and any pointer is implicitly convertible to void*
.
Also note that CString
is based on TCHAR
, which maps to either char
or wchar_t
depending on whether you are compiling your project for ANSI/MBCS or Unicode. So, reading from the stream directly into a CString
will only work correctly if:
the stream contains ANSI characters and TCHAR
maps to char
.
the stream contains UTF-16 characters and TCHAR
maps to wchar_t
.
If the stream's character type does not match the character type used by CString
, you would have to first read the stream into an intermediate buffer and then convert that to TCHAR
before it can be stored in the CString
.
Try something more like this:
static HRESULT UTL_ReadStreamTxt(MyStorage* pSrcStg, const char* pszStream, CString* myCStr, int size)
{
HRESULT hrRet = STG_E_INVALIDPARAMETER;
LPSTREAM lpSrc = NULL;
ULONG ul;
LPVOID buffer;
TRY
{
USES_CONVERSION;
HRESULT hrSrc = pSrcStg->GetStg()->OpenStream(CT2COLE(strSrc),
NULL,
STGM_READ | STGM_SHARE_EXCLUSIVE,
0,
&lpSrc);
if (hrSrc != S_OK)
{
hrRet = hrSrc;
}
else
{
// if the stream's character type matches TCHAR...
buffer = myCStr->GetBuffer(size / sizeof(TCHAR));
hrRet = lpSrc->Read(buffer, size, &ul);
myCStr->ReleaseBuffer(ul / sizeof(TCHAR));
// else, if the stream's character type is 'char' and TCHAR is 'wchar_t'...
CStringA tmp;
buffer = tmp.GetBuffer(size);
hrRet = lpSrc->Read(buffer, size, &ul);
tmp.ReleaseBuffer(ul);
*myCStr = CString((LPSTR)tmp, tmp.GetLength());
// else, if the stream's character type is 'wchar_t' and TCHAR is 'char'...
CStringW tmp;
buffer = tmp.GetBuffer(size / sizeof(wchar_t));
hrRet = lpSrc->Read(buffer, size, &ul);
tmp.ReleaseBuffer(ul / sizeof(wchar_t));
*myCStr = CString((LPWSTR)tmp, tmp.GetLength());
// alternatively, you can do the above 2 cases more generically...
typedef CStringT<char or wchar_t> CStreamString;
CStreamString tmp;
buffer = tmp.GetBuffer(size / sizeof(CStreamString::XCHAR));
hrRet = lpSrc->Read(buffer, size, &ul);
tmp.ReleaseBuffer(ul / sizeof(CStreamString::XCHAR));
*myCStr = CString((CStreamString::PXSTR)tmp, tmp.GetLength());
}
}
CATCH_ALL(e)
{
hrRet = STG_E_UNKNOWN;
}
END_CATCH_ALL
_AfxRelease((LPUNKNOWN*)&lpSrc);
return hrRet;
}