|
Asking people to do your homework is not a long term method of succeeding in the computer industry. How about you post your code and I'm sure we'll all be happy to give advice on where you are stuck.
Christian
The content of this post is not necessarily the opinion of my yadda yadda yadda.
To understand recursion, we must first understand recursion.
|
|
|
|
|
Thanks to those who helped me previously, my WTL project is progressing nicely (when I find time )....
My latest problem is very odd. I have tried using a number of functions that take an ATL CString, such as GetDlgItemText, GetWindowText ( no prize for guessing what I want to do ) and also GetItemText in my CListCtrl. In EVERY case, I get a message like this:
c:\my programs\wtl\mp3cat\inputdlg.h(52) : error C2664: 'int __thiscall ATL::CWindow::GetDlgItemTextA(int,unsigned short *& ) const' : cannot convert parameter 2 from 'class WTL::CString *' to 'unsigned short *& '
A reference that is not to 'const' cannot be bound to a non-lvalue
I don't see why it's asking me for a pointer to an unsigned short, but it appears an almost universal problem. I've found ways around it in other places ( by doing things differently ) but in this case I have a dialog which has the sole purpose of getting input from a CEdit. I tried fiddling with the DDX sample, but to no avail, and while I'm guessing a bit more time with that might do it, I *hate* DDX anyhow and *never* use it in MFC, so I'd much prefer to get to the bottom of this problem.
FWIW I have the version of ATL in the last SDK ( October from memory )
Thanks to one and all 
Christian
The content of this post is not necessarily the opinion of my yadda yadda yadda.
To understand recursion, we must first understand recursion.
|
|
|
|
|
Sounds like something's screwy with your ANSI/Unicode settings. GetDlgItemTextA() is expecting an unsigned short* (ie Unicode string) which isn't right. 
--Mike--
http://home.inreach.com/mdunn/
The preferred snack of 4 out of 5 Lounge readers.
|
|
|
|
|
*sigh* Thanks - now I know what the problem is. Could it be I need to reinstall WTL ? I haven't #define'd UNICODE anywhere, and in fact doing so breaks a lot of other stuff. I also have tried passing in char*'s, etc., but I keep getting the same messages... I can get this particular code to work by doing this:
char * pChar = new char[255];
::GetDlgItemText(m_hWnd, IDC_EDIT, pChar, 255);
string = pChar;
but I'm still left with the feelings something is fundamentally broken 
Christian
The content of this post is not necessarily the opinion of my yadda yadda yadda.
To understand recursion, we must first understand recursion.
|
|
|
|
|
Wait, y'know what, I was wrong. You are calling the version of GetDlgItemText() that takes a BSTR&. That's where the unsigned short*& mentioned in the error message is coming from -- BSTR is typedef'ed as unsigned short*
You need to call the other version, which takes an LPTSTR. Note that CString doesn't have an LPTSTR operator - you need to call GetBuffer() to get a non-const pointer.
--Mike--
http://home.inreach.com/mdunn/
The preferred snack of 4 out of 5 Lounge readers.
|
|
|
|
|
I've noticed that BSTR keeps coming up in my list of available inputs and I've been trying to use GetBuffer to get to the LPTSTR version. Why would I not have access to the CString versions ? It appears to be right across the board...
Christian
The content of this post is not necessarily the opinion of my yadda yadda yadda.
To understand recursion, we must first understand recursion.
|
|
|
|
|
Calling GetBuffer() returns an LPTSTR. Is this not working for you? The code would be something along these lines...
CString s;
wndSomeWindow.GetDlgItemText ( IDC_SOME_CONTROL, s.GetBuffer(64), 64 );
s.ReleaseBuffer();
--Mike--
http://home.inreach.com/mdunn/
The preferred snack of 4 out of 5 Lounge readers.
|
|
|
|
|
Oh, um... *embarrassed grin*
I thought ReleaseBuffer released the ownership of the pointer, i.e. if I passed the value to a char* or a list on a permanent basis, I would release it from the string. As my wife has been heard to tell me, I've been doing it wrong !!!
So I release after every GetBuffer to restore ownership to the string, is that right ?
Christian
The content of this post is not necessarily the opinion of my yadda yadda yadda.
To understand recursion, we must first understand recursion.
|
|
|
|
|
Yup, this also goes for MFC CString
MSDN says:
"Use ReleaseBuffer to end use of a buffer allocated by GetBuffer."
- Anders
|
|
|
|
|
I think you get it now. When you call GetBuffer(), you get access to the char buffer and you can modify it however you want (as long as you don't overrun the end of the buffer). Calling ReleaseBuffer() means you no longer need this direct access, and the CString object once again becomes responsible for managing the memory associated with the buffer.
--Mike--
http://home.inreach.com/mdunn/
The preferred snack of 4 out of 5 Lounge readers.
|
|
|
|
|
Thanks for the help - all this WTL stuff, combined with trying to read/write ID3 tags, I got a little sidetracked and didn't bother checking out some of the basics. You know what they say - when you make an assumption, you make an ass out of u and umption. ( yeah, it doesn't work, but it's from one of my favourite movies, so I'll quote it anyhow )
Thanks again
Christian
The content of this post is not necessarily the opinion of my yadda yadda yadda.
To understand recursion, we must first understand recursion.
|
|
|
|
|
Hey all. I'm using GlobalAlloc to allocate memory for a structure. One member of that structure is a 'char *'. I'm using 'malloc' to allocate memory for the string. My question is does GlobalFree free my string, or just the HGLOBAL returned by GlobalAlloc? Or put another way, do I have to call free(m_MyString) before calling GlobalFree?
Thanks in advance for any help.
Jamie Nordmeyer
Portland, Oregon, USA
|
|
|
|
|
Well, yes - you should call free() - GlobalFree() won't know about your char*.
Interesting stuff. malloc on 32 bit systems uses HeapAlloc, I think - the GlobalAlloc / LocalAlloc stuff is 16 bit. The new functions (HeapAlloc et al) can do more in terms of actually creating/managing heaps.
How have you come to use GlobalAlloc?
|
|
|
|
|
Hi Tim. Thanks for the reply. I'm trying to write my own window class. Why? One, so I don't have to link to MFC, and two, so I can study how Windows works from the ground level (using API calls and such). I'm using GlobalAlloc to create a pointer to my window classes 'this' pointer for use in its WndProc. I'd never really tried anything like this before, so when I first noticed GlobalAlloc, it looked like what I needed.
I'll use HeapAlloc instead, since it's 32-bit, and go from there! 
Jamie Nordmeyer
Portland, Oregon, USA
|
|
|
|
|
Visual C++ 6.0, sp5
NT4, sp5
(I also posted this on the ADO board)
I'm working on an app that uses ADO to access an oracle database. The following code is part of the function that opens a recordset (try/catch stuff removed for the sake of brevity):
Fields* m_pFields = NULL;
SAFE_CALL(m_piRecordSet->Open(vSQL, vConnection, adOpenDynamic, adLockOptimistic, adCmdText));
int m_nRecordCount = m_piRecordSet->GetRecordCount();
SAFE_CALL(m_piRecordSet->get_Fields(&m_pFields));
...and this is a code snippet from the function we use to close the recordset:
if (m_pFields)
{
m_pFields->Release();
}
SAFE_CALL(m_piRecordSet->Close());
m_piRecordSet.Release();
m_piRecordSet = NULL;
The problem is with the m_pFields variable.
The first 10 or so times we call the function containing this code, it's a valid pointer. On or about the nth time, the variable is NULL. Actually "not initialized" would be a better description of it's state, because if we don't set it to NULL before calling the get_Fields() function, it's value is 0xcccccccc.
As far as I can tell, we close the recordset when we're done with it. Otherwise, we'd get an error message from Oracle saying we have too many open cursors.
Has anyone got any idea why this might be happening?
|
|
|
|
|
Hmmm, just a shot in the wild......
I had similar strange problems with Win2k, when I used the connection from a service app.
I removed the m_piRecordSet.Release(); because the = operator actually calls m_piRecordSet.Release();.
It worked for me 
A bit strange though...
- Anders
|
|
|
|
|
Hi, 
One of my programs uses keyboard hook and shell hook. However, under win9x/ME it does not work as expected.
What's the point:
When Winscheduler is active and it is minimised on try bar...
If you minimize another program (e.g. Word, Acad 14, Excel ...), the icon of minimised program disappear from task bar. If you use ALT+TAB you can steel switch between programs.
Moreover, when you open a new window, it is not seen on the taskbar.
Here is the source code for the dll file:
<br />
<br />
#include "stdafx.h"<br />
#include "SystemHook.h"<br />
#include <stdlib.h><br />
<br />
#define ID_CMD_KEYPRESSED WM_USER+200<br />
#define ID_CMD_WNDCREATED WM_USER+201<br />
<br />
#pragma data_seg("SHARDATA")<br />
static HWND hwndMain = NULL;
#pragma data_seg()<br />
<br />
HWND ghWndMain;
HHOOK hHookKey;
HHOOK hHookShell;
<br />
static BOOL bHookInstalled;
<br />
HINSTANCE ghDLLInst;
<br />
<br />
BOOL APIENTRY DllMain( HANDLE hModule, <br />
DWORD ul_reason_for_call, <br />
LPVOID lpReserved)<br />
{<br />
switch (ul_reason_for_call)<br />
{<br />
case DLL_PROCESS_ATTACH:<br />
case DLL_THREAD_ATTACH:<br />
case DLL_THREAD_DETACH:<br />
case DLL_PROCESS_DETACH:<br />
break;<br />
}<br />
ghDLLInst=(HINSTANCE)hModule;<br />
return TRUE;<br />
}<br />
<br />
CSystemHook::CSystemHook()<br />
{ <br />
return; <br />
}<br />
<br />
SYSTEMHOOK_API LRESULT CALLBACK KeyboardProc(int nCode,WPARAM wParam,LPARAM lParam)<br />
{<br />
if (nCode >= 0) <br />
{<br />
if(nCode == HC_NOREMOVE)<br />
return CallNextHookEx(hHookKey, nCode, wParam, lParam);<br />
<br />
if (!(lParam & 0x80000000 || lParam & 0x40000000))<br />
{<br />
{<br />
static HWND h=::FindWindow(NULL,"WinScheduler");<br />
PostMessage(h , ID_CMD_KEYPRESSED , wParam, lParam);<br />
}<br />
}<br />
}<br />
return CallNextHookEx(hHookKey, nCode, wParam, lParam); <br />
}<br />
<br />
SYSTEMHOOK_API LRESULT CALLBACK ShellProc(int nCode,WPARAM wParam,LPARAM lParam)<br />
{<br />
return CallNextHookEx(hHookShell, nCode, wParam, lParam); <br />
}<br />
<br />
SYSTEMHOOK_API int WINAPI InstallHook (HWND hWnd, BOOL bCode )<br />
{<br />
int nReturn = 1;<br />
ghWndMain = hWnd;
<br />
if (bCode == bHookInstalled) <br />
return 0; <br />
<br />
if (bCode)<br />
{<br />
hHookKey=(HHOOK)SetWindowsHookEx(WH_KEYBOARD,(HOOKPROC)KeyboardProc,ghDLLInst,0);<br />
hHookShell=(HHOOK)SetWindowsHookEx(WH_SHELL,(HOOKPROC)ShellProc,ghDLLInst,0);<br />
if (!hHookKey)<br />
return 0; <br />
if (!hHookShell)<br />
return 0;<br />
bHookInstalled = TRUE; <br />
}<br />
else <br />
{ <br />
nReturn = UnhookWindowsHookEx(hHookKey);<br />
nReturn &= UnhookWindowsHookEx(hHookShell);<br />
bHookInstalled = FALSE; <br />
}<br />
return nReturn; <br />
}<br />
As it is seen, even if I am not doing anything in the ShellProc, the program does not work correctly. If i do not install the WH_SHELL hook, everything works fine.
Please help me. What may be the reason for this stupid behavior?
Best regards
Mustafa Demirhan
|
|
|
|
|
I'm not sure I understand what your code is supposed to do but I can spot something wrong :
hHookKey, hHookShell, bHookInstalled should be inside your SHARDATA section. They should be shared between processes.
I think ghWndMain should be inside as well. (hwndMain doesn't seem to be used ?)
#pragma data_seg("SHARDATA")
HHOOK hHookKey = NULL;
HHOOK hHookShell = NULL;
BOOL bHookInstalled = FALSE;
HWND ghWndMain = NULL;
#pragma data_seg()
-------------------------
Benoit Devigne
http://www.bdevigne.com
contact@bdevigne.com
|
|
|
|
|
I'm using MFC programming to implement a software. Of course there are "View, Doc, mainframe...etc" file.....
I've added a dialog box with a textbox for me to input some values. And I have separate file called ABC.cpp to do other things. In the ABC.cpp, I would like to pop up that dialog box to let user to input the value and then use that value in ABC.cpp. However, i got the following error message:
C:\XYZ\InputNum.h(21) : error C2065: 'IDD_INPUTNUM' : undeclared identifier
C:\XYZ\InputNum.h(21) : error C2057: expected constant expression
I did everything already:
CInputNum InputNum;
InputNum.DoModal();
num = InputNum.m_inputnumber;
Do you know why? and how to fix it? I don't want to make a global variable......
|
|
|
|
|
Looks like you are including inputnum.h without including resource.h for that translation unit. Its nicer if you can keep #includes out of header files, but including resource.h inside inputnum.h would probably solve the problem.
|
|
|
|
|
Wow........you are superb!!!
Thanks....
It works!!!!
|
|
|
|
|
Hi everybody,
I am developping a program named WinScheduler. I am using Win2000 and the program runs well under Win2000. However, there are some problems in Win9x. The program normally uses WH_KEYBOARD and WH_SHELL hooks. WH_SHELL is used to catch the newly opened top level windows.
However, under win9x it does not work as expected.
What's the point:
When Winscheduler is active and it is minimised on try bar...
If you minimize another program (e.g. Word, Acad 14, Excel ...), the icon of minimised program disappear from task bar. If you use ALT+TAB you can steel switch between programs.
Moreover, when you open a new window, it is not seen on the taskbar.
Please help me. What may be the reason for this stupid behavior? Everything works fine on Win2000 but nothing works on Win9x.
Best regards
Mustafa Demirhan
|
|
|
|
|
Hello, the codegurus around the world. 
Life seems not to be easy among Windows OS. 
At our company, we have had the problems among Windows OS, especially Windows Me and NT 4.0.
In that case, we had written the different code for each Windows OS.
So, before we write the code, we always check that this API supports by all or not.
In fact, I can't give you the answer, but Window 2000 gives us the useful API function.
But, it may not work for Windows 9x.....
Have a nice day!
-Masaaki Onishi-
|
|
|
|
|
This may be old news to everyone else, but I can't seem to compile anything with an STL header such as <vector> included without warnings. I've been asked to develop a project to compile with warnings set at level 4, but even a bare source file with a single line...
#include <vector>
...gives rise to 6 warnings.
Is there some way around this? Can I get better source code for the STL under Visual C++?
|
|
|
|
|
one line will fix it. put it before your STL includes.
#pragma warning(disable:4786)
-c
------------------------------
Smaller Animals Software, Inc.
http://www.smalleranimals.com
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
