|
|
Hello,
I develop add-ons for MS Flight Simulator and I use the poorly documented SDK for this. The SDK provides a .h file in which an interface class is defined (with pure virtual methods only), something like this:
class IPanelCCallback {
public:
virtual IPanelCCallback* QueryInterface (PCSTRINGZ pszInterface) = 0;
virtual bool ConvertStringToProperty (PCSTRINGZ keyword, SINT32* pID) = 0;
virtual bool ConvertPropertyToString (SINT32 id, PPCSTRINGZ pKeyword) = 0;
};
In my code, I use this interface like this:
IPanelCCallback* pCallBack = panel_get_registered_c_callback("fs9gps");
...
SINT32 id;
pCallBack->ConvertStringToProperty(propertyName, &id);
Everything works fine, but I don't understand why... I thought the linker would stop with an "undefined symbol" error because the IPanelCCallback methods, such as ConvertStringToProperty, are declared as pure virtual but defined nowhere, and I don't use any library for linking. With such an interface class, I thought I would have to defined a subclass of IPanelCCallback and define the ConvertStringToProperty method. My code works, so I shouldn't complain, but I would like to know more about the use of interface classes, this is a subject I don't know much.
Any information is greatly welcome !!
Thank you,
Eric
|
|
|
|
|
The essence of virtual functions/methods is that you don't know and don't have to know what is the actual implementation of the given function. This can be used to design code that uses polymorphism (object oriented programming), or to hide actual implementation from the user (this is probably what happens in your case).
The interface is implemented by the flight simulator. They have a class that implements that interface and that class is already compiled to the flight simulator. You dont need the source of that implementation either. That interface is just an instruction to your compiler, it tells your compiler how to call the already compiled implementation (binary) inside the flight simulator. Maybe you should look up some tutorials on how virtual functions work in c++ (vtables), its useful to know how things work in detail.
|
|
|
|
|
Simplifying the previous answer a bit.
You think that pCallBack is a IPanelCCallback.
It isn't. It is some other class.
That class is derived from IPanelCCallback so your code works.
|
|
|
|
|
Hello!
My question is, do STL containers store a copy or they keep the object itself? Hence, can I insert a local variable into a container?
typedef struct{
int number;
string name; } myData;
class myClass{
void InsertData(myData &a);
void oneFunction ();
void otherFunction ();
collection <myData> m_memberCollection;
};
void oneFunction() {
myData localData;
localData.name = "Paul";
InsertData(localData);
}
void myClass::InsertData (myData &a){
a.number = 5;
m_memberCollection.push_back(a);
}
void otherFunction (){
printf ("%s %d", m_memberCollection[0].name.c_str(), m_memberCollection[0].number);
}
Or do I need to use a collection of the type collection<*myData>?
|
|
|
|
|
piul wrote: Or do I need to use a collection of the type collection<*myData>?
That would be a collection of pointers, which would probably lead to memory leaks and/or loss of information. Your original code takes copies of the objects and adds them to the collection, so the data is retained.
|
|
|
|
|
|
Not particularly pertinent to the question but just a quick note... You don't need to write typedef structs in C++ the way you do in C to introduce a new type. Instead you can just write:
struct myData
{
};and it'll do almost the same thing as your typedef (try adding some member functions to see what the difference is).
|
|
|
|
|
I wonder how many people will catch that hint, so here goes:
The only difference is that the typedef'd struct is unnamed, so you cannot add a constructor or destructor! You could fix that by adding a name explicitely:
typedef struct myData_struct {
...
} myData;
But of course then you wouldn't need the typedef at all since myData would be synonymous to myData_struct.
|
|
|
|
|
Absolutely right, but for those of us with K&R hardwired into the brain it's a tough habit to kick.
|
|
|
|
|
hi
i am calling the video player exe in my application that will show preview for the user.
now the problems is that i want to disable the child window or window untill this exe run.after completion of preview the window should enable .
Here is my code :
void CPreview::OnpreviewBtnClk()
{
::MessageBox(m_hWnd,_T("here is the preview "),_T("INFORMATION"),MB_ICONINFORMATION);
TCHAR szCmdline[MAX_PATH];
CString szWindowTitle = _T("rs-mplayer");
VERIFY(::GetModuleFileName(AfxGetApp()->m_hInstance, szCmdline, _MAX_PATH));
szCmdline [StrRChr (szCmdline, NULL, _T('\\')) - szCmdline] = _T('\0');
_tcscat_s(szCmdline, _T("\\rs-mplayer.exe /"));
ShellExecute(NULL,_T("open"),szCmdline,NULL,sourceFileName,SW_SHOWNORMAL);
Sleep(500);
CWnd *pOtherWnd = CWnd::FindWindow(NULL, szWindowTitle);
h_Rs_playerProcessId =pOtherWnd;
if (pOtherWnd)
{
MYDATA m_mydata;
wcscpy(m_mydata.m_pName,sourceFileName);
m_mydata.m_bNameLength = (BYTE)sourceFileName.GetLength();
m_mydata.m_ullNodeSize = m_li64FileSize.QuadPart;
COPYDATASTRUCT cpd;
cpd.dwData = 1;
cpd.cbData = sizeof(MYDATA);
cpd.lpData = (void*)&m_mydata;
copyDataResult = pOtherWnd->SendMessage(WM_COPYDATA,
(WPARAM)AfxGetApp()->m_pMainWnd->GetSafeHwnd(),
(LPARAM)&cpd);
}
ProcessID;
<pre><pre>
}
Please reply me soon
got struct from past two days
Best Regards
sarfaraz
|
|
|
|
|
1. Instead of ShellExecute, use CreateProcess[^]. This will give you a handle to the process in the struct specified for lpProcessInformation.
2. Use EnableWindow[^] to disable whatever you wish to disable.
3. Run a timer with SetTimer[^], e.g. let it run every 5 seconds.
4. In the timer function, use WaitForSingleObject[^] with the handle of the process and a timeout of zero. When this gives you WAIT_OBJECT_0, it means that the process has exited. If it did not, do nothing special.
5. If the process DID exit (you got the event signalled in step 4), then kill the timer with KillTimer[^], use EnableWindow to re-enable what you need and don't forget to close the handles you got in that structure in step 1.
Wouldn't this work for you?
> The problem with computers is that they do what you tell them to do and not what you want them to do. <
> If it doesn't matter, it's antimatter.<
|
|
|
|
|
while(pOtherWnd!=0)
{
pOtherWnd = CWnd::FindWindow(NULL, szWindowTitle);
CPreview::EnableWindow(FALSE);
if(pOtherWnd == NULL)
{
CPreview::EnableWindow(TRUE);
break;
}
sleepmode=5000;
Sleep(sleepmode);
}<pre<pre> lang="c++"></
i got the pOtherWnd is handle
i set the timer sleepmode in sleep(sleepmode)
its working fine after some time the window will strat not responding ..
i have added the code with respect to the earlier code which i had posted .
i tried with waitsingleobject by getting the handle og main window even though its not working
help me out this
Best
Regards
sarfaraz
|
|
|
|
|
Noone said you should use WaitForSingleObject with the window handle, of coure that doesn't work. And your aproach is wrong because while you are running this while loop -i assume you do this in the main GUI thread- your application is not processing any incoming messages, thus it will not respond.
> The problem with computers is that they do what you tell them to do and not what you want them to do. <
> If it doesn't matter, it's antimatter.<
|
|
|
|
|
hey bro i got it ..
i added ontimer msg to GUI thread.
thanks to you for sharing your knowledge
|
|
|
|
|
My CMainFrame class derive from CFrameWnd,and create toolbar but can't display tooltips, the problem confuse me for long time.
int CMainFrame::OnCreate(LPCREATESTRUCT lpCreateStruct)
{
if (CFrameWnd::OnCreate(lpCreateStruct) == -1)
return -1;
if (!m_wndToolBar.CreateEx(this, TBSTYLE_FLAT, WS_CHILD | WS_VISIBLE | CBRS_TOP
| CBRS_GRIPPER | CBRS_TOOLTIPS | CBRS_FLYBY | CBRS_SIZE_DYNAMIC) ||
!m_wndToolBar.LoadToolBar(IDR_MAINFRAME))
{
TRACE0("Failed to create toolbar\n");
return -1;
}
}
|
|
|
|
|
|
For each button in the toolbar, have you added a string resource with the same ID and added the text after the newline ('\n') character?
"One man's wage rise is another man's price increase." - Harold Wilson
"Fireproof doesn't mean the fire will never come. It means when the fire comes that you will be able to withstand it." - Michael Simmons
"Show me a community that obeys the Ten Commandments and I'll show you a less crowded prison system." - Anonymous
|
|
|
|
|
Hi,
I have a dialog based application. I have a picture control on a dialog. I am drawing a line in OnMouseMove handler. I want to hande Undo /Redo events on that control. If a user has drawn a line then clicking on Undo button will erase the line and Redo will again draw the same line deleted. I have seen that events handled in a SDI application. But I want to do the same in a dialog based application.
Any help will be appreciated.
Regards,
|
|
|
|
|
a very basic single-step undo/redo scheme is :
get two off-screen images: call them undo and redo.
before every action that can be undone, copy the current image to the 'undo' image, clear the 'redo' image.
to undo: copy the current image to 'redo', copy 'undo' to the current image
to redo: copy the current image to 'undo', copy 'redo' to the current image
to do more steps, you'll need an array or queue of images.
|
|
|
|
|
Hi,
Thanx for the reply.
But for this we need to save the image at every instance we modify the image. I was actually looking at Command Design pattern. But I have no idea, how to handle the MouseMove event in case I am using Command design pattern.
if you have any sample code using command design patern with dialog based apps, help will be appreciated.
Can u give me a sample code for the idea you had given. I was doing the same way but still it was difficult with MouseMove handler event to save the images. I was saving two images, when you do any modification on again, I'll copy the third image to second and second to first, but it was hadnling the MouseMove event properly.
Also the reason I want to use command pattern is that I want to Undo/Redo to some specified level, say 10 or 20 level.
Regards,
modified 9-Jul-12 8:39am.
|
|
|
|
|
mbatra31 wrote: But for this we need to save the image at every instance we modify the image.
or, you can store just the part which was modified. that will reduce memory usage.
this Colorizing edit control[^] has unlimited undo/redo, using something like the command pattern. but text handling is much different from images.
modified 9-Jul-12 9:31am.
|
|
|
|
|
You have to think things over.
The undo is usually the harder thing to do because you need to restore the previous drawing stage.
For example, if you have a "spray-can" tool, the undo of that might be quite difficult without storing the previous bitmap; unless you redraw all the steps from the start (could be fast, could be slow depending on the drawing, but that solution (redrawing every thing) is quite easy to implement.
Saving the bitmap is not incompatible with having a command pattern,
Watched code never compiles.
|
|
|
|
|
|
Hi,
Thanx for the reply.
I was actually looking at Command Design pattern. But I have no idea, how to handle the MouseMove event in case I am using Command design pattern.
if you have any sample code using command design patern with dialog based apps, help will be appreciated.
Regards,
modified 9-Jul-12 8:40am.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
