C / C++ / MFC

7-Nov-02 2:43

I need to avoid that a dialog get the focus when it's called.
How can I make this ??

[]'s
Cris.

includeh107-Nov-02 2:53

7-Nov-02 2:53

do not pass WM_SETFOCUS to default window process.
i.e. in MFC
MyCls::OnSetFocus(*)
{
//cancel this line
//CWnd::OnSetFocus(*);
}

includeh10

Cris7-Nov-02 7:30

7-Nov-02 7:30

In the ClassWizard don't have the WM_SETFOCUS event. But I did write the event and it's not trigged. Do you know why ??

[]'s

includeh107-Nov-02 8:30

7-Nov-02 8:30

goto last page, select window, it will appear.

includeh10

Cris8-Nov-02 0:17

8-Nov-02 0:17

What ??
Last page ??

The WM_SETFOCUS don't appear int the MCF Class Wizard dialg.

[]'s

includeh108-Nov-02 1:25

8-Nov-02 1:25

What ??
Last page ??

Last page of Class wizard!

if still don't understand, let me know with no hesitate!

includeh10

Cris8-Nov-02 1:36

RegisterWindowMessage vs. WM_USER

8-Nov-02 1:36

Ok, thanks for your help ...

[]'s
Cris.

CListCtrl Oddity

Richard Green7-Nov-02 2:02

Richard Green

7-Nov-02 2:02

'ello,
I have a problem with a CListView derived class, it refuses to show the Header control when in report view. I've tried a few things including commenting out all except

GetListCtrl().MofifyStyle(LVS_TYPEMASK,LVS_REPORT);<br />
GetListCtrl().InsertColumn(0,"Test",LVCFMT_LEFT,100);

in OnInitalUpdate().

The app is a appwizard produced SDI app with the "Windows Explorer" option enabled" and has the Listview view options buttons on the toolbar, the selected mode shows as the Details mode when the app starts but the CHeaderCtrl() is never created ( GetListCtrl()->GetHeaderCtrl() returns NULL ).

Another appwizard produced app with all the same modes works ok, anyone else as confused as me? WTF | :WTF:

[EDIT]
Never mInd i forgot to delete the defualt (empty) implmentation of OnStyleChanged [/EDIT]
Richard Green

LowLevelMouseProc

Gleb7-Nov-02 0:58

Gleb

7-Nov-02 0:58

Does anyone knows how can i capture mouse movements from the driver of the mouse and not by capturing WM_MOUSEMOVE.
I need somethig like "LowLevelMouseProc" for all windows and not only for 2000 and XP.
Thanks.

Re: LowLevelMouseProc

Stephane Rodriguez.7-Nov-02 1:52

Stephane Rodriguez.

7-Nov-02 1:52

::GetCursorPos

Back to real work : D-23.

Alexinuk7-Nov-02 0:30

Alexinuk

7-Nov-02 0:30

Is there any reason why i should not use RegisterWindowMessage() to define messages specific to my applications (i.e. not used for inter-process communication). MSDN says that I should use WM_USER + i to do this, and not RegisterWindowMessage(), but having read several discussions on the topic I am not sure that I see why...

Any ideas?

Alex Confused | :confused:

jhwurmbach7-Nov-02 1:21

jhwurmbach

7-Nov-02 1:21

In fact, RegisterWindowMessage() is the most desireable version of a private message.
(WM_USER+const being the least favorable, WM_APP+const intermediate).
You just have to remember to make your handler-function HRESULT name(WPARAM, LPARAM) and use the ON_REGISTRED_MESSAGE() macro.

Alexinuk7-Nov-02 1:26

Alexinuk

7-Nov-02 1:26

jhwurmbach wrote:
In fact, RegisterWindowMessage() is the most desireable version of a private message.

Is that because they're guaranteed to be unique to the string used to identify them and therefore you get less potential conflicts?

Are you saying that there is no situation where a RegisterWindowsMessage is less preferable to a WM_USER + i one?

Cheers, Alex.

jhwurmbach7-Nov-02 1:34

jhwurmbach

7-Nov-02 1:34

You should never use WM_USER+i, because some i are already used by some (newer) windows controls. You get almost undebuggable errors when you are reacting to messages that are not meant for you.
Use WM_APP+i, this is guaranteed not to be used by Windows internally.
You then only have conflicts with controls from other sources (commertial or third party).
Use of RegisterWindowsMessage() avoids all this at no cost.
No, I cant think of a situation where WM_USER is more preferable than RegisterWindowsMessage().

Tim Smith7-Nov-02 2:44

Tim Smith

7-Nov-02 2:44

If you are making your own control, WM_USER+x is fine.

However, for application level messages, I have to agree that RWM is better.

Tim Smith

I'm going to patent thought. I have yet to see any prior art.

Alexandru Savescu7-Nov-02 2:48

Alexandru Savescu

7-Nov-02 2:48

You should use RegisterWindowMessage only. Read this[^] for details!

Best regards,
Alexandru Savescu

Complex String Comparison

Alexinuk7-Nov-02 2:53

Alexinuk

7-Nov-02 2:53

Alexandru Savescu wrote:
You should use RegisterWindowMessage only. Read this[^] for details!

This (and that on message management) was one of the articles I was referring to in my originaly post. I was wondering why MSDN says that you should only use RegisterWindowsMessage() for inter-process messages and WM_USER + i for the others.

I suppose it's just like a lot of the other bits in MSDN - looks absolutely fine until you talk to other people about it...;)

James Spibey7-Nov-02 0:15

James Spibey

7-Nov-02 0:15

What I need to do is compare a series of strings (actually filenames) and find potential duplicates. The strings, however, are unlikely to contain literal duplicates so what I need to do is find similar strings. Has anyone ever had experience of this before?

Cheers

James

Re: Complex String Comparison

Daniel Turini7-Nov-02 0:23

Daniel Turini

7-Nov-02 0:23

Yes, this is a common need in credit applications (find names mistyped). In your case, I suggest the Levenshtein String (or Edit) Distance algorithm for this. You can find tons of implementations in C++ on google.

lazy isn't my middle name.. its my first.. people just keep calling me Mel cause that's what they put on my drivers license. - Mel Feik

Re: Complex String Comparison

James Spibey7-Nov-02 0:42

James Spibey

7-Nov-02 0:42

Thanks for that Daniel. I took a look at it and that solves part of my problem. The other part of the problem is that I need to find duplicates based on whether a string has some elements of another.For example, I would want to match the following filenames

c:\Music\Albums\Vines\01 - Highly Evolved.mp3
c:\Music\Singles\The Vines - Highly Evolved.mp3
f:\Stuff\Vines - Highly Evolved.wma

Do you know if there is a standard way to do this? I don't think there is but I just wanted to make sure.

Cheers

James

Re: Complex String Comparison

jhwurmbach7-Nov-02 1:35

jhwurmbach

7-Nov-02 1:35

If you are really into informatics as a science, maybe some algorithms from bioinformatics can help you. Finding substring in one another and computing similarity-distance is quite common in DNA-sequence handling.
You will find more stuff about that in the net than you are able to read in the entire rest of your life.

Help! about PreTranslateMessage

Simon.W7-Nov-02 0:05

Simon.W

7-Nov-02 0:05

Hi:

I meet a problem, it has puzzled me for a long time. I create a COM dll which support MFC, several

methods was exposured. The implement code of two methods was shown as follow:

STDMETHODIMP CMyComObj::OnConfiguration(long iReserved, BOOL *pbSuccess)<br />
{<br />
	AFX_MANAGE_STATE(AfxGetStaticModuleState())<br />
	<br />
	CMyCfgDlg dlg;<br />
	int rtn = dlg.DoModal();<br />
	*pbSuccess = (rtn == IDOK) ? TRUE : FALSE;<br />
	return S_OK;<br />
}<br />
 <br />
STDMETHODIMP CMyComObj::ShowUIFace(long hParentWnd)<br />
{<br />
	AFX_MANAGE_STATE(AfxGetStaticModuleState())<br />
	<br />
//	declare in header file as CMyUIDlg *pdlg;<br />
	pdlg = new CMyUIDlg();<br />
	if(pdlg != NULL)<br />
	{<br />
		pdlg->Create(ID_MYUIDLG,CWnd::Fromhandle((HWND)hParentWnd));//Create success<br />
	}<br />
	*pbSuccess = (rtn == IDOK) ? TRUE : FALSE;<br />
	return S_OK;<br />
}<br />

The problem is in the first method call, the dialog shown normal and I can use TAB switch focus from

one control to another control. When I use the second method call, the dialog shown normal but can't

use TAB to switch focus(It does not take effect). The reason is virtual function PreTranslateMessage

was not be processed, What's wrong? Please help me!!!

I am seeking...
For what?
Why did you ask me for what? I don't know!

Re: Help! about PreTranslateMessage

includeh107-Nov-02 3:09