OK, I got it. Thanks for your patience. Thanks a lot.
"Joseph M. Newcomer" wrote:
Generally, when you are adding new views, you do not call AddDocTemplate. See my essay on
MSDN Errors and Omissions, where I show the steps required to add a view.
http://www.flounder.com/msdn_documentation_errors_and_omissions.htm#CDocument::AddView
Why do you want a doc2/view2 pair? Presumably all your views are to the same document.
See more, below
joe
On Tue, 13 Jan 2009 02:03:01 -0800, 97612 <97612@discussions.microsoft.com> wrote:
Another question:
I'm trying to create only region 2,3,4,5(a 2 x 2 layout spilt in MainFrame)
in my working project because I can successfully create the region 2 and 3
with thumbnail functionality. I'm going to leverage my working project.
And the code is the following:
//******** code *********************
m_wndSplitter.CreateStatic(this, 2, 2);
m_wndSplitter.CreateView(0, 0, RUNTIME_CLASS(CTreeBrowserView),
CSize(200,400), pContext);
m_wndSplitter.CreateView(0, 1, RUNTIME_CLASS(CThumbView), CSize(400,400),
pContext);
m_wndSplitter.CreateView(1, 0, RUNTIME_CLASS(CTreeBrowserView),
CSize(200,400), pContext);
m_wndSplitter.CreateView(1, 1, RUNTIME_CLASS(CThumbView), CSize(200,400),
pContext);
//******** code *********************
It'll cause a problem that when I click the folder showed in region 2, the
thumbnail of image files in that folder are showed in region 5, but not
region 3 as I wanted. I found the reason of the problem is in:
//******** code *********************
at CXXXDOC:
while( pos != NULL )
{
pView = GetNextView( pos );
if( pView->IsKindOf(RUNTIME_CLASS(CThumbView)) == TRUE )
pThumbView = (CThumbView*)pView;
}
*****
There is something seriously wrong if your document knows about any of the view types. It
should neither know nor care that these types exist. If it does, rethink your design. As
Tom already pointed out, just do an UpdateAllViews with an lHint/pHint pair that
communicates to the views that they have something to update. For example, a project I'm
working on this week has
void CMyView::OnUpdate(CView * pSender, LPARAM lHint, CObject * pHint)
{
if(lHint == 0 && pHint == NULL)
{
CView::OnUpdate(pSender, lHint, pHint);
return;
}
switch(lHint)
{
case UPDATE_RECORD:
...
break;
case UPDATE_FIELD:
...
break;
case UPDATE_SELECTION:
...
break;
}
}
where each of those codes causes specific actions to be taken.
If you ever write #include "SomeView.h" in your CDocument class, you have made a serious
design error. This is your indication that you did something wrong. Don't do it.
*****
//******** code *********************
The pThumbView finally gets the pointer of region 5(CThumbView), but not the
pointer of region 3(CThumbView).
Is there any solutions for the problem?
*****
Rewrite the code so the document knows NOTHING about the types of the views!
*****
May I write a CXXXDoc2(the same function of CXXXDoc) and a CThumbView2(the
same function of CThumbView).
And then do the following:
//******** code *********************
at CXXXDoc2:
while( pos != NULL )
{
pView = GetNextView( pos );
if( pView->IsKindOf(RUNTIME_CLASS(CTumbView2)) == TRUE )
pCTumbView2 = (CTumbView2*)pView;
}
*****
This is completely wrong, because it suggests there is a totally different document, and
as already indicated, the document should never know anything about the types of the
views!
*****
at MyApp:
CSingleDocTemplate* pDocTemplate;
pDocTemplate = new CSingleDocTemplate(
IDR_MAINFRAME,
RUNTIME_CLASS(CXXXDoc),
RUNTIME_CLASS(CMainFrame),
RUNTIME_CLASS(CTreeBrowserView));
AddDocTemplate(pDocTemplate);
****
Do not AddDocTemplate
****
CSingleDocTemplate* pDocTemplate;
pDocTemplate = new CSingleDocTemplate(
IDR_MAINFRAME,
RUNTIME_CLASS(CXXXDoc2),
RUNTIME_CLASS(CMainFrame),
RUNTIME_CLASS(CTreeBrowserView));
AddDocTemplate(pDocTemplate);
//******** code *********************
Is the way I thought correct? Or there's a normal way to achieve that?
"97612" wrote:
OK, I got what you mean.
But another problem is that how to access the six VIEW's pointer in DOC?
And how the six VIEWs access the DOC?
My beginning work(just split MainFrame into two regions(or windows?)) is the
following:
//**** code *****
at MainFrm.cpp:
BOOL CMainFrame::OnCreateClient(LPCREATESTRUCT lpcs, CCreateContext*
pContext){
m_wndSplitter.CreateStatic(this, 1, 2);
m_wndSplitter.CreateView(0, 0, RUNTIME_CLASS(CXXXView1), CSize(200,400),
pContext);
m_wndSplitter.CreateView(0, 1, RUNTIME_CLASS(CXXXView2), CSize(400,400),
pContext);
}
*******************************************
at DOC:
CXXXView* pXXXView = NULL;
POSITION pos = GetFirstViewPosition();
CView* pView = NULL;
while( pos != NULL )
{
pView = GetNextView( pos );
if( pView->IsKindOf(RUNTIME_CLASS(CXXXView)) == TRUE )
pXXXView = (CXXXView*)pView;
}
*************************************************************
at VIEW1 and VIEW2:
CXXXDoc *pDoc = (CXXXDoc*)GetDocument();
//**** code *****
It works in my working project. But my question is that is it the correct way?
If yes, can I conclude one fact that the split six regions in MainFrame is
bounded with DOC by the following codes?! And I can use the way above to get
the pointer of DOC in VIEW1~VIEW6, also I can get the pointer of VIEW1~VIEW6
in DOC?
//***** code ******
at MyApp.cpp:
CSingleDocTemplate* pDocTemplate;
pDocTemplate = new CSingleDocTemplate(
IDR_MAINFRAME,
RUNTIME_CLASS(CXXXDoc),
RUNTIME_CLASS(CMainFrame),
RUNTIME_CLASS(CXXXView1));
AddDocTemplate(pDocTemplate);
//****** code ********
"Scott McPhillips [MVP]" wrote:
There are two ways to do it. If you want the user to be able to resize the
regions then you would use splitter windows. If resizing is not important
you could do it all by placing 6 controls on a single CFormView.
For splitters, you create splitter windows inside splitter windows. For
example, the first splitter contains region 1 and regions 23456. The second
splitter fills 23456 and splits 6 and 2345. Then a third splitter splits up
2, 3, 4 and 5. The VIEWEX MFC sample has an example of splitters in
splitters.
If you don't need the splitter features then create all controls on a
CFormView. In CMainFrame::OnSize you would reposition and resize all 6
controls to fill the CMainFrame new size.
"HSNU_976" <HSNU_976@discussions.microsoft.com> wrote in message
news:391EDDD4-3A17-4927-82E8-F43CD25C8037@microsoft.com...
Hi, all
The following is my desired UI layout:
--------------------------------------
| 1 |
--------------------------------------
| 2 | 3 | |
| | | |
------------------------- 6 |
| 4 | 5 | |
| | | |
--------------------------------------
The region 2 is the tree view of the folders in my local PC, and region 3
will show the thumbnails of image files that in the folder selected in the
region 2.
The region 4 is the tree view of the folders in remote PC(I've created a
local folder in my PC for it, so it just a local folder in my PC in
developement stage), and region 5 wil show the thumbnails of image files
that
in the folder selected in the region 4.
The region 1 will show the selected image files in the region 3.
The region 6 have some functions that can perform on the selected image
file
in the region 3.
Is any one can tell how to achieve that!?
Below are what I thought:
Use SDI.
Also write a MyMainFrame class with CSplitterWnd that split the main frame
into six region.
Then write a DOC class and six VIEWs for each region.
But I don't know how to bound and use the VIEWs and DOC with MyMainFrame
that can achieve my desired goal.
Thanks for your anwsers.
--
Scott McPhillips [VC++ MVP]
Joseph M. Newcomer [MVP]
email: newcomer@flounder.com
Web: http://www.flounder.com
MVP Tips: http://www.flounder.com/mvp_tips.htm