Edited by OP.
My program is in need of a lot of cleanup and restructuring.
In another post I asked about leaving the MFC DocView framework and going to the WinProc & Message Loop way (what is that called for short?). Well at present I am thinking that I should clean up what I have in Doc View and perhaps later convert to non-MFC it that even makes sense. My Document class currently has almost nothing useful in it.
I think a place to start is the InitInstance() function (posted below).
In this part:
POSITION pos=pDocTemplate->GetFirstDocPosition();
CLCWDoc *pDoc=(CLCWDoc *)pDocTemplate->GetNextDoc(pos);
ASSERT_VALID(pDoc);
POSITION vpos=pDoc->GetFirstViewPosition();
CChildView *pCV=(CChildView *)pDoc->GetNextView(vpos);
This seem strange to me. I only have one doc and one view. I feel like I am going about it backwards with GetNextDoc() and GetNextView(). To try to use a silly analogy; it's like I have a book in my hand but I have to look up in it's index to find out what page the Title of the book is on. I'm tired of feeling embarrassed about my code. I either need correction or reassurance, or both. :)
Also, all the miscellaneous items are in no particular order. I would like to rearrange them into an order that may be more standard, structured or straightforward.
ALL suggestions welcome!
BOOL CLCWApp::InitInstance()
{
InitCommonControls();
if(!AfxOleInit())
return FALSE;
// Initialize the Toolbar dll. (Toolbar code by Nikolay Denisov.)
InitGuiLibDLL(); // NOTE: insert GuiLib.dll into the resource chain
SetRegistryKey(_T("Real Name Removed"));
// Register document templates
CSingleDocTemplate* pDocTemplate;
pDocTemplate = new CSingleDocTemplate(
IDR_MAINFRAME,
RUNTIME_CLASS(CLCWDoc),
RUNTIME_CLASS(CMainFrame),
RUNTIME_CLASS(CChildView));
AddDocTemplate(pDocTemplate);
// Parse command line for standard shell commands, DDE, file open
CCmdLineInfo cmdInfo;
ParseCommandLine(cmdInfo);
// Dispatch commands specified on the command line
// The window frame appears on the screen in here.
if (!ProcessShellCommand(cmdInfo))
{
AfxMessageBox("Failure processing Command Line");
return FALSE;
}
POSITION pos=pDocTemplate->GetFirstDocPosition();
CLCWDoc *pDoc=(CLCWDoc *)pDocTemplate->GetNextDoc(pos);
ASSERT_VALID(pDoc);
POSITION vpos=pDoc->GetFirstViewPosition();
CChildView *pCV=(CChildView *)pDoc->GetNextView(vpos);
if(!cmdInfo.m_Fn1.IsEmpty() && !cmdInfo.m_Fn2.IsEmpty())
{
pCV->OpenF1(cmdInfo.m_Fn1);
pCV->OpenF2(cmdInfo.m_Fn2);
pCV->DoCompare(); // Sends a paint message when complete
}
// enable file manager drag/drop and DDE Execute open
m_pMainWnd->DragAcceptFiles(TRUE);
m_pMainWnd->ShowWindow(SW_SHOWNORMAL);
m_pMainWnd->UpdateWindow(); // paints the window background
pCV->bDoSize=true; //Prevent a dozen useless size calculations
return TRUE;
}
Thanks
Hard to give you good recommendations without knowing what your program shall do. I have only a few general remarks:
Your InitInstance does not look very messed up for me. It's pretty much standard with a bit of custom code in it.
Also the ugly construction to retrieve the first view from the application class (the chain GetDocTemplate -> GetDoc -> GetView) is standard to my knowledge. I actually don't know another way. You might think about moving it into a separate method like CChildView* CLCWApp::GetFirstView() but well, that's only cosmetic as long as you need it only at one place.
What you are doing and which data you are placing in your Document class and in your View class(es) is more a semantic question if you only have one view. (You have only one document anyway because it's an SDI application.). From a technical viewpoint often both is possible.
But to be open for (perhaps) later extensions to more than one view and to follow the standard pattern of a doc/view architecture there are a few rules of thumb:
Data which exist and have a meaning independent of the way to present and view them (a document file, a database handle, etc.) belong to the document class. I don't know what your pCV->OpenF1(cmdInfo.m_Fn1) ... and so on does but if it's something like a file or filename or a parameter to be used to access data in any way OpenF1 might be better a method of the document class.
Methods which do any kind of data processing or modification of your underlying data belong to the document class as well
Data and methods which are only needed for a specific way to display a document belong to a view class (for instance a selected font, colours, etc.)
On the other side: If you have a fixed number of views which open with the document it might not be wrong to put view specific data into the document, especially if you want to make those view parameters persistent. An example would be a file with some statistical data - your document - and a splitter frame with two views: one displays the data as a grid table and the other as a pie chart. The table has "view data" describing the order of and width of columns, the pie chart has data to configure the colours of the pie pieces and the legend location, for instance. If you want to make sure that the user gets the last view configuration displayed when he opens the document file you have to store these view parameters somewhere. It wouldn't be wrong or bad design in my opinion to store those parameters in the document too, to store and retrieve them from any permanent storage, even if you need them only in the view classes.
If your application allows to open an unlimited number of views for a document dynamically and those views are only temporary as long as the application runs, storing all view configuration parameters directly in the view classes seems more natural to me. Otherwise in the document you would need to manage any kind of dynamic data structure and establish a relationship between a View and an entry in this data structure (an index in an array, or a key in a map, etc.)
If you are in doubt whether to place any data in the document or view class I'd prefer the document because you always have the easy GetDocument() accessor in the View class to retrieve members or call methods of the Doc. To fetch data from the View into the Document requires to iterate through the list of views. (Remember: Doc-View is a 1-n relationship, even in a SDI application.)
Just a few cents.
Related
Is it possible to associate a view with different documents (not simultaneously of course)?
What I want to achieve: the application can have multiple documents of the same type, that can be added during runtime, and set of suitable views, the main of which is CFormview-based. The user can choose the number of the document to show per combobox in the toolbar. The views then associate themselves with this document and update with data from it.
Is it possible to achieve with CMultiDocTemplate?
If yes, how? especially is it possible to retrieve a document by it order number?
If not I'll probably have to abstain from using the templates altogether, which of course means more coding (and more questions to friendly community), but it will do what I want it to do and not Microsoft thought I should be doing
Yes, you could but not out of the 'box'. Instead I'd recommend a different strategy. Rather than trying to attach a View to an existing CDocument derived class LOAD the Views CDocument with the information you want. That could done by adding a simple Copy method to the CDocument or you could move your actual data into a separate class then just point the CDocument to 'data' you want.
Trying to change the CDocument instance for a CView is going against the MFC grain which normally means lot of ASSERTS.
I have a view defined on an xpage. I also have several filters (based on the columns) that the user can select and combine to filter the results in the view. I generate a query string based on this that I construct in dominoView.search (doing a complete refresh). What I would like to do is get the results of the search so that I can then update some counts displayed elsewhere on the page. I'm having a hard time figuring out where I can perform this logic, though. I'm trying to use view.getAllEntries() and then iterating over the collection. Sometimes it seems like it works, but other times I seem to be getting the unfiltered view. Someone suggested I explicitly call view.FTSearch inside one of the events (beforePageLoad?) and immediately after do my getAllEntries call, saving the results in viewScope, but I get an "Error while browses Notes view" runtime error when I try to do that. Any pointers? TIA!
EDIT: After studying the xpages lifecycle a bit (which is still a little confusing), I think I can fine-tune my question. This is my first stackoverflow question, so I hope this is okay to do and productive....
As I described, I have a dominoView defined on my xpage. A repeat iterates over the rows of the view, displaying certain fields from the documents. If I define a query in the search property, then the repeat correctly displays the reduced set of documents rather than the complete set. (The query is computed in the search property via SSJS from some variables defined in the viewScope in a combobox's eventHandler.) However, if I try to access the current entries in the view inside of the repeat's rendered section (with SSJS) using myView.getAllEntries (where myView is what's defined as the "value" of the repeat), I am still getting all of the documents, even if a query has been done. It seems like at that point, the view variable has already had its search applied (since the repeat works), so why the differing results? Is there another way to access the view's rows? To complicate this further, this is just a simple experiment that might clarify the problem; as I indicated earlier, I don't actually want to access the view data within the repeat, I want to access it in the rendered or value sections of some comboboxes defined before the repeat in the xpage file.
I hope that makes more sense now....
EDIT #2: I forgot to add that if I manually call FTSearch (or FTSearchSorted) before calling myView.getAllEntries, then I think I can make this work. It just seems unnecessary to have to do that in addition to the view's built-in search.
From what I get you want to iterate over the entries in a view that before has been filtered, i.e. whose resulting entry collection is smaller than the the view itself.
What I don't get (yet) is what you want to do with the result, or what you're axpecting to get from the iteration over your filtered view (you're mentioning some counts to be displayed somewhere else).
Probably a good way is to use the view's .getAllEntriesByKey method which returns a NotesViewEntryCollection object which then can be used for your iteration.
Don't forget to recycle the resulting NotesViewEntry objects; reason for this has been explained several times here at stackoverflow.
I have two windows, one is a table view to display content. One is a window for inputting content.
I have an NSTextField binded to a property in my app delegate. When I change a value in the text field, the app delegate property will change. However, if I go the other way and programmatically change the property's binded key, the value of the text field does not update. However, it does in the table view.
What is going wrong here? How can I update text field?
EDIT:
I tried all 3 of these with the same result
[_addEntry setValue:#"Chet" forKey:#"payee"];
[_addEntry setPayee:#"chet"];
_addEntry.payee = #"chet";
EDIT:
Here's a simple example to elaborate on my point
https://dl.dropboxusercontent.com/u/48014761/test.zip
the label and the textfield are bound to the "str" property. It is initially null. press log to see that in the console. press change str button to change the string. Log to verify. Note that the label and the textfield do not update!
EDIT:
"What am I trying to accomplish?"
Here's the project I am working with:
https://dl.dropboxusercontent.com/u/48014761/Write-Offs.zip
I am trying to use Cocoa Bindings as much as possible. I have a table with a bunch of data. However, each entry is going to have an array of images associated with it. Thus, when I add a new entry, I need to open up a new window so I can set the properties of that entry (rather than through the table) along with upload some images.
I am not sure I am implementing the addEntryWindow correctly. It seems like I should create and destroy the window each time I open and close it. This doesn't appear to be happening.
Also, [[self addEntry] setDate:[NSDate date]]; doesn't seem to fix the problem for me.
Thanks
Chet
How are you programmatically changing the field? Are you using KVO? Is the field inside of a NSManagedObject instance?
Bindings work through KVO. Core Data disables some aspects of KVO for its own internal uses and you might be tripping over one of those.
Update your question with the code sample and lets see what is going on.
Update
Both windows don't necessarily need to be in the same xib but they do need to be talking to the same instances. Ideally they should both be talking to the same instance of NSManagedObject and therefore talking to the same NSManagedObjectContext.
Who is the owner of each window?
Is the owner the same?
If not, is a new Core Data stack being created?
Update
In your test, I was able to correct it by how you were editing the property
- (IBAction)press:(id)sender {
[self setStr:#"this"];
}
You were accessing the iVar directly instead of the property. When you access the iVar directly KVO does not fire.
In addition, in your xib files you were accessing self.str which is unnecessary. It should be just str.
Are you doing any direct property access in your actual project?
Update
You can do a get accessor and then a set accessor:
[[self addEntry] setPayee:xxx];
Based on the variables you are using I wonder what you are trying to accomplish. Can you post the exact code of the programmatic change you are trying to enact?
This question covncerns my lack of understanding of how to use the core data undo manager and how to restore a NSManagedObject to its state before editing was done.
I am just learning my way around Core Data. I have my NSManagedObject classes set up with their dynamic accessors. I perform a fetch that returns several NSManagedObject entity results. Content from each of these entity results (first name, last name) get put into a table view, and then the user picks one out of the table for detailed view and then editing.
The detail view controller receives a pointer to the selected NSManagedObject entity. As the user edits the fields, the corresponding property value in the NSManagedObject entity is updated. This seemed like the cleanest way to manage these changes.
Now, rather than committing the changes using save, I want to provide a cancel-editing feature that rolls back to what is in the data base for that entity. I really only want to restore the one entity and not perform the entire refetch.
I tried rollback and I tried NSUndoManager (with beginUndoGrouping and endUndoGrouping), and that is not working. I don't think I understand what rollback is really supposed to do.
But in any case, I still want to restore the property values in just that single entity (taking the lazy approach to only fetch what is needed, which is the one entity) so that my detail view controller can refill its view with the correct information. Right now it is using the NSManagedObject entity values, which contain the edited values, which were cancelled.
I suppose I could just start the edit process by creating a copy of the NSManagedObject. If the cancel-editing button is pressed, I could copy it back into the original. (I might even be able to just replace the original with the copy by moving the pointer. But since the pointer has actually been passed through several objects, I'm not sure how to manage the retain number on the copy.)
Does anyone have any other suggestions?
Thanks
Using rollback should accomplish what you want and I'm not sure what it doesn't. It is probably an implementation detail error.
You can find the specific managed object/s that were updated but not yet saved by calling the context's updatedObjects.
I have core data app with an entity OBSERVATION that has as one of its attributes DEALNAME.
I want to reference through Interface Builder or by making custom modifications to an NSArrayController a list of unique sorted dealnames so that I can use them in a pop-up.
I have attempted to use #distinctUnionOfSets (and #distinctUnionOfArrays) but am unable to locate the proper key sequence.
I can sort the ArrayController by providing a sort descriptor, but do not know how to eliminate duplicates.
Are the #distinct... keys the right methodology? It would seem to provide the easiest way to optimize the use of IB.
Is there a predicate form for removing duplicates?
Or do I need to use my custom controller to extract an NSSet of the specific dealnames, put them back in an array and sort it and reference the custom array from IB?
Any help would be appreciated. I am astounded that other have not tried to create a sorted-unique pop-up in tableviews.
You need to take a look at -[NSFetchRequest returnsDistinctResults]. That is the level you need to be handling the uniquing of data.
Although I do not have a definitive answer for you, I think there are two ways you can go about it.
The way you already started. You need to bind the contents array of the PopUp button, not just against the arrayController.arrangedObjects, but continue on the path and somehow filter only objects with distinct "DealName"s. This means - the arrayController presents ALL the entities (and may sort them for you) but the PopUp button will have its contents filter via some sophisticated binding to the array controller.
Make your filtering at the ArrayController level (as suggested in another answer here). Here it depends how you set up the array controller. If It is set up to use an "Entity" (vs. "Class") which means the array controller will fetch CoreData entities directly - you can modify its "Fetch" to only bring a subset of the "OBSERVATION" entities with distinct values of "DEALNAME". I don't know how to control WHICH entities are filtered out in this case. Otherwise, you can setup the arrayController to work with "Class" objects, and then you can fetch the entities yourself (in code) and populate the arrayController programmatically, with just the entities you like.
In the second option, the Popup button should be bound normally to the arrayController's arrangedObjects.