Right now in order to remove the ComboBox element from the FormLayout I use the following method:
formLayout.getElement().removeChild(locationComboBox.getElement());
I still have a validation binder associated with locationComboBox. I may remove binding in the following way:
binder.removeBinding(locationComboBox);
that's works fine. But I'm thinking about the locationComboBox ? Is it enough to just remove the component via removeChild method, or I should call some additional logic in order to completely destroy the locationComboBox element and remove it from the memory? Or this is a task for JVM GC?
formLayout.getElement().removeChild(locationComboBox.getElement()) is equivalent to formLayout.remove(locationComboBox) minus some check that the element is actually a child of the parent.
Both will effectively remove the combo box from the layout, but there may some other hard references to the component (from the binder, or from some listeners) that prevents it from being garbage collected.
Related
I inherited from QGraphicsItemGroup and made a class that keeps a pointer to its contained items so that I can later refer to them and change properties. It has an ellipse item and a line item and I want only the ellipse to be clickable. I need that press event of the ellipse to propagate to the QGraphicsView so that I can send a signal to some surrounding widgets.
So far I tried inheriting also from QGraphicsObject to have signals available but got stuck with ambigous base error when trying to use scene->addItem. I tried casting to QGraphicsItemGroup but I still get the error. I also tried inheriting from QObject with no success.
I'm new to QGraphics and I know the QGraphics framework has a lot of tools for user interaction and even interaction between GraphicsItems but this is really kicking my butt.
What would be the proper way to get this behavior?
Create a separate "emitter" class
To allow your subclass of QGraphicsItemGroup to emit signals, you can create a separate "emitter" class that inherits from QObject. Then, you can add an instance of this emitter class within your subclass of QGraphicsItemGroup. The emitter object can then emit signals for your subclass as needed.
QGraphicsItemGroup is treated as a single item
Unfortunately, an instance of QGraphicsItemGroup is treated as a single item, so each mousePressEvent will belong to the entire group rather than one of the members of that group (i.e., the ellipse item or the line item). If you want the mousePressEvent to behave differently depending on which item is clicked, they will need to be separate items, or you could try using line->setParentItem(ellipse) to link up the 2 items without using QGraphicsItemGroup.
I have made a HomeViewModel which contains some other ViewModels to be presented in a TabParPresenter. This works fine and the ViewModels associated Views are presented correctly in their corresponding tabs. However on of the ViewModels have a check in the ctor that tells it in when some conditions apply it needs to navigate to SettingsViewModel , which is not a part of the ViewModels contained in HomeViewModel.
The navigation event gets caught by the TabBarPresenterHost, which is simply the Presenter of the application and if a TabBarPresenter is present and not null it is passed to the TabBarPresenter.ShowView() method. All this happens before the TabBarPresenter is done loading and SelectedViewController is set in ViewDidLoad. Hence the TabBarPresenter relies on using the SelectedViewController to push new Views I obviously get a Null Reference Exception.
In short what I want is:
if (conditionForShowingHome == true)
GoToHome();
else
GoToSettings();
Then when inside SettingsViewModel I have set the stuff I need when going back you return to HomeViewModel.
What breaks this is that the ViewModels are loaded before the View is done loading and the navigation event is executed before the View is done loading. Any ideas how to go around this?
I'm afraid that putting this code inside a ViewModel constructor is likely to lead to problems.
The ViewModel constructors are called during navigations - and not all platforms will support starting new navigations while existing ones are still in progress.
To workaround this, I personally opt to put this type of behaviour in the code before the navigation.
In the case of a normal navigation, this can go inside an MvxCommand execution handler.
In the case of application startup, I do this in a custom app start - see some notes in https://speakerdeck.com/cirrious/appstart-in-mvvmcross
I have a question regarding coded ui UIMap.
Every time I record an action on the same application, coded ui generates a new object for the same window in the application.
It looks like:
UIAdminWindow
UIAdminWindow1
UIAdminWindow2
and so on...
every window class holds different buttons, even though it's the same window.
Thus it's very hard to keep code maintenance.
What i would like is that every time i perform actions and records on a window, even if not at the same time, the already generated class for this window, will be updated with the new controls.
any suggestions to why it happens?
Thanks a lot!
You can clean up your UIMaps by doing two things:
Use the UIMap Toolbox (from codeplex) to move controls within the UIMap so they are all under one control tree.
When you have duplicate UI controls, go to the properties for the action that references the duplicate control and change the UI Control property to point to the original control in the UIMap.
The duplicate trees should now be unreferenced and you can delete it from your map, keeping things clean.
And yes, it's a pain to do, but it's worth it for maintainability.
In UIMap.uitest you can change the action name and the control name for better maintenance.
For example: you can set UIAdminWindow as FirstAcessWindow or other name that will express comfortably the control or the action.
What I can guess is that there is some randomly generated content or element identification data such as class or title that may be causing it. This may be caused by different username for example. Also you can update the element from UI map element tree.
I have a MainView which has an associated MainViewModel in my WPF application, assigned via its contructor.
App > start up uri > MainWindow.xaml
public MainWindow()
{
InitializeComponent();
var viewModel = new MainViewModel();
DataContext = viewModel;
}
My MainView holds as many as four nested views or child views who are hidden and displayed based upon the button that has been clicked on the MainView. So we toggling the visibility property vi a binding which gets updated via command bindings assigned to each button on the MainView.
Each nested View does not have an associated ViewModel, all bindings found on child views find their information in the MainViewModel. So binding system ends waling up the UI tree of the app to find out that the parent 'MainView' has an associated ViewModel.
So overall there is 'ONE' -> ViewModel. This seems to work fine and but as expected this VM has gotten to big and needs re-factoring. It holds information that contextually it should not. But this is a proof concept application. So i decided to keep it simple and make sure it was do-able.
PROBLEM:
When i tried assigning a empty view with an empty view model I noticed binding errors in the output window and as expected weird and broken behaviour. Which makes no sense ... Is there a more clear and concise way of letting WPF know how to handle any bindings it finds in a nested view control? I thought if each view's constructor assigned itself a corresponding VM as shown above, then it should work as this logically makes sense. Unfortunately all buttons on the MainView stop working when the corresponding view it is designated to switch on and hide the others has an associated ViewModel. On some buttons it works and the others it does not? This is really weird?
As mentioned in my answer above, the problem was that WPF binding system was struggling to to resolve bindings at run time. The main view has its associated view model instantiated and assigned via the Main View contructor and this pattern is repeated for all nested views that the MainView also houses.
By default, I tend to use the implied binding syntax which means that without explicitly specifiying a source the binding system will try to resolve the name you supply in the binding. So it's all implied and nothing is explicitly set!
Upgrading each nested view to have its own view model makes this auto discovery/resolution of bindings go a little crazy and I have not explicitly told the binding system where to find the property I am looking for hence the output window binding errors.
This leads to unexpected behaviour as the output window was telling that it was trying to resolve binding expressions in nested views --> viewmodels. When in actual fact that VM it is looking in, IS EMPTY!
So clearly the binding system is good when you do not explicitly set a source property inside the binding syntax. It is clever enough to find things on its own. In my case it needed some extra help, as it was unsure as to where to find things.
SOLUTION:
Remove the constructor declaration for the MainViewModel in the MainView constructor.
Scope in an xmlns for ViewModels namesapce into your MainView.xaml
Create a window resource inside the MainView .xaml
Give the resource a key.
Upgrade all your bindings in the MainView xaml file to include source property.
Give the source property a static resource binding that points to your ViewModel key value set up in step 4.
Only do step 6 for bindings who refer to the ViewModel that is associate with the MainView.
All nested views are to be left alone, they should handle their own bindingds in their own xaml files. The MainView simply instantiates them and places them onto the UI. For me this was the case, I did not have any more bindings concerning my nested views. All Bindings that lived on the MainView.xaml file referred to data in the MainViewModel.cs. This makes alot easier when viewing your problem.
For some reason the editor was being awkward so I chose to omit any sample code but the steps above are descriptive enough to follow the steps that I took. Above is what worked for me.
ANOTHER WAY OF SUMMING UP THIS PROBLEM
Most books teach the shorter binding syntax
What happens when more than one data context is available?
How is the binding system supposed to know how to resolve your short hand binding expressions.
I'm looking for a ready made, freely available component that provides the following features:
editing list entries
deleting entries
adding a new entry - possibly invoked multiple times
bind to property in managed bean of type List, or similar
optional: reordering entries
optional, simple: allow verification that at least one element is entered (required="true"? required="1"?)
optional, better: restrict by minimum and maximum allowed number of elements
The simplest way to allow editing entries would be to represent them as input element. Preferably the element won't need to sync to the server for every change in the number of elements.
Alternatively: is there a trivial way to create or compose such a component oneself?
If this if of importance: currently I'm using Richfaces 3.3.3 and MyFaces 2.0.
Edit: The component is expected mostly to be used for creating new lists from scratch, so that should be as simple for the user as possible. Preferably just tabbing to the next empty input element automatically created at the end of the list. I'd like to use the component for editing these lists, too, but that feature is expected to be used much less often.
You could take a look at Primefaces p:collector. The example in the Primefaces showcase is a good starting point. It can be used for adding elements to a list and for editing and removing these elements.
Probably the simplest way that I would know of going about almost all of these is the PrimeFaces dataTable component. With the exception of adding and deleting entries inline, all of this comes out of box. Deleting entries can be as simple as adding a delete button in a row and adding entries can be done through invoking a modal dialog with a form to create a new entry, add it to the backing bean list for the table, then asynchronously updating the dataTable.
http://www.primefaces.org/showcase/ui/datatableHome.jsf
It really does seem to be the most fully featured JSF dataTable component out there.