Goal: To be able to find point(X,Y) of element on windows desktop application which been clicked without performing click itself.
Not working solution
I've added mousekeyhook Nuget package to my .net core WPF application, it working fine ( when click and event raising with correct position), but each time that program hit breakpoint( for specific below example - at M_Events_MouseDown method) the mouse starts to lag for about 20 seconds.
Also i have found that in WinForms application it works smoothly
MouseKeyHook implementation:
public partial class MainWindow : Window
{
private IKeyboardMouseEvents m_Events;
public MainWindow()
{
InitializeComponent();
SubscribeGlobal();
}
private void SubscribeGlobal()
{
Unsubscribe();
Subscribe(Hook.GlobalEvents());
}
private void Subscribe(IKeyboardMouseEvents events)
{
m_Events = events;
m_Events.MouseDown += M_Events_MouseDown;
}
private void M_Events_MouseDown(object sender, System.Windows.Forms.MouseEventArgs e)
{
Console.WriteLine($"Mouse down :{e}");
}
private void Unsubscribe()
{
if (m_Events == null)
return;
m_Events.MouseDown -= M_Events_MouseDown;
}
}
This is a known issue and low-level hook behavior by design.
When you hook into the mouse event processing chain, your code is one of many other processors. Every hooked piece of code has a possibility to suppress or modify further processing. Nevertheless, there is a time limit to do so. If you set a breakpoint, windows call your code, your code does not respond, a timeout occurs and it starts over.
See a similar question for details. The second answer with more upvotes is the better one.: Low level mouse hook - mouse freeze on breakpoint
Related
I want to run integration UI tests on my WPF application, and I'm not sure how to detect when the current test has finished so that I can proceed to the next one.
Simplifying, suppose I have a button on my window. When the button is clicked I disable it, I modify the model, and I re-enable the button. Once it detects that the model has changed, WPF changes the view on the screen.
Now I want to run a test that simulates clicking the button again and again. To click the button I’ll use automation, as described in this SO question. But how do I know when the work is finished and the display updated, so as to "click" the button again? Do I hook the botton’s IsEnabledChanged, or is there some global indication that the current cycle of processing has finished?
Edit: What was missing in my description is that I want the user to see the interim results on the screen. For example, if the test has 10 phases I want the user to see something like a Step Counter label with values 1 .. 10 appearing on the screen, and not just the number changing immediately from 1 to 10. See my answer below.
how do I know when the work is finished and the display updated, so as to "click" the button again?
According to your description, you said When the button is clicked I disable it, I modify the model, and I re-enable the button.
Therefore, I can only assume that when the model has changed, the Button will be re-enabled. So you could either attach a handler to the model's NotifyPropertyChanged event, or as you suggested, add a handler for the IsEnabledChanged event.
Here is how I managed to get it working. This might be trivial - I'm a novice with GUI. I'm just posting it here in the hope it'll help other novices like me :)
Anyhow, what I used is a two button solutions: Test and Step. Test starts the testing sequence, Step runs each step of the tests. The Step buttons interact with an Integration Tester By Steps helper.
The helper receives an Init with the Number Of Commands as parameter, (currently the helper generates random commands by itself, so it just needs to know how many commands to generate). The helpe provides a Step method to execute the next command, and a Needs More Steps property to indicate whether testing should continue.
The helper derives form INotifyPropertyChanged and has a Counter dependency property that is displayed on the main window.
The states of the Test and Step buttons are controlled by three helper methods: SetButtonsFor_OutsideTesting, SetButtonsFor_InsideTestingOutsideAnyStep and SetButtonsFor_InsideTestingInsideAStep.
First, I verified that everything is working manually, and then I added a timer and automated the process using the Stack Overflow suggestions on how to programmatically click a button in WPF and how to make a WPF Timer Like C# Timer.
Now, here's the Main Window's code:
private void Test_Click(object sender, RoutedEventArgs e)
{
SetButtonsFor_InsideTestingOutsideAnyStep();
RunTheTestBySteps();
}
public readonly IntegrationTesterBySteps _integrationTesterBySteps =
new IntegrationTesterBySteps();
void RunTheTestBySteps()
{
SetButtonsFor_InsideTestingOutsideAnyStep();
IntegrationTesterBySteps.Init(10);
StartTheTimer();
}
private void StartTheTimer()
{
DispatcherTimer = new DispatcherTimer();
DispatcherTimer.Tick += DispatcherTimer_Tick;
DispatcherTimer.Interval = new TimeSpan(0, 0, 1);
DispatcherTimer.Start();
}
private void StopTheTimer()
{
DispatcherTimer.Stop();
DispatcherTimer.Tick -= DispatcherTimer_Tick;
}
private DispatcherTimer DispatcherTimer { get; set; }
private void DispatcherTimer_Tick(object sender, EventArgs e)
{
if (!BtnStep.IsEnabled) return;
ClickTheStepButton();
CommandManager.InvalidateRequerySuggested();
}
private void BtnStep_Click(object sender, RoutedEventArgs e)
{
SetButtonsFor_InsideTestingInsideAStep();
IntegrationTesterBySteps.Step();
if (this.IntegrationTesterBySteps.NeedsMoreSteps)
SetButtonsFor_InsideTestingOutsideAnyStep();
else
{
SetButtonsFor_OutsideTesting();
StopTheTimer();
}
}
private void ClickTheStepButton()
{
var peer = new ButtonAutomationPeer(BtnStep);
var invokeProv = peer.GetPattern(PatternInterface.Invoke)
as IInvokeProvider;
if (invokeProv != null)
invokeProv.Invoke();
}
void SetButtonsFor_InsideTestingInsideAStep()
{
BtnTest.IsEnabled = false;
BtnStep.IsEnabled = false;
}
void SetButtonsFor_InsideTestingOutsideAnyStep()
{
BtnTest.IsEnabled = false;
BtnStep.IsEnabled = true;
}
void SetButtonsFor_OutsideTesting()
{
BtnTest.IsEnabled = true;
BtnStep.IsEnabled = false;
}
I have a WPF application with a page with some code as shown below
public partial class MyPage : Page
{
public MyPage ()
{
InitializeComponent();
}
private void btnClose_Click(object sender, RoutedEventArgs e)
{
this.Cursor = Cursors.Wait;
this.btnClose.Content = "Cancel";
// some long time consuming processing
this.Cursor = Cursors.Arrow;
this.btnClose.Content = "Close";
}
}
I am doing two things here on the Close button click hander which are causing problems. Before long processing I change the button context text to Cancel. I also want to change cursor for whole page to wait. Once long processing is done I set the cursor state and button content back to where it was. However I am facing following two issues.
When application is doing long running operation, I don't get to see the button content as Cancel. It just keep showing me original content CLose.
The cursor changes to Arrow only on the button. However on rest of page,I still keep getting same arrow cursor.
Any ideas how can these issue be solved?
Your code runs on the UI thread by default, so nothing else can be executed on the UI thread (such as re-rendering the UI) until the thread finishes executing.
There are many ways of releasing control of the UI thread before the code finishes executing, but I find the simplest is to use a Task from the Task Parallel Library which can be used to run code on a separate thread.
For example,
// this runs on the main UI thread
this.Cursor = Cursors.Wait;
this.btnClose.Content = "Cancel";
Task.Factory.StartNew(() =>
{
// this code runs on a background thread
// some long time consuming processing
})
.ContinueWith((e) =>
{
// this code runs from the UI thread again
this.Cursor = Cursors.Arrow;
this.btnClose.Content = "Close";
});
It should be noted that UI objects can only be modified on the UI thread, which is why I put the second UI update in the .ContinueWith(...) of the task. An alternative to this would be to use the Dispatcher to ensure code gets executed on the UI thread. If you decide you need this instead and can't find an easy example via Google, let me know and I'll write one here.
This has to be a duplicate some where
public class WaitCursor : IDisposable
{
private Cursor _previousCursor;
public WaitCursor()
{
_previousCursor = Mouse.OverrideCursor;
Mouse.OverrideCursor = Cursors.Wait;
}
#region IDisposable Members
public void Dispose()
{
Mouse.OverrideCursor = _previousCursor;
}
#endregion
}
using (new WaitCursor())
{
// long blocking operation
}
I am having trouble using the WPF Extended Toolkit (version 2.1.0.0) MessageBox from other threads. The namespace is: Xceed.Wpf.Toolkit.MessageBox
I replaced my regular MessageBoxs (System.Windows.MessageBox) with the Toolkit MessageBox and get errors when I launch one from another thread. The System.Windows.MessageBox has no such problems. I saw this posting that reports the problem, but there seems to be no follow up:
https://wpftoolkit.codeplex.com/workitem/21046
I'm guessing there is a work around. An example is presented there that shows the problem, but here is my simple example:
First, I wrap the Toolkit.MessageBox. I do this primarily because I'm applying style (although I've commented that out to show that's not the problem)
public class CustomMessageBox
{
//static DummyUserControl1 _ctrl = new DummyUserControl1();
public static MessageBoxResult Show(string msgText, Style styleArg = null)
{
Cursor saveCursor = Mouse.OverrideCursor;
Mouse.OverrideCursor = null;
//Style style = styleArg != null ? styleArg : _ctrl.FindResource("MessageBoxStyle1") as Style;
// MessageBoxResult result = Xceed.Wpf.Toolkit.MessageBox.Show(msgText, "", MessageBoxButton.OK, style);
MessageBoxResult result = Xceed.Wpf.Toolkit.MessageBox.Show(msgText, "", MessageBoxButton.OK);
Mouse.OverrideCursor = saveCursor;
return result;
}
}
The main window just has two buttons on it, and here's the code behind:
public partial class MainWindow : Window
{
public MainWindow()
{
InitializeComponent();
}
private void btnMainThreadMsgBox_Click(object sender, RoutedEventArgs e)
{
CustomMessageBox.Show("Hello on main thread");
}
private void btnAltThreadMsgBox_Click(object sender, RoutedEventArgs e)
{
Thread altThread1 = new Thread(new ThreadStart(AltThread1Proc));
altThread1.SetApartmentState(ApartmentState.STA);
altThread1.Priority = ThreadPriority.AboveNormal;
altThread1.IsBackground = true;
altThread1.Start();
}
public void AltThread1Proc()
{
MessageBox.Show("Hello on Alt Thread");
CustomMessageBox.Show("Hello on alt thread");
}
}
The problems occur in AltThreadProc() with CustomMessageBox.Show(...). The curious behavior I referred to is this: If you hit the main thead button and then the Alt thread button, you get the error:
Cannot access Freezable 'System.Windows.Media.SolidColorBrush' across threads because it cannot be frozen.
However, if you skip the main thread button and just hit the Alt thread button, you get the error:
The calling thread cannot access this object because a different thread owns it.
I'm curious what the "Freezable" error is all about and why you can get different errors based on what would seem to be an innocuous event: clicking/not clicking button that produces message box on main thread.
Ideally, it would be nice to just replace System.Windows.MessageBox with Xceed.Wpf.Toolkit.MessageBox, but if there is some sort of extra code to write, that might be acceptable. The documentation, and the link I provided hints at using a WindowContainer, but I can't really see any examples of how you do that. I was attracted to the Toolkit MessageBox as it allows one to do some cool stuff with MessageBox (which I don't show here) such as apply styles, change the text of the OK, CANCEL button, etc.
Any ideas would be much appreciated.
Thanks,
Dave
Extra info:
User1341210 suggestion works well if you just have one window. However, if you have a second window in it's own thread it doesn't work so well. Perhaps someone can tell me what I'm doing wrong. I use the suggestion of the TaskScheduler, but the code throws an exception if the TaskScheduler used is the one of the second window. That is, all works fine if I use the TaskScheduler of the first window, but throws an exception if I use the TaskScheduler of the second window. Here is the code behind for my second window:
public partial class AltThreadWindow : Window
{
private TaskScheduler _ui;
public AltThreadWindow()
{
InitializeComponent();
_ui = TaskScheduler.FromCurrentSynchronizationContext();
}
// This constructor is for passing in the TaskScheduler of the mainwindow and works great
public AltThreadWindow(TaskScheduler scheduler)
{
InitializeComponent();
_ui = scheduler;
}
private void btnWindowsMsgBox_Click(object sender, RoutedEventArgs e)
{
MessageBox.Show("Standard Windows message box");
}
private void btnCustomMsgBox_Click(object sender, RoutedEventArgs e)
{
MessageBoxResult result;
Task.Factory.StartNew(() => { result = CustomMessageBox.Show("Custom MessageBox on separate window"); }, CancellationToken.None,
TaskCreationOptions.None,
_ui);
}
}
Notice the two constructors. The default one assigns the TaskScheduler of the second window. The other constructor allows one to pas in the TaskScheduler of the main Window.
Here's the code I use to launch the second window from the main window. Again, I'm launching the second window on another thread, and I pass in the TaskScheduler of the main window. It would be nice to use the TaskScheduler of the second window instead.
_altWindowThread = new Thread(new ThreadStart(AltWinThreadProc));
_altWindowThread.SetApartmentState(ApartmentState.STA);
_altWindowThread.Priority = ThreadPriority.AboveNormal;
_altWindowThread.IsBackground = true;
_altWindowThread.Start();
And the actual threadproc:
[EnvironmentPermissionAttribute(SecurityAction.LinkDemand, Unrestricted = true)]
public void AltWinThreadProc()
{
// Create our context, and install it:
SynchronizationContext.SetSynchronizationContext(
new DispatcherSynchronizationContext(
Dispatcher.CurrentDispatcher));
_altWindow = new AltThreadWindow(_ui);
_altWindow.Show();
System.Windows.Threading.Dispatcher.Run();
}
Notice here I pass in the TaskScheduler of the MainWindow.
we had the same issue in our application (I created the work item on codeplex).
The error messages are quite confusing and I cant provide you an answer to that.
But:
We didn't used a separated WindowContainer to solve it. Instead came up with calling the separate task/thread with the UI scheduler:
Task.Factory.StartNew(
() => { result = CustomMessageBox.Show(messageText); },
CancellationToken.None,
TaskCreationOptions.None,
_ui);
Where _ui is assigned in a method that is executed from UI context (e.g. Constructor of your Window/Control:
_ui = TaskScheduler.FromCurrentSynchronizationContext();
Hope this helps for solving the "replace System.Windows.MessageBox with Xceed.Wpf.Toolkit.MessageBox" part of your question.
If you want that the messagebox shows up on another Window you have to set the "Owner" property of the message box to the other window.
Best regards.
As part of my App's startup procedure, it checks data integrity, and if it finds a problem it pops up a message to the user telling them that it might take a while to repair things.
I'm showing the message using MessageBox.Show. Because the data check is done from a worker thread, I'm switching over to the UI thread to make that call, and then setting a ManualResetEvent to tell the worker thread when the user has acknowledged the message.
I kick off the data check/load very early in the app's lifecycle from the constructor in the main Application class, by spinning off a worker thread (using the ThreadPool).
When I run with the debugger, and the message is displayed, the app just waits for input. When I run without the debugger, the app terminates after displaying the dialog for 10 seconds.
That 10 seconds is a big clue - it tells me that the OS thinks the app took too long to initialize (the OS kills apps that take too long to start up).
I think that my MessageBox.Show is blocking the UI thread before the App.RootFrameNavigating has a chance to be invoked.
My questions:
Does my diagnosis sound right?
I'd prefer to kick off my data load early, because it is almost entirely IO, except for this Message Box, and the sooner I can get my Model loaded, the better, but do you normally delay your data load until later in the app lifecycle?
Any other ideas/suggestions? I can't guarantee which page will be the start page, because the app could be resuming to any page. I'm also thinking of having the MessageBox.Show delay itself until the app has initialized, perhaps polling away for a flag set by App.RootFrameNavigating - does that make sense?
I think your problem is a result of kicking off the worker thread in the Application constructor. You should use the appropriate life-cycle event, in this case: PhoneApplicationService.Activated Event
So, the solution I've come up with is to still kick off the data load in a worker-thread from the Application's constructor, but in my PhoneService's class ShowDialog method that I invoke to invoke MessageBox.Show, I check to see if the initial navigation has occurred:
private readonly ManualResetEvent _appInitialized = new ManualResetEvent(false);
public void AppInitialized()
{
_appInitialized.Set();
}
public void ShowDialog(string caption, string text, Action<MessageBoxResult> callback, MessageBoxButton button = MessageBoxButton.OKCancel)
{
_appInitialized.WaitOne();
DispatcherHelper.CheckBeginInvokeOnUI(() =>
{
var result = MessageBox.Show(text, caption, button);
if (callback != null)
{
callback(result);
}
});
}
Then in my Application class:
private bool _firstNavigate = true;
private void RootFrameNavigating(object sender, NavigatingCancelEventArgs e)
{
if (_firstNavigate)
{
_firstNavigate = false;
var navigationService = (NavigationService) sender;
navigationService.Navigated += NavigationServiceNavigated;
}
....
private void NavigationServiceNavigated(object sender, NavigationEventArgs e)
{
var navigationService = (NavigationService)sender;
navigationService.Navigated -= NavigationServiceNavigated;
PhoneServices.Current.AppInitialized();
}
Anyone see any issues with this approach? Anyone come up with a better way?
In my WPF app (csharp) I have an event handler that when triggered will open a new window (window B) of the application and display some data. However, when the event is triggered again, if the new window (window B) is still open, I don't want to spawn another instance of window B but just update the data being displayed in the current instance. So the question is: How to detect if window B is already and only open if it is not already, otherwise just update the data?
I found the Application.Current.Window collection but somehow that isn't working for me yet. Ideas?
You could create a LoadWindow() method in WindowB that you can call to load (or refresh) the data & that will work regardless of if the window is already open or not. Have it take a delegate to call when this window gets closed:
private Action ParentCallbackOnClose;
public void LoadWindow( Action parentCallbackOnClose ) {
// load the data (set the DataContext or whatever)
ParentCallbackOnClose = parentCallbackOnClose;
// Open the window and activate/bring to the foreground
Show( );
Activate( );
}
and have your window closed event call the close delegate:
private void WindowClosed( object sender, EventArgs e ) {
ParentCallbackOnClose.Invoke( );
}
Now, from your class that opens Window B, have it hold onto that instance it opens, so that if WindowB is already open when someone tries to reload it, it just calls LoadWindow on the existing instance. Something like...
private WindowB WinB;
private void LoadWindowB(Content content)
{
if (WinB == null ){
WinB = new WindowB( );
}
WinB.LoadWindow(content, WindowBClosed);
}
And then you can just have it null out WinB on that close callback so if WinB is closed, then the next time LoadWindowB() is called it will create a new instance of it:
private void WindowBClosed( ){
WinB = null;
}
Since this is the first link Google listed, which posted several years ago, for a solution to check if a Window is already open, I'll post my answer, for others, which I find easier to implement. The ChildWindow is only called from MainWindow so no other Window will need to do any checks.
private void OpenChildWindow()
{
if (this.OwnedWindows.OfType<ChildWindow>().Count() > 0)
{
ChildWindow Win = this.OwnedWindows.OfType<ChildWindow>().First();
Win.Activate();
}
else
{
ChildWindow Win = new ChildWindow();
Win.Owner = this;
Win.Show();
}
}
There is an old school way to do this using an interface. I see this in Java a lot as a way to compensate for not having delegates (correct me if I am wrong). This method will allow you to check if there is a window already open (of any kind). The original response works very well, but you can also do it the following way:
Create the interface
public interface IWindowTracker
{
void WindowIsOpened();
void WindowIsClosed();
}
Implement the interface on the parent (from where you are opening):
public partial class MainWindow : Window, IWindowTracker
In your constructor, accept an object that is of the IwindowTracker interface. Save the instance for future use
IWindowTracker windowTracker;
public ProjectManager(IWindowTracker parentWindowTracker)
{
windowTracker = parentWindowTracker;
InitializeComponent();
}
Setup the calls to the window tracker object
protected override void OnActivated(EventArgs e)
{
windowTracker.WindowIsOpened();
base.OnActivated(e);
}
protected override void OnClosed(EventArgs e)
{
windowTracker.WindowIsClosed();
base.OnClosed(e);
}
and finally implement the IWindowTracker in your parent WPF window
bool windowIsOpen = false;
public void WindowIsOpened()
{
windowIsOpen = true;
}
public void WindowIsClosed()
{
windowIsOpen = false;
}
This will allow you to keep track of if the window is still open and if it is, there is no need to open a new instance of it:
if (!windowIsOpen)
{
remoteProjectManager = new ProjectManager(this);
remoteProjectManager.Show();
}
remoteProjectManager.Focus();
Calling show() on a closed window seems to throw an exception, so my guess is that there is some other way or that if you have closed the window, the window is technically "destroyed"
The nice thing to this is that I can detect if the window is still open and focus on it (so that it comes to the front again).
NOTE: There is a draw back to this, in that in this setup it limits you to opening only one window at a time (assuming that all your windows are implemented like this). In my case, I only ever want to have one window open besides the main window.
You might also want to check if your window is null or not, considering that it probably isn't the only window you will have to open.
edit: oops, my answer is specific to Windows Forms. i just now saw the WPF mention. i'm not sure what the specific code would be for WPF, but i would imagine that it's not all that different conceptually. I think in WPF the property is called IsVisible instead of Visible
You could hold on to the instance of your window (or make it a Singleton) and then when you need to determine if it is visible or not, check it's Visible property.
for example:
if(myWindow.Visible){
myWindow.Hide();
}else{
myWindow.Show();
}
This article it the best I found for passing data between WPF pages. The author used KISS approach to provide a simple solution.