Done button not firing Completed event on Xamarin Entry - c#

After I've added the Done button on iOS numeric keyboard in Xamarin Forms, I encountered another problem: the Done button not firing Completed event (like return button does).
On my way to implement this, I found the following code on Xamarin Forums:
using System;
using System.Drawing;
using System.Reflection;
using Xamarin.Forms.Platform.iOS;
using Xamarin.Forms;
using UIKit;
using KeyboardTest.iOS;
using KeyboardTest;
[assembly: ExportRenderer(typeof(EntryDone), typeof(EntryDoneRenderer))]
namespace KeyboardTest.iOS
{
public class EntryDoneRenderer : EntryRenderer
{
// used to cache the MethodInfo so we only have the reflection hit once
private MethodInfo baseEntrySendCompleted = null;
public EntryDoneRenderer ()
{
}
protected override void OnElementChanged(ElementChangedEventArgs<Entry> e)
{
base.OnElementChanged (e);
if (this.Element.Keyboard == Keyboard.Numeric)
{
// only want the Done on the numeric keyboard type
UIToolbar toolbar = new UIToolbar (new RectangleF (0.0f, 0.0f, (float)Control.Frame.Size.Width, 44.0f));
var doneButton = new UIBarButtonItem (UIBarButtonSystemItem.Done, delegate {
this.Control.ResignFirstResponder ();
Type baseEntry = this.Element.GetType();
if(baseEntrySendCompleted==null)
{
// use reflection to find our method
baseEntrySendCompleted = baseEntry.GetMethod("SendCompleted",BindingFlags.NonPublic|BindingFlags.Instance);
}
try
{
baseEntrySendCompleted.Invoke(this.Element,null);
}
catch (Exception ex)
{
// handle the invoke error condition
}
});
toolbar.Items = new UIBarButtonItem[] {
new UIBarButtonItem (UIBarButtonSystemItem.FlexibleSpace),
doneButton
};
this.Control.InputAccessoryView = toolbar;
}
}
}
}
I don't know why, but I receive the error:
System.NullReferenceException: Object reference not set to an instance of an object
at myGame.iOS.DoneEntryRenderer.<OnElementChanged>m__0 (System.Object , System.EventArgs ) [0x0005d] in /Users/silviu/Projects/myGame/iOS/DoneEntry.cs:37
On that line I have the code:
baseEntrySendCompleted.Invoke(this.Element, null);
I’ve tried to debug the problem and I found that the SendCompleted method does not exist, but I don’t understand how to solve this problem in the lastest version of Xamarin, because I think on the moment when that guy posted the code, worked.
Thanks!

SendCompleted() was actually added for IEntryController so you don't need to use reflection for this any longer. In fact it appears that way no longer works. Just call SendCompleted() directly from your button press like so.
protected override void OnElementChanged(ElementChangedEventArgs<Entry> e)
{
base.OnElementChanged(e);
var toolbar = new UIToolbar(new CGRect(0.0f, 0.0f, Control.Frame.Size.Width, 44.0f));
toolbar.Items = new[]
{
new UIBarButtonItem(UIBarButtonSystemItem.FlexibleSpace),
new UIBarButtonItem(UIBarButtonSystemItem.Done, delegate {
Control.ResignFirstResponder();
((IEntryController)Element).SendCompleted();
})
};
this.Control.InputAccessoryView = toolbar;
}

Related

Programmatically click a button in Xamarin.Forms

I have 2 buttons on a xamarin form,
scannerButton and checkOrderButton
scannerButton opens the scanner page, scans a QRCode and populates it into a order entry field
checkOrderButton reads whatever is in the order entry field and processes validations and sends it to server for verification
what I want - is to call the checkOrderButton.Click from within the scannerButton.Click - after it has scanned the text
code:
private async void scanCameraButton_Clicked(object sender, EventArgs e)
{
var options = new ZXing.Mobile.MobileBarcodeScanningOptions();
options.PossibleFormats = new List<ZXing.BarcodeFormat>() {
ZXing.BarcodeFormat.QR_CODE,ZXing.BarcodeFormat.EAN_8, ZXing.BarcodeFormat.EAN_13
};
var scanPage = new ZXingScannerPage(options);
scanPage.OnScanResult += (result) =>
{
//stop scan
scanPage.IsScanning = false;
Device.BeginInvokeOnMainThread(() =>
{
//pop the page and get the result
Navigation.PopAsync();
orderNoEntry.Text = result.Text;
});
//invoke checkOrderButton.Click here
};
what would be the best approach to do this?
one alternate is to dump all the functionality from checkOrderButton.Click handler into a function and then call that function from both button clicks, but I'm interested in learning how I can invoke the click event programmatically
What I would do is having a viewmodel with a command that performs whatever logic would be done when pressing the button .
Then bind the Command property of the button to the command property in the ViewModel.
At this stage you will have a command that you can execute programmatically just as if you called "Button.Click()" if there will be such thing .
I wrote an extension method for this purposes. This not only works in Xamarin-Forms-projects but also in WPF projects.
using System;
using System.Reflection;
using System.Windows;
using System.Windows.Input;
#if !NETSTANDARD
using System.Windows.Controls;
#else
using Xamarin.Forms;
#endif
public static class ButtonExtensions
{
public static void PerformClick(this Button sourceButton)
{
// Check parameters
if (sourceButton == null)
throw new ArgumentNullException(nameof(sourceButton));
// 1.) Raise the Click-event
#if !NETSTANDARD
sourceButton.RaiseEvent(new RoutedEventArgs(System.Windows.Controls.Primitives.ButtonBase.ClickEvent));
#else
sourceButton.RaiseEventViaReflection(nameof(sourceButton.Clicked), EventArgs.Empty);
#endif
// 2.) Execute the command, if bound and can be executed
ICommand boundCommand = sourceButton.Command;
if (boundCommand != null)
{
object parameter = sourceButton.CommandParameter;
if (boundCommand.CanExecute(parameter) == true)
boundCommand.Execute(parameter);
}
}
#if NETSTD
private static void RaiseEventViaReflection<TEventArgs>(this object source, string eventName, TEventArgs eventArgs) where TEventArgs : EventArgs
{
var eventDelegate = (MulticastDelegate)source.GetType().GetField(eventName, BindingFlags.Instance | BindingFlags.NonPublic).GetValue(source);
if (eventDelegate == null)
return;
foreach (var handler in eventDelegate.GetInvocationList())
{
#if !(NETSTANDARD1_6 || NETSTANDARD1_5 || NETSTANDARD1_4 || NETSTANDARD1_3 || NETSTANDARD1_2 || NETSTANDARD1_1 || NETSTANDARD1_0)
handler.Method?.Invoke(handler.Target, new object[] { source, eventArgs });
#else
handler.GetMethodInfo()?.Invoke(handler.Target, new object[] { source, eventArgs });
#endif
}
}
#endif
}
In fact you should pass two parameters, so it's not just supposed to be Button_Click(), you should call it like this: Button_Click(null, null) since it needs sender and e as two required parameters, take a look at the method definition: Button_Click(object sender, EventArgs e){..}
As Sami already mentions I would suggest to extract the functionality of the checkOrderButton.Click to a separate method so you can call that some method from the checkOrderButton.Click as wel as the scannerButton.Click.

NEW TITLE -- The application called an interface that was marshalled for a different thread

After making the project simpler, I believe I identified the problem is actually a result the async marshalling.
UPDATE: I made the code simpler to try to figure out what was going on. So here is an update... The Observable collection is being populated on a new thread (async method). I tried moving the assigning of the ItemsSource to after the ObservableCollection is loaded as seen below
async void LoadAllData(object sender, EventArgs e)
{
if (sender != null)
{
App.GeoLocationComplete -= LoadAllData;
}
await ViewModelObjects.NearbyLocations.LoadLocationData();
lvPlaces.ItemsSource = ViewModelObjects.NearbyLocations.GBSLocationDetails;
}
The definition for the data load method is a follows:
public async Task LoadLocationData()
{....}
When I run this code I get the following error:
The application called an interface that was marshalled for a different thread. (Exception from HRESULT: 0x8001010E (RPC_E_WRONG_THREAD))
I know what is causing the error (the data was loaded on a thread other than the UI thread) but I don't know how to fix it. Suggestions?
UPDATE UPDATE: So I believe I have identified the root cause of the problem but have not figured out how to fix it. I started by simplifying my code as follows and it worked.
public nearbyplaces()
{
InitializeComponent();
NavigationPage.SetHasNavigationBar(this, false);
LoadAllData(null, null);
}
void LoadAllData(object sender, EventArgs e)
{
lobj_Places = new ObservableCollection<GBSLocationDetail>()
{
new GBSLocationDetail()
{
Title = "Location 1",
Distance = "20 Miles",
AddInfo = "Something Else",
AttributesTexts="Gay, Bar, Dance"
}
};
lvPlaces.ItemsSource = lobj_Places;
}
HOWEVER, what I need is for the LoadAllData method to be called once I have the GPS location from the device. So in my App.XAML.cs I have the following delegate event declared:
public static Plugin.Geolocator.Abstractions.IGeolocator gobj_RealGeoCoordinator;
public static event GeoLocationCompleteEventHandler GeoLocationComplete;
public static bool gb_WaitingForLocation = true;
Then I have the following code call the event once I get the location back from the device:
private async void ProcessStartupandResume()
{
if (gobj_RealGeoCoordinator == null)
{
gobj_RealGeoCoordinator = CrossGeolocator.Current;
ViewModelObjects.AppSettings.CanAccessLocation = App.gobj_RealGeoCoordinator.IsGeolocationEnabled;
if (!ViewModelObjects.AppSettings.CanAccessLocation)
{
await MainPage.DisplayAlert(ResourceStrings.GetValue("NoLocationServicesTitle"), ResourceStrings.GetValue("NoLocationServicesMessage"), ResourceStrings.GetValue("OKButtonText"));
}
//Only add the events if the object has to be created.
gobj_RealGeoCoordinator.PositionChanged += gobj_RealGeoCoordinator_PositionChanged;
gobj_RealGeoCoordinator.PositionError += (sender, e) =>
{
ProcessException(new Exception(e.Error.ToString()));
};
}
//Set this to null to trigger the first check
ib_GPSReenabled = null;
if (gobj_RealGeoCoordinator.IsListening)
await gobj_RealGeoCoordinator.StopListeningAsync();
gobj_RealGeoCoordinator.DesiredAccuracy = 50;
await gobj_RealGeoCoordinator.StartListeningAsync(10000, 20);
}
private static void gobj_RealGeoCoordinator_PositionChanged(object sender, PositionEventArgs e)
{
var pos = e.Position;
ViewModelObjects.AppSettings.Latitude = pos.Latitude;
ViewModelObjects.AppSettings.Longitude = pos.Longitude;
if (gb_WaitingForLocation)
{
gb_WaitingForLocation = false;
GeoLocationComplete?.Invoke(new object() , null);
}
}
Then in my page I subscribe to the GeoLocationComplete event using the LoadAllData method as seen below. Even when I use a local object and try to set the ItemsSource for the ListView in the code when executed as a result of the event being raised, I receive the error. See code below which subscribed to the event:
public nearbyplaces()
{
InitializeComponent();
NavigationPage.SetHasNavigationBar(this, false);
if (App.gb_WaitingForLocation)
App.GeoLocationComplete += LoadAllData;
else
LoadAllData(null, null);
}
Any suggestions on how to fix this?
OK so I figured it out. I needed to invoke the event on the main thread and I did that with the following code:
Device.BeginInvokeOnMainThread(() =>
{
GeoLocationComplete?.Invoke(new object(), null);
});
After inserting this code, the error was gone. Changing the code back to simply
GeoLocationComplete?.Invoke(new object(), null);
cause the error to occur again. Thus I believe this resolved my problem. Hope this helps someone else. :)

MonoTouch memory issue with Threading / Timers

I am having an issue with MonoTouch where UIViewControllers are remaining in-memory forever, even after they have been popped from the navigation stack.
I have a UINavigationController which contains a UIViewController with a button. Clicking the button pushes a custom UIViewController called ThreadingViewController onto the navigation stack.
The ThreadingViewController uses a NSTimer.CreateRepeatingScheduledTimer and a Thread to update the text of a label every one second.
When the user clicks "back" to pop back to the root view, the Mono Profiler says that my ThreadingViewController still exists in memory. The Profiler tells me it has something to do with NSAction and/or ThreadStart which has a reference to the ThreadingViewController, keeping it alive. I can see this by checking the "inverse references" checkbox in the profiler.
This means that if the user clicks backwards and forwards 100 times between the root ViewController and the custom ThreadingViewController, there will be 100 instances of this ViewController in memory. It isn't being garbage collected.
In ViewDidDisappear I have tried aborting the thread, setting it to null, to no avail.
What do I need to do to get this ThreadingViewController to be properly cleaned up / GC'ed by MonoTouch?
Here is the full (C# only) source code to reproduce the problem:
using System;
using System.Collections.Generic;
using System.Drawing;
using System.Linq;
using MonoTouch.Foundation;
using MonoTouch.UIKit;
using System.Threading;
namespace MemoryTests
{
[Register("AppDelegate")]
public partial class AppDelegate : UIApplicationDelegate
{
private UIWindow window;
private UINavigationController rootNavigationController;
private RootScreen rootScreen;
public override bool FinishedLaunching(UIApplication app, NSDictionary options)
{
window = new UIWindow(UIScreen.MainScreen.Bounds);
rootScreen = new RootScreen();
rootNavigationController = new UINavigationController(rootScreen);
window.RootViewController = rootNavigationController;
window.MakeKeyAndVisible();
return true;
}
}
public class RootScreen : UIViewController
{
private UIButton button;
public override void ViewDidLoad()
{
base.ViewDidLoad();
this.Title = "Root Screen";
// Add a button
button = new UIButton(UIButtonType.RoundedRect);
button.SetTitle("Click me", UIControlState.Normal);
button.Frame = new RectangleF(100, 100, 120, 44);
this.View.Add(button);
}
public override void ViewWillAppear(bool animated)
{
base.ViewWillAppear(animated);
button.TouchUpInside += PushThreadingViewController;
}
public override void ViewDidDisappear(bool animated)
{
base.ViewDidDisappear(animated);
button.TouchUpInside -= PushThreadingViewController;
}
private void PushThreadingViewController(object sender, EventArgs e)
{
var threadingViewController = new ThreadingViewController();
NavigationController.PushViewController(threadingViewController, true);
}
}
public class ThreadingViewController : UIViewController
{
private UILabel label;
private NSTimer timer;
private int counter;
public override void ViewDidLoad()
{
base.ViewDidLoad();
this.Title = "Threading Screen";
// Add a label
label = new UILabel();
label.Frame = new RectangleF(0f, 200f, 320f, 44f);
label.Text = "Count: 0";
this.View.Add(label);
// Start a timer
var timerThread = new Thread(StartTimer as ThreadStart);
timerThread.Start();
}
public override void ViewDidDisappear (bool animated)
{
base.ViewDidDisappear(animated);
timer.Dispose();
timer = null;
// Do I need to clean up more Threading things here?
}
[Export("StartTimer")]
private void StartTimer()
{
using (var pool = new NSAutoreleasePool())
{
timer = NSTimer.CreateRepeatingScheduledTimer(1d, TimerTicked);
NSRunLoop.Current.Run();
}
}
private void TimerTicked()
{
InvokeOnMainThread(() => {
label.Text = "Count: " + counter;
counter++;
});
}
}
}
Here is a screenshot of the profiler telling me that we have 3 instances of ThreadingViewController in memory:
Cheers.
I think the problem is here:
timer.Dispose();
Try changing it to:
timer.Invalidate();
NSTimer is the only thing here that uses an NSAction.
Also, I don't think this is very helpful:
var timerThread = new Thread(StartTimer as ThreadStart);
timerThread.Start();
Reasons:
It's not a background thread. Not sure how foreground .NET threads behave on iOS. My guess is, it never finishes.
You don't need to have a separate thread live all the time, within from you are starting the NSTimer. NSTimers are light, bulletproof and provide all the options you need. They work great even if you are starting them on the UI thread.
(Please ignore #2 above if you have another reason you are doing it this way, which is not directly visible from your code. But, nothing else comes to mind).
PS: I haven't tested your code above. But I'm 100% sure that you must Invalidate() NSTimers so that they stop running, when you no longer need them.
Anytime you connect a click handler such as:
button.TouchUpInside += PushThreadingViewController;
You must also disconnect it. I would suggest moving the line of code above to the ViewDidAppear override, and adding the line of code below to the ViewDidDisappear override:
button.TouchUpInside -= PushThreadingViewController;
I would also recommend that you check to see if threadingViewController is null and only creating a new instance if it is, before pushing it onto the navigation stack. That way it will use the same one each time if it hasn't been garbage collected yet.
Be careful with ViewDidDisappear. This method may called at any time your View-Controller yields control to another View-Controller (calling PresentViewController, invoking the sharing UI, display an iAd etc). If you only want to react when your View-Controller is removed from the navigation stack, I'd recommend this:
public override void DidMoveToParentViewController(UIViewController parent)
{
base.DidMoveToParentViewController(parent);
if(parent == null)
Cleanup();
}
Try invoking ReleaseOutlets() (generated from the Xamarin Studio XIB processor) from within your Cleanup() method in addition to unwiring the touch event handler improves the situation.

Playing two video with axWindowsMediaPlayer

I try to play jpg (in loop), after click mp4 should be played after end, that jpg should play again. I dont know why but after I play in axWindowsMediaPlayer1_PlayStateChange vido play and then stop. Help.
using System;
using System.Collections.Generic;
using System.ComponentModel;
using System.Data;
using System.Drawing;
using System.Linq;
using System.Text;
using System.Windows.Forms;
namespace Video
{
public partial class Form1 : Form
{
bool clicked = false;
public Form1()
{
InitializeComponent();
axWindowsMediaPlayer1.URL = "wait2.JPG";
}
private void axWindowsMediaPlayer1_PlayStateChange(object sender, AxWMPLib._WMPOCXEvents_PlayStateChangeEvent e)
{
if (axWindowsMediaPlayer1.playState == WMPLib.WMPPlayState.wmppsMediaEnded & clicked== true)
{
clicked = false;
axWindowsMediaPlayer1.settings.setMode("Loop", true);
axWindowsMediaPlayer1.URL = "wait2.JPG";
axWindowsMediaPlayer1.Ctlcontrols.play();
}
}
private void axWindowsMediaPlayer1_ClickEvent(object sender, AxWMPLib._WMPOCXEvents_ClickEvent e)
{
axWindowsMediaPlayer1.settings.setMode("Loop", false);
axWindowsMediaPlayer1.URL = "video.MP4";
axWindowsMediaPlayer1.Ctlcontrols.play();
clicked = true;
}
}
}
I wish someone had replied to this question the time it was posted. It took me a lot of time to figure out why I was not able to start a new video by setting the URL property. I finally found the answer to this issue here:
http://msdn.microsoft.com/en-us/library/windows/desktop/dd562470%28v=vs.85%29.aspx
The problem is with setting the URL property from within the axWindowsMediaPlayer1_PlayStateChange() event handler. According to the above msdn document:
"Do not call this method from event handler code. Calling URL from an event handler may yield unexpected results."
So the URL property has to be set outside of the even handler. I also tried Dispatcher.Invoke() and even starting a new thread from within the event handler to set the URL property; but that too did not help. It really has to come from outside of the event handler!

.NET custom form designer: How to implement IMenuCommandService?

I've got a report form designer written long ago for a database project. It used a lot of winapi magic hence i was forced to rewrite some parts 'in proper way'.
Thanks to some articles from MSDN magazine (here and here) and CodeProject i was able to implement designer surface, toolbox and undo/redo engine.
Every resource i discovered on the topic so far is a bit outdated. Can you point to fresh/comprehensive article?
Code from article mentioned above seems not working.
MenuCommandService.ShowContextMenu is called but nothing appears since there aren't any DesignerVerbs in globalVerbs collection. Should i add 'standard' ones, corresponded to designer actions such as cut/paste, manually? If yes, how can i accomplish this?
Thanks to SharpDevelop sources i was able to figure out the solution
This minimal implementation (some standart commands, no more) worked for me
using System;
using System.Collections.Generic;
using System.ComponentModel.Design;
using System.Diagnostics;
using System.Windows.Forms;
using System.Drawing;
namespace DesignerHost
{
class MenuCommandServiceImpl : MenuCommandService
{
DesignerVerbCollection m_globalVerbs = new DesignerVerbCollection();
public MenuCommandServiceImpl(IServiceProvider serviceProvider)
: base(serviceProvider)
{
m_globalVerbs.Add(StandartVerb("Cut", StandardCommands.Cut));
m_globalVerbs.Add(StandartVerb("Copy", StandardCommands.Copy));
m_globalVerbs.Add(StandartVerb("Paste", StandardCommands.Paste));
m_globalVerbs.Add(StandartVerb("Delete", StandardCommands.Delete));
m_globalVerbs.Add(StandartVerb("Select All", StandardCommands.SelectAll));
}
private DesignerVerb StandartVerb(string text, CommandID commandID)
{
return new DesignerVerb(text,
delegate(object o, EventArgs e)
{
IMenuCommandService ms =
GetService(typeof(IMenuCommandService)) as IMenuCommandService;
Debug.Assert(ms != null);
ms.GlobalInvoke(commandID);
}
);
}
class MenuItem : ToolStripMenuItem
{
DesignerVerb verb;
public MenuItem(DesignerVerb verb)
: base(verb.Text)
{
Enabled = verb.Enabled;
this.verb = verb;
Click += InvokeCommand;
}
void InvokeCommand(object sender, EventArgs e)
{
try
{
verb.Invoke();
}
catch (Exception ex)
{
Trace.Write("MenuCommandServiceImpl: " + ex.ToString());
}
}
}
private ToolStripItem[] BuildMenuItems()
{
List<ToolStripItem> items = new List<ToolStripItem>();
foreach (DesignerVerb verb in m_globalVerbs)
{
items.Add(new MenuItem(verb));
}
return items.ToArray();
}
#region IMenuCommandService Members
/// This is called whenever the user right-clicks on a designer.
public override void ShowContextMenu(CommandID menuID, int x, int y)
{
// Display our ContextMenu! Note that the coordinate parameters to this method
// are in screen coordinates, so we've got to translate them into client coordinates.
ContextMenuStrip cm = new ContextMenuStrip();
cm.Items.AddRange(BuildMenuItems());
ISelectionService ss = GetService(typeof (ISelectionService)) as ISelectionService;
Debug.Assert(ss != null);
Control ps = ss.PrimarySelection as Control;
Debug.Assert(ps != null);
Point s = ps.PointToScreen(new Point(0, 0));
cm.Show(ps, new Point(x - s.X, y - s.Y));
}
#endregion
}
}
Update
found similar solution
I ran into same problem a month ago and got something really intersting and helpful from CodePlex here. I read about this on Brad Abram's Blog Post. These includes many examples on framework extensibility and Custom Windows Form Designer Interface is one of them.
Link Txt 1: http://mef.codeplex.com/
Link Txt 2: http://blogs.msdn.com/brada/archive/2009/04/13/managed-extensibility-framework-preview-5-released.aspx
Thanks.

Categories