So I've made this menu for my game in godot. I also made this window which shows up when a button is clicked. This window is WindowDialog in godot, and I want the window to be resizeable. However I get an error when I put the code for the window to be resizeable. Here is my code:
extends WindowDialog
var newSize
func _ready():
# Change 'hide()' to 'show()' to show the window.
hide()
set_process_input(true)
newSize = self.get_size()
pass
func _input(event):
if(event.type == InputEvent.MOUSE_MOTION):
if(Input.is_mouse_button_pressed(1)):
if (self.get_local_mouse_position().x > (self.get_size().x - 30)) && (self.get_local_mouse_position().y > (self.get_size().y - 30)) :
newSize += event.relative_pos
self.set_size(newSize)
print(self.get_size().x)
print(event.relative_x)
The error I get is this:
Invalid get index 'type' (on base: 'InputEventMouseMotion').
To answer the question the title, to make a WindowDialog that can be resized, set its resizable to true.
Now, the error. Do not check a type property, check the type of the event object, like this:
if event is InputEventMouseMotion:
pass
Or, assuming you want to read the properties of the InputEventMouseMotion, you can do this:
var mouse_motion_event := event as InputEventMouseMotion
if mouse_motion_event != null:
pass
Something else: You probably don't need to call set_process_input(true), at least not in _ready. If you have an _input function defined, Godot call it (and if it does not have _input, then it won't). Thus, the only reason to call set_process_input(true) is if you called set_process_input(false) before.
Related
I am writing an UI test case, in which I need to perform an action, and then on the current page, scroll the only UITableView to the bottom to check if specific text shows up inside the last cell in the UITableView.
Right now the only way I can think of is to scroll it using app.tables.cells.element(boundBy: 0).swipeUp(), but if there are too many cells, it doesn't scroll all the way to the bottom. And the number of cells in the UITableView is not always the same, I cannot swipe up more than once because there might be only one cell in the table.
One way you could go about this is by getting the last cell from the tableView. Then, run a while loop that scrolls and checks to see if the cell isHittable between each scroll. Once it's determined that isHittable == true, the element can then be asserted against.
https://developer.apple.com/documentation/xctest/xcuielement/1500561-ishittable
It would look something like this (Swift answer):
In your XCTestCase file, write a query to identify the table. Then, a subsequent query to identify the last cell.
let tableView = app.descendants(matching: .table).firstMatch
guard let lastCell = tableView.cells.allElementsBoundByIndex.last else { return }
Use a while loop to determine whether or not the cell isHittable/is on screen. Note: isHittable relies on the cell's userInteractionEnabled property being set to true
//Add in a count, so that the loop can escape if it's scrolled too many times
let MAX_SCROLLS = 10
var count = 0
while lastCell.isHittable == false && count < MAX_SCROLLS {
apps.swipeUp()
count += 1
}
Check the cell's text using the label property, and compare it against the expected text.
//If there is only one label within the cell
let textInLastCell = lastCell.descendants(matching: .staticText).firstMatch
XCTAssertTrue(textInLastCell.label == "Expected Text" && textInLastCell.isHittable)
Blaines answer lead me to dig a little bit more into this topic and I found a different solution that worked for me:
func testTheTest() {
let app = XCUIApplication()
app.launch()
// Opens a menu in my app which contains the table view
app.buttons["openMenu"].tap()
// Get a handle for the tableView
let listpagetableviewTable = app.tables["myTableView"]
// Get a handle for the not yet existing cell by its content text
let cell = listpagetableviewTable.staticTexts["This text is from the cell"]
// Swipe down until it is visible
while !cell.exists {
app.swipeUp()
}
// Interact with it when visible
cell.tap()
}
One thing I had to do for this in order to work is set isAccessibilityElement to true and also assign accessibilityLabel as a String to the table view so it can be queried by it within the test code.
This might not be best practice but for what I could see in my test it works very well. I don't know how it would work when the cell has no text, one might be able to reference the cell(which is not really directly referenced here) by an image view or something else. It's obviously missing the counter from Blaines answer but I left it out for simplicity reasons.
I've a really simple UI with a single NSPopUpButton. Its selectedIndex is bound to an int value ViewController.self.my_selection. As soon as I change the selection from the UI (i.e. a select the third item of the NSPopUpButton) I see that my_selection value changes. So far so good, what I'm trying to obtain is the opposed direction though. I want to change the my_selection value programmatically and see the NSPopUpButton selecting the item a the index that I've defined in my_selection. I erroneously supposed that behaviour was the default behaviour for bindings...
This is what I'm obtaining now:
NSPoPUpButton ---> select item at index 2 ----> my_selection becomes equal to 2
This is what I want to achieve (keeping also the previous behaviour)
my_selection ---> set value to 3----> NSPoPUpButton selected index = 3
Without a bit more info (see my comment) it's hard to see exactly what you're doing wrong. Here's how I got it working: First, create a simple class...
// Create a simple class
class Beatle: NSObject {
convenience init(name: String) {
self.init()
self.name = name
}
dynamic var name: String?
}
Then, in the AppDelegate I created a four-item array called beatles:
dynamic var beatles: [Beatle]?
func applicationDidFinishLaunching(aNotification: NSNotification) {
beatles = [Beatle(name: "John"),
Beatle(name: "Paul"),
Beatle(name: "Ringo"),
Beatle(name: "George")]
}
In Interface Builder I set things up so that this array provides the pop-up with its content:
This class also has a selectedIndex property that is bound to the pop-up button's selectedIndex binding - this property provides read-write access to the pop-up button's selection:
// Set the pop-up selection by changing the value of this variable.
// (Make sure you mark it as dynamic.)
dynamic var selectedIndex: Int = 0
I want to verify if an element is visible or not depending on its .hidden property but I don't find a valid way to do that using the new Xcode 7 UI test stuff.
I've tried with myelement.exists and myelement.hittable but they doesn't seems to work as I expected. I suppose they work on conjunction with the hidden property. An hidden element shouldn't exists and is not hittable... but this is not the current behaviour (I can understand the exists behaviour... but a hidden element should be not hittable IMO).
Is there another way to verify the "hidden" property value?
As of Xcode 7.1 Beta 3, UI Testing does not currently support validating the visibility of an element. I suggest filing a radar to bring the necessary attention to Apple.
Xcode 7.1 has fixed this issue. hittable now checks to see if the element is correct.
1) I am testing the UI with swift in Xcode 7.3 and I using both .hittable and .exists for testing whether a label is hidden or not and they both work. I test for 'true' and 'false' to make sure either way agree with the result.
I have a label whose static text is "Track Info" and set to be hidden when app is first loaded, then later on I press a button to show the label, and here is the result after the label is shown.
// test fails
let trackInfoLabel = app.staticTexts["Track info"]
XCTAssertEqual(trackInfoLabel.exists, true)
XCTAssertEqual(trackInfoLabel.hittable, true)
// test passes
XCTAssertEqual(trackInfoLabel.exists, false)
XCTAssertEqual(trackInfoLabel.hittable, false)
// test passes
let trackInfoLabel = app.staticTexts["Track Info"]
XCTAssertEqual(trackInfoLabel.exists, true)
XCTAssertEqual(trackInfoLabel.hittable, true)
// test fails
XCTAssertEqual(trackInfoLabel.exists, false)
XCTAssertEqual(trackInfoLabel.hittable, false)
Leter on when I press the button to hide the label, all the results turned opposite. This confirms that both properties (hittable and exists) works for label.hidden setting.
2) Another way to find out if an element is hidden, you can do is element.frame.size.width == 0 || element.frame.size.height == 0
XCUIElement.hittable works for me (in my particular test case where I am checking several UIButton elements for visibility) - quite sure it is not a right way to do it though
Next code worked for me.
At the end of the test you can past code as follow:
while ([app.staticTexts matchingIdentifier:#"accesibilityId"].count > 0) {
sleep(1);
}
I agree hittable doesn't always work for buttons (Swift 2.0, XCode 7.2)
I just discovered that if button is visible, you can find it ONLY among buttons, while if button is hidden, you can find it's tag in staticTexts as well!
XCTAssertFalse(app.buttons["Log out"].exists && app.staticTexts["Log out"].exists) // This button is visible (hidden = false)
XCTAssert(app.buttons["Log in"].exists && app.staticTexts["Log in"].exists) // This one is hidden
Hacky, but works for buttons. Apple should just introduce .hidden or .visible along .hittable and .exists
The syntax is now .isHittable:
isHittable only returns true if the element is already visible and
hittable onscreen. It returns false for an offscreen element in a
scrollable view, even if the element would be scrolled into a hittable
position by calling click(), tap(), or another hit-point-related
interaction method.
Using the .isHittable property will work because hidden elements are not visible or hittable on screen.
My solution is to add accessibilityIdentifier dynamicly
func someMethod() {
label.isHidden = true
label. accessibilityIdentifier = "isHidden"
}
func someOtherMethod {
label.isHidden = false
label. accessibilityIdentifier = "isVisible"
}
and the in UITest you can use it as
if app.staticTexts["MyLabel"].identifier == "isHidden" {
dosomething()
}
I have an array of buttons and an array of associated NSTextViews (embedded in NSScrollViews), all created programmatically. All this in a program (a calendar) that compiles and works fine as far as it does.
However, I would like clicking a button to cause the associated text field to behave as if I had clicked directly in it (cursor and focus ring appear and it accepts text). Right now, I have to click the button AND the text box before I begin entering text.
It appears that “makeFirstResponder” should do what I want, but it won’t compile as I am trying to do it.
Here is the relevant code:
(All in a single View Controller)
Global Declarations:
var arrayOfButtons:[NSButton] = []
var arrayOfFields: [NSTextView!] = []
var arrayOfWindows: [NSScrollView!] = []
Creation of array of fields:
var i = 0
var myLocalText: NSTextView! = NSTextView(…
var myLocalWindow: NSScrollView! = NSScrollView(…
for i = 0; i <= 6; i++ {
var myLocalText: NSTextView! = NSTextView(…
var myLocalWindow: NSScrollView! = NSScrollView(…
view.addSubview(myLocalWindow)
myLocalWindow.hasVerticalScroller = false
myLocalWindow.focusRingType = NSFocusRingType(rawValue: UInt(2))! // I didn’t get a focus ring until I did this
myLocalWindow.addSubview(myLocalText)
myLocalWindow.documentView = myLocalText
myLocalText.editable = true
myLocalText.selectable = true
myLocalText.verticallyResizable = true
self.arrayOfFields.append(myLocalText)
}
Routine that responds to clicking a button:
…
view.makeFirstResponder(arrayOfFields[tag]) // THIS IS THE SUBJECT OF MY QUESTION
// IT GIVES A COMPILER ERROR AS FOLLOWS: “Cannot invoke ‘makeFirstResponder’ with an argument list of type ’NSTextView’”
// as far as I can tell from the documentation, makeFirstResponder should accept an argument type of NSResponder, and NSTextView should inherit from that
In Mac OS X, makeFirstResponder is an NSWindow method, not an NSView method.
I noticed that in Windows, if you maximize a window you can not resize it until you un-maximized it again. This appears to be a normal behaviour, so I would like to remove my resize gripper when the window is maximised.
At the moment I can't find a property to detect if a window is maximized, and although I could add a boolean in my controller, it wouldn't necessarily catch requests to maximize from the OS.
So if you know of a reliable way to test if a window is maximized please let me know.
On a related note, I am using custom chrome, and when I maximize a window it overlaps the windows task bar. I can think of hacks to detect available screen size (using a transparent system chrome window), but it would be good to know of a better method.
Thanks
Rob
In your application (MXML) on the in the init method you ussually call on creationComplete:
<mx:WindowedApplication xmlns:mx="http://www.adobe.com/2006/mxml" layout="absolute"
creationComplete="init()" >
Add the following code:
this.addEventListener(NativeWindowDisplayStateEvent.DISPLAY_STATE_CHANGE, trackState);
the method looks like this:
public function trackState(event:NativeWindowDisplayStateEvent):void
{
if (event.afterDisplayState == NativeWindowDisplayState.MAXIMIZED)
{
isMaximised = true;
} else {
isMaximised = false;
}
}
I have figured out how this can best be done thanks to some pointers from TheBrain.
Firstly you need to watch for resize events to the window your want to control:
NativeApplication.nativeApplication.activeWindow.addEventListener(NativeWindowBoundsEvent.RESIZE, onWindowResize);
Then handle that event to decide if the window is maximised or not:
public function onWindowResize(event:NativeWindowBoundsEvent):void
{
if (event.afterBounds.height >= Screen.mainScreen.visibleBounds.height && event.afterBounds.width >= Screen.mainScreen.visibleBounds.width)
isMaximised = true;
else
isMaximised = false;
}
You then need to catch or create your own maximize button, and when clicked perform the following code:
if (isMaximised)
{
var bounds:Rectangle = Screen.mainScreen.visibleBounds;
NativeApplication.nativeApplication.activeWindow.bounds = bounds;
}
else
{
NativeApplication.nativeApplication.activeWindow.bounds = new Rectangle(100, 100, 500, 600);
}
You can modify the bounds to over maximize (which is handy for custom chrome windows with shadows), and you can also set the application to reset to a default size if the maximize button is clicked when it's already maximized (or do nothing).
I had issues about when to assign the window resize listner, and ended up removing and adding it every time the maximize button was clicked. It's a bit of overkill, but not too bad.
There is Win32 API Call that will do this for you:
BOOL IsZoomed( HWND hWnd );
to get the actual usable space from the screen, use the flash.display.Screen class, or you can use the systemMaxSize() which returns the largest window size allowed by the OS. For maximization you have some events that the window is dispaching when maximized/minimized/restored. You can find more info on the adobe pages (the link under systemMaxSize).
To detect if window is maximized...I don't think there is such a function (I might be wrong) but you can test if the app size is equal with the available screen size which means it's maximized. Or hook on the resize event which is triggered when the app is maximized/minimized/resized
Here is an easier way of checking if a window is maximized:
if(stage.nativeWindow.displayState == NativeWindowDisplayState.MAXIMIZED)
{
//do something
}
The following worked for me. No need to set event listeners, this code can be used to check the real-time state of the native window:
if (nativeWindow.displayState == 'maximized')
{
trace('Maximized');
}
else
{
trace('Minimized');
}
Can you use something like this to hook the maximize() event?