I have a couple of MovieClips in swf files loaded with Loader.loadBytes. They are 10-th version, as3-enabled and have scale9Grid defined (I can see the tag DefineScaleGrid in them with SWiX). But when loaded, loader.content has scale9Grid == null (maybe it is normal?) and scaling is not 9-slice. I can't even set scale9Grid myself - any attempt is met with exception "One of the parameters is invalid". How to make 9-slice grid work?
I finally found this article to explain it all. So, in short, scale9Grid works only on DisplayObjects without children. I dug down to Shape in my clip and managed to set that grid.
Related
I'm importing someone's createjs application into an application loader which imports each developers application javascript to a global canvas. The application is working standalone but when I load the javascript dynamically I'm getting some strange behavior surrounding the getBounds() function.
I have some text created like this:
dialogText = new createjs.Text(text, 'bold ' + (28 * scale) + 'px Calibri', '#FFFFFF');
Then later on I use dialogText.getBounds() which returns null.
I try console logging dialogText and I can see that it is set and there is a value _bounds which reads null.
At what point do these values get set when the element in drawn to the canvas?
Note: I'm already adding the text to a createjs.Container() which has its bounds set before I run getBounds().
For a quick reference, I pasted EaselJS's update regarding getBounds():
(This dates back from "August 21, 2013", not sure if this is still necessary nowadays, but using setBounds(x, y, w, h) helped me with Shapes today [in 2017] actually!)
Source: http://blog.createjs.com/update-width-height-in-easeljs/
The _bounds property reflects a value that is manually set on the DisplayObject - which is helpful if you know the bounds of something that can't calculate it (like a Shape). It will always be null, unless you set it directly.
I tested your code, and it seems fine. Are you sure the scale or text values are not null? Maybe hard-code them to see if it works.
Do you have a sample somewhere? I noticed in the GitHub issue you posted that you mentioned it worked fine without RequireJS (though you don't mention that here). I would definitely ensure that the values are being set properly.
Here is the fiddle, as well as a simpler one with no EaselJS stage. Note that I have Calibri installed on my machine, so you might get different results if you are loading it in.
Sample code:
var dialogText = new createjs.Text("This is some text", 'bold ' + (28 * 1.5) + 'px Calibrix', '#FFFFFF');
console.log(">>",dialogText.getBounds());
I'm trying to check if a value is contained in the innertext of a webelement but I'm having a little problem: frames seem to change at every refresh of the pages.
These are the steps I've recorded:
Browser("SystemPage").Page("SystemP").Frame("dme2_header").Image("Gestione Anagrafiche").Click<br>
Browser("SystemPage").Page("SystemP").Frame("dme2_appl").WebEdit("pdrBean.pdrPod").Set parameter("POD")<br>
Browser("SystemPage").Page("SystemP").Frame("dme2_appl").WebButton("Cerca").Click
Browser("SystemPage").Page("SystemP").Frame("dme2_appl_2").Image("show_files").Click
Browser("SystemPage").Page("SystemP").Frame("dme2_appl_6").Image("Lente").Click
cctype = Browser("SystemPage").Page("SystemP").Frame("dme2_appl_7").WebElement("arrow_down").GetROProperty("innertext")<br>
DataAct = Browser("SystemPage").Page("SystemP").Frame("dme2_appl_7").WebElement("arrow_down_2").GetROProperty("innertext")<br>
Browser("SystemPage").Page("SystemP").Frame("dme2_header").Image("Gestione Anagrafiche").Click
The frames "dme2_appl6" and "dme2_appl7" changes at every refresh of the two pages.
The question is simple: how can I rewrite these two actions to make them universal?
I've tried to do something like:
Set objFrame = Frame("title:=dme2_appl_.")
and then
Browser("SystemPage").Page("SystemP").objFrame.Image("Lente").Click
But QTP gives me an error in return: "property or method not supported by the object"
Please try using the below code
Browser("SystemPage").Page("SystemP").Image("Lente").Click
Avoid using the "Frame" and if you really want to use it, put regex for name property of Frame.
Go to Object Properties--> Click on the Frame object --> Mandatory Properties--> Change name property as
like iFrame_213123123 to i.*
Hope this will solve your problem.
I don't think you can use a frame object that way. When you write Page().Frame() UFT sets the frame's parent in different way than if you first create the Frame.
Think of Frame() as a function that behaves differently when called on a Page or as a free function.
Try:
Browser("SystemPage").Page("SystemP").Frame("title:=dme2_appl_.")
I am creating a script that involved searching for a record and then updating the record. On the search screen, the user has the option of viewing advanced search options. To toggle showing or hiding advanced search is controlled by one button.
<a title="Searches" href="javascript:expandFilters()"><img border="0" align="absmiddle" alt="Advanced" src="****MASKED URL****"></a>
The only difference between the properties of the search button when it is showing or hiding the advanced search is the img src:
When advanced search is hidden the IMG src ends with "/Styles/_Images/advanced_button.jpg", when advanced search is visible, the IMG src ends with "/Styles/_Images/basic_button.png"
When I open the page, sometimes the Advanced search options are showing, sometimes they aren't. The value that I want to search on appears in the Advanced section, so for my script to work I have added an IF statement.
<input type="text" value="" maxlength="30" size="30" name="guiSystemID">
The IF statement looks for the fields that I need to enter data into, and if the field does not exist then that would indicate that the Advanced options are not visible I need to click on the button to expand the search option.
I created the following IF statement.
if (!driver.findElement(By.name("guiSystemID")).isDisplayed()) {
driver.findElement(By.cssSelector("img[alt='Advanced']")).click();
}
When I run the script and the Advanced search is expanded then the script runs successfully. However, when I run the script and the Advanced search is not expanded, the script fails, advising me that it could not find the object "guiSystemID". This is frustrating because if it can't find it then I want the script to continue, entering into the True path of the IF statement.
Has anyone got any suggestions about how else I could assess if the field is appearing without having the script fail because it can't find the field.
Thanks in advance
Simon
I might be late in answering this, but it might help someone else looking for the same.
I recently faced a similar problem while working with isDisplayed(). My code was something like this
if(driver.findElement(By.xpath(noRecordId)).isDisplayed() )
{
/**Do this*/
}
else
{
/**Do this*/
}
This code works pretty well when the element that isDisplayed is trying to find is present. But when the element is absent, it continues looking for that and hence throws an exception "NosuchElementFound". So there was no way that I could test the else part.
I figured out a way to work with this(Surround the {if, else} with try and catch block, say something like this.
public void deleteSubVar() throws Exception
{
try
{
if(driver.findElement(By.xpath(noRecordId)).isDisplayed() )
{
/**when the element is found do this*/
}
}
catch(Exception e)
{
/**include the else part here*/
}
}
Hope this helps :)
I've had mixed results with .isDisplayed() in the past. Since there are various methods to hide an element on the DOM, I think it boils down to a flexibility issue with isDisplayed(). I tend to come up with my own solutions to this. I'll share a couple things I do, then make a recommendation for your scenario.
Unless I have something very specific, I tend to use a wrapper method that performs a number of checks for visibility. Here's the concept, I'll leave the actual implementation approach to you. For general examples here, just assume "locator" is your chosen method of location (CSS, XPath, Name, ID, etc).
The first, and easiest check to make is to see if the element is even present on the DOM. If it's not present, it certainly isn't visible.
boolean isPresent = driver.findElements(locator).size() > 0;
Then, if that returns true, I'll check the dimensions of the element:
Dimension d = driver.findElement(locator).getSize();
boolean isVisible = (d.getHeight() > 0 && d.getWidth() > 0);
Now, dimensions, at times, can return a false positive if the element does in fact have height and width greater than zero, but, for example, another element covers the target element, making it appear hidden on the page (at least, I've encountered this a few times in the past). So, as a final check (if the dimension check returns true), I look at the style attribute of the element (if one has been defined) and set the value of a boolean accordingly:
String elementStyle = driver.findElement(locator).getAttribute("style");
boolean isVisible = !(elementStyle.equals("display: none;") || elementStyle.equals("visibility: hidden;"));
These work for a majority of element visibility scenarios I encounter, but there are times where your front end dev does something different that needs to be handled on it's own.
An easy scenario is when there's a CSS class that defines element visibility. It could be named anything, so let's assume "hidden" to be what we need to look for. In this case, a simple check of the 'class' attribute should yield suitable results (if any of the above approaches fail to do so):
boolean isHidden = driver.findElement(locator).getAttribute("class").contains("hidden");
Now, for your particular situation, based on the information you've given above, I'd recommend setting a boolean value based on evaluation of the "src" attribute. This would be a similar approach to the CSS class check just above, but used in a slightly different context, since we know exactly what attribute changes between the two states. Note that this would only work in this fashion if there are two states of the element (Advanced and Basic, as you've noted). If there are more states, I'd look into setting an enum value or something of the like. So, assuming the element represents either Advanced or Basic:
boolean isAdvanced = driver.findElement(locator).getAttribute("src").contains("advanced_button.jpg");
From any of these approaches, once you have your boolean value, you can begin your if/then logic accordingly.
My apologies for being long winded with this, but hopefully it helps get you on the right path.
Use of Try Catch defies the very purpose of isdisplayed() used as If condition, one can write below code without using "if"
try{
driver.findElement(By.xpath(noRecordId)).isDisplayed();
//Put then statements here
}
Catch(Exception e)
{//put else statement here.}
I am learning HP UFT.
Recently I came across fireevent and I tried to implement it on the website of Flipkart. I was trying to use firevent("onmouseover") for the link Men on the homepage of the website.
I used ChildObjects to find out the Link and WebElement (in two different tests), first Highlighted it and then used object.fireevent("onmouseover") as well as object.fireevent("OnClick"). The OnClick is working and it is showing the link as selected (i.e. the dotted box covering the link when we press tab), but it is not showing the Menu Under Men Section.
I had googled and bingged a lot. But was unable to find the exact working of FireEvent in QTP/UFT.
Please Help me solving the above problem as well as some tutorials on FireEvent.
EDIT: I am using IE 11 for testing.
Motti has already answered the technical definition, but I shall attempt to give you a solution to your functional issue.
In my experience .FireEvent often doesn't work as you would expect. An alternative for something like onmouseover is to simulate user behaviour a bit more closely by actually moving the mouse to the desired location. In our framework we have a little extension function to do just that, a pared-down version of which is shown here:
Sub My_MouseOver(objSender)
Dim absX : absX = objSender.GetROProperty("abs_x")
Dim absY : absY = objSender.GetROProperty("abs_y")
Dim width : width = objSender.GetROProperty("width")
Dim height : height = objSender.GetROProperty("height")
Dim x : x = (absX + (width / 2))
Dim y : y = (absY + (height / 2))
Dim deviceReplay : Set deviceReplay = CreateObject("Mercury.DeviceReplay")
deviceReplay.MouseMove x, y
Reporter.ReportEvent micDone, "A step name", "A useful step description"
End Sub
RegisterUserFunc "Link", "MouseOver", "My_MouseOver"
RegisterUserFunc "WebButton", "MouseOver", "My_MouseOver"
RegisterUserFunc "WebElement", "MouseOver", "My_MouseOver"
As an example you can then do as follows to bring up the "ELECTRONICS" menu overlay on flipkart.com (obviously substitute your own Browser and Page definitions):
Browser("Flipkart").Page("Main Nav").Link("xpath:=//a[#data-tracking-id='electronics']").MouseOver
In the original version there's various extra bits of error handling and custom reporting so it tells you what you clicked on, but the essence is the same. It locates the object on screen, calculates the centre and moves the mouse there. You might want to wait a small amount of time for the menu overlay to appear after doing so before calling .Click on one of the newly-displayed sub-elements.
I found a solution to my problem and it is working perfectly.
Setting.WebPackage("ReplayType") = 2
object.FireEvent "onmouseover"
Setting.WebPackage("ReplayType") = 1
In this case, the object will be:
Browser("name:=Online Shopping.*").Page("name:=Online Shopping.*").Link("innertext:=Men")
I have tried this and it is working fine. I guess, we do not need any alternatives. But I really don't know is, Ctrl+Space is not working for this in UFT. Don't know the reason.
This actually depends on what browser you're using.
Warning: There are exceptions to the information presented in this answer and it also may change in the future. The answer is meant to give a basic understanding but don't depend on it to be true without checking the behaviour for your specific version/use-case.
Originally QTP's FireEvent was supposed to call IE's non-standard fireEvent method.
On Firefox and Chrome this is implemented using the standard dispatchEvent. You should check which events the web site expects to get.
Things get complicated if you mix the event models (the standard DOM level 2 and Microsofts) as explained in this blog post.
I'm doing some drag and drop js testing with selenium-client. It works (taking screengrabs before and after clearly show the elements to switch places), but I'm having trouble programatically asserting the change happened.
Am I mental, or can I not do something like:
selenium.assert_equal("css locator", "expected id of element")
which, in this case, would look something like:
selenium.assert_equal("css=li:nth-child(1)", "li#list_item_2")
Any tips on how to implement this would be great.
Thanks,
Adam
Edit: if I had selenium.get_element that would take a selector and return what it was, I could then perform the assertion in the next step.
i.e.
element = selenium.get_element("css=li:nth-child(1)")
assert_equal(element, "li#list_item_2")
(I think).
Your example won't work because you're comparing two strings that aren't equal. One way to assert that your element has moved would be to use isElementPreset as demonstrated below:
//before drag and drop
assertTrue(selenium.isElementPresent("css=#source li:nth-child(1)"));
assertFalse(selenium.isElementPresent("css=#target li:nth-child(1)"));
//drag and drop code here
...
//after drag and drop
assertTrue(selenium.isElementPresent("css=#target li:nth-child(1)"));
assertFalse(selenium.isElementPresent("css=#source li:nth-child(1)"));
This example uses the Java client API, but it should be easy to work out how to do it in your preferred client API. It will also depend heavily on your application, as the above will check that the element with id of 'target' has one child li element before the drag and drop, and none afterwards. If you have a snippet of your HTML source I might be able to prove a more robust example.