I'm trying to understand how UIAutomation can be used to perform automatic application tests, so I've created a temperature converter app that can be driven from the JavaScript interface. Here's the UI:
The top text field has an Accessibility label of "Celsius", similarly the bottom field has "Fahrenheit" as its Accessibility label. I'm trying to drive it through this script:
UIALogger.logStart("Test -40ºC == -40ºF");
var window = UIATarget.localTarget().frontMostApp().mainWindow();
var celsiusField = window.textFields()["Celsius"];
var fahrenheitField = window.textFields()["Fahrenheit"];
var convertButton = window.buttons()["Convert"];
celsiusField.setValue("-40");
convertButton.tap();
var fahrenheitValue = fahrenheitField.value();
if (fahrenheitValue == "-40.0") {
UIALogger.logPass("-40C == -40F");
} else {
UIALogger.logFail("-40C == " + fahrenheitValue + "F");
}
This correctly sets the text in the Celsius field to "-40", and when the button is tapped, the app updates the text in the Fahrenheit field to "-40.0". However, fahrenheitValue
has the value "Fahrenheit" (the name/label of the text field), so the test fails. I wondered whether this was maybe a timing issue, so put in a one-second delay after the tap, which didn't change the behaviour.
Is it possible to get the text out of the text field and compare it with the expected value?
Is it getting "Fahrenheit" because that's the accessibility label or because it's the placeholder text?
It looks like value() might return the placeholder text. One workaround could be to modify the placeholder value once the user starts typing to set it to be the text the user entered and then reverting it to Fahrenheit when they delete? I assume that even when you call setValue it acts as it would when the UITextField delegate methods get called?