Sikuli for system testing

if you want to do serious testing with Sikuli, use the Python/Jython unit test feature directly.

The IDE test feature is a nice gimmick, but not really for professional use.

Depending on your pre-knowledge, you might decide, to use the Java API and make your tests using JUnit package.

Sorry, there is no specific Sikuli oriented testing docs.

For Python/Jython have a look at the official docs.

Same goes for Java: you are only using the Sikuli Java API to do the visual searching on the screen and may be the click and type/paste actions. This is how many people use Sikuli fetters as an add-on to the Selenium approach.

Cheers for the reply. Still would like to know the what the difference is between Error, and Failure in the unit test window. Just in case it is something I can work with.

What I definitely know:
Failure means, that a test ended with an assert exception (assert(False)).

I guess, that an error means any other exception during run of a test. But that is exactly, what does not work: If your test scripts produce an exception (e.g. a simple syntax error), the test IDE will simply hang, you have to interrupt it (hopefully using the little stop button in the transparent mini window) and in many cases you have to fully restart the IDE to run your tests again.

Conclusion: Errors? I have never seen them ;-)

If interested: I have made a script containing a hack, that allows you to fully simulate the test run in normal IDE until it is error free. Then you can run it as real unit test without modification and get your statistics.
This example runs calc.exe on Windows.
The problem you might have is this line:
 ac = self.ac = App("Rechner")
It was scripted on a german Windows. Exchange the word "Rechner" with the title text of the window of calc.exe.

the last line
weAreInIDE(("testAdd",))
could be changed to run both tests in normal IDE:
weAreInIDE(("testAdd","testTypeNumber"))

download: https://files.me.com/rhocke/zpsan6