Workaround for supporting Qt desktop applications.

[TMenting]

No description provided.

[ferpasri]

System Under Test: https://github.com/tonis-lusmagi/QT_Notepad

1. Master and this Qt solution branch : Seems that both branches are wrongly detecting some menu items widgets
   (Maybe this should be a different issue)

2. Master branch derives actions in some Qt widgets of the main windows but is not deriving an action in the OK button.
   Also master is correctly ignoring the action derivation of blocked widgets from the posterior panel

3. This Qt branch derives actions in the main window and also in the OK button.
   But is also deriving actions in the widgets of the posterior panel (that should be blocked)

[ferpasri]

@TMenting What Qt SUT did you use?
So we can test multiple Qt SUTs

[TMenting]

@ferpasri That's strange.. I've tested it against my own Qt application.
I just checked it in Spy mode, and it looked fine. This PR mainly enables TESTAR to detect actions. without this fix, TESTAR will not identify available actions because everything will me marked as blocked.

[ferpasri]

Thanks, with your Qt app the difference of action derivation is clear (left master, right this Qt branch)

Think I need to test deep with more Qt examples

[TMenting]

Note: This PR is embedded in PR 299.

[ferpasri]

This feature is already included in the #299
We will work on integrating the complete visual validation feature, and we close this PR.