Updated the ActivityIndicator tests to also check for visibility when… #2838
Conversation
… started and stopped Signed-off-by: Moises <moises.aranas@gmail.com>
Signed-off-by: Moises <moises.aranas@gmail.com>
Signed-off-by: Moises <moises.aranas@gmail.com>
| @@ -5,3 +5,7 @@ | |||
|
|
|||
| class ActivityIndicatorProbe(SimpleProbe): | |||
| native_class = NSProgressIndicator | |||
|
|
|||
| @property | |||
| def is_hidden(self): | |||
There was a problem hiding this comment.
It would be worth being explicit that we're talking about the spinner visibility here. Depending on the implementation, it would be possible to have a widget that is "visible" in the sense of Toga's drawing instructions, but that that displays no content because the spinner is hidden. iOS got caught on this because spinner visibility and widget visibility are treated as the same thing; but that's no necessarily the same.
There's already a probe for evaluating visibility; spinner visibility isn't necessarily the same thing.
(edit) not sure what happened... this got submitted before I finished writing the comment.
There was a problem hiding this comment.
Hey @freakboy3742 ! I will have a look deeper into this - it looks like it is possible to do in iOS and macOS but I still need to figure out for the Linux and Windows implementations how it would be done
There was a problem hiding this comment.
Winforms doesn't have an ActivityIndicator, so the solution there will be easy :-)
As for Linux - if we can verify that the GTK widget behaves as expected (i.e., hidden and running are different properties, and a stopped indicator is defined as not visible, but not hidden), then it might be sufficient to just check the active property.
… started and stopped
Updates the tests (based on #2829 review comments) to include visibility checks when testing the start/stop function of the ActivityIndicator widget
PR Checklist: