On disable, dont try to pop a key that doesn't exist

[psobolewskiPhD]

Closes napari/napari#7135

This is a naive fix. If the key doesn't exist when disabling, just return None.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 100.00%. Comparing base (eeeeacc) to head (73995ca).

Additional details and impacted files

@@            Coverage Diff            @@
##              main      #356   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           37        37           
  Lines         2758      2758           
=========================================
  Hits          2758      2758           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[DragaDoncila]

This might fix the issue but I'm concerned that the early return in that function didn't save us - I think that check only works across a single session, but it's still not clear to me what's going on exactly. I was also able to reproduce this back to 0.4.18 which seems unbelievable?? So I'm looking at this a bit closer to figure out what's going on

[DragaDoncila]

Ahh yess, I see what's going on. It's my own code 🤣 And that's why it's broken back to 0.4.18, because it's in the new npe2 release...

[DragaDoncila]

Confirmed this addresses the issue. Let's merge and release npe2 asap