I have also lead in with two tests of type B and then it is
B - good
B - good
A - bad
So it seems like the ones that want to evaluate the log for the raised "bad" can only go first and once. So it still could be that python no more re-loads the code at all and thereby no more triggered "bad".
That leads it to continue and then come up with the "not found".
I have also lead in with two tests of type B and then it is
B - good
B - good
A - bad
So it seems like the ones that want to evaluate the log for the raised "bad" can only go first and once. So it still could be that python no more re-loads the code at all and thereby no more triggered "bad".
That leads it to continue and then come up with the "not found".