Bug: announce_highest in hog project can't pass all tests about cs61a

[czahie]

Hi! Thanks for your issue. I just checked my code I think the problem is because of the line:
previous_score = score0 or previous_score = score1. The previous_score is updated every time I call say. Therefore, if you call the function for the second time, there would be no print messages. If I remembered it correctly, my code passed all the test cases back to the course of 2017 Fall. So maybe they have changed the rule. Thanks again for your report anyway!

[Mike-Dai]

Thanks! I take the version of 2018 Spring and they might add more tests.

[pangjac]

Hi @czahie , it is very nice for sharing the notes with us. I have a quick question, where do we get the OK credential if not enrolled in Berkeley, how do you pass all tests if without OK? I read cs61 account trouble shooting, but still confused about their autograder system.
Thank you so much!

[czahie]

@pangjac Actually, I used ok. I think there's a way to use ok locally, meaning that you can use ok without a Berkeley email. But I'm not sure whether they've updated this or not.

[gclintj]

Hi! Thanks for your issue. I just checked my code I think the problem is because of the line:
previous_score = score0 or previous_score = score1. The previous_score is updated every time I call say. Therefore, if you call the function for the second time, there would be no print messages. If I remembered it correctly, my code passed all the test cases back to the course of 2017 Fall. So maybe they have changed the rule. Thanks again for your report anyway!

Thanks!
I don't fully understand the command "nonlocal", but I believe it causes the problem that the function gives a different result when later called.
@Mike-Dai 's solution also makes it unnecessary to use the "nonlocal" command.
I think it is what the instruction means: "1) Rather than reassigning [var] to its new value, create a new variable to hold that new value. Use that new variable in future calculations."
Thank you both! Now I understand much better about this problem.
BTW, potentially "previous_high" is another headache, with similar two issues.