I've worked out since that the markers '-' and '^' don't get removed correctly if they're used as markers from the expected results ie they're interpreted as regular text. Thus, removing them from the series of comment markers fixes this!

I think it is because of the escape sequence, but I don't know what their intention is here. The sequence is for '§'. I got round this by allowing comment markers of more than one character ("\xc2\xa7" has 8 characters). I guess handling unicode properly would have been another way of doing it. (Python2)

trailing space after "pears"?

there seems to be issues with the Test suite presenting incorrect solutions when the '-' symbol is involved. There are some cases where it doesn't correctly comment out the answers.

I second that! Agreed that it doesn't handle that edge case well.