queen-attack: update tests to v2.1.0

[cmccandless]

Canonical Data

[mrcfps]

@cmccandless This exercise had been implemented before canonical data came out. As for updating to canonical data, here are three questions worth discussing:

• Should we change function board to create? If so, what is its new signature?
• How should we deal with those tests beyond canonical set, such as checking for two pieces being placed on the same position?
• Should we still keep tests of graphical representation of chess board?

[cmccandless]

• Should we change function board to create? If so, what is its new signature?

This is how I understand the spec:

def create(x, y):
    # validate position here
    return (x, y) # return queen pseudo-object


# can_attack() unchanged
def can_attack(white_queen, black_queen):
    pass

• How should we deal with those tests beyond canonical set, such as checking for two pieces being placed on the same position?

These can be checked within can_attack.

• Should we still keep tests of graphical representation of chess board?

I think generation of the graphical representation is a) trivial and b) outside the scope of this exercise, but I'm open to debate on those points. @m-a-ge , @N-Parsons , you guys have an opinion on this?

[mrcfps]

From my perspective, if we leave create(x, y) in the stub, the answer would be quite obvious:

def create(x, y):
    if not 0 <= x <= 7 or not 0 <= y <= 7:
        raise ValueError("invalid position")

I don't think there is any point of doing this. So instead I suggest a stub like this:

class Queen(object):
    def __init__(self, row, column):
        pass

    def can_attack(self, other_queen):
        pass

It looks pretty easy, so I guess we should move this exercise forward in config.json.

[cmccandless]

In preparation for the next version of Exercism (known as "nextercism" for now), we are working to categorize and reorganize exercises into a more natural progression. So for now, we'll just leave this exercise where it is in the track.