Objective

Implement a function that compares an int against an unsigned int by their real numeric values and returns -1, 0 or 1.

Steps

• a is an int and may be negative. b is an unsigned int, so it never is.
• Return -1 when a is numerically smaller than b, 0 when both stand for the same number, 1 when a is larger. No other value is accepted.
• Writing a < b gives a wrong answer. Both operands have the same rank but differ in signedness, so C converts a to unsigned int before it compares anything, and a negative a arrives as a huge positive value: -1 becomes 4294967295. Settle the sign of a yourself, before any comparison.
• Equality has the same hole. a == b claims -1 and 4294967295u are the same number, and it says the same of INT_MIN and 2147483648u.
• A negative a is smaller than every possible b, 0 included, and INT_MIN is no exception.
• Once you know a is zero or positive, casting it to unsigned int preserves its value, so from there a direct comparison is safe.
• Do not negate a to get its magnitude. -INT_MIN does not fit in an int, and that overflow is undefined behavior.
• The tests print the returned value on its own line.

Objective

Implement a function that compares an int against an unsigned int by their real numeric values and returns -1, 0 or 1.

Steps

• a is an int and may be negative. b is an unsigned int, so it never is.
• Return -1 when a is numerically smaller than b, 0 when both stand for the same number, 1 when a is larger. No other value is accepted.
• Writing a < b gives a wrong answer. Both operands have the same rank but differ in signedness, so C converts a to unsigned int before it compares anything, and a negative a arrives as a huge positive value: -1 becomes 4294967295. Settle the sign of a yourself, before any comparison.
• Equality has the same hole. a == b claims -1 and 4294967295u are the same number, and it says the same of INT_MIN and 2147483648u.
• A negative a is smaller than every possible b, 0 included, and INT_MIN is no exception.
• Once you know a is zero or positive, casting it to unsigned int preserves its value, so from there a direct comparison is safe.
• Do not negate a to get its magnitude. -INT_MIN does not fit in an int, and that overflow is undefined behavior.
• The tests print the returned value on its own line.