Swap inconsistent assert_eq! argument order in testing chapter

[regexident]

The chapter starts off by using assert_eq!(…); with assert_eq!(actual, expected); argument ordering, but then suddenly swaps the order to assert_eq!(expected, actual); with assert_eq!(4, add_two(2));

Further more it then states …

We could write the assertion in this test as assert_eq!(add_two(2), 4), which would result in the same failure message
that displays `` assertion failed: (left == right).

… which seems weird as in the chapter assert_eq!(4, add_two(2)) actually is the odd one out, not assert_eq!(add_two(2), 4).

By swapping the args the message of "which would result in the same failure message" remains true, but the text feels a bit more consistent and predictable.

[chriskrycho]

Yeah, this seems like a reasonable change for consistency’s sake. Thanks!