Fix order in test assertions

[juxtin]

The convention for clojure.test is (is (= expected actual)), which is the reverse of what the template uses in handler_test.clj. This commit switches the arguments to = so that failures will be reported correctly.

[weavejester]

The convention for clojure.test is (is (= expected actual))

Could you provide a source for that?

[juxtin]

@weavejester sure thing, the clojure.test API reference has a bunch of examples. I also noticed this when I wrote a failing test and saw that "actual" and "expected" were backwards in the output.

[weavejester]

I don't see anything in the API reference. The example it gives would work either way around.

user> (is (= 5 (+ 2 2)))

FAIL in  (:1)
expected: (= 5 (+ 2 2))
  actual: (not (= 5 4))

[juxtin]

Yeah, perhaps the docs aren't as explicit on this as I thought. But if you look at examples like this one:

(testing "Arithmetic"
  (testing "with positive integers"
    (is (= 4 (+ 2 2)))
    (is (= 7 (+ 3 4))))
  (testing "with negative integers"
    (is (= -4 (+ -2 -2)))
    (is (= -1 (+ 3 -4)))))

They consistently put the expression with the function call second, and the expected return value first.

[weavejester]

Sure, but that doesn't mean that the library favours one way over the other, just that the examples just happen to be written like that.

[juxtin]

You're right, it turns out that humane-test-output was responsible for making the output look actually wrong when the reverse order is used. With that installed, I get:

user=> (is (= 5 (+ 2 2)))

expected: 5
  actual: 4
    diff: - 5
          + 4

That won't be an issue for many people, but I do think that following the convention is for the best. If it helps, this is also the way that Clojure's own tests are written (see example).