cider-test-run-test fails with (wrong-type-argument stringp (dict ...

[sktoiva]

I'm using Emacs 25.0.50.1 and CIDER 0.10.0snapshot.

Given the following example:

(deftest example1
  (checking "this-will-fail" 100
            [n gen/pos-int]
            (is (> 0 n))))

Running test with cider-test-run-test or cider-test-run-tests, the reporting will fail with following error message:

Debugger entered--Lisp error: (wrong-type-argument stringp (dict "fail" (0) "failing-size" 0 "num-tests" ...

I have tested it without the checking macro:

(deftest example2
  (testing "this-will-fail"
    (is (> 0 1))))

This works as intended, so currently I'm assuming the error has something to do with test.chuck and not CIDER itself.

However, invoking the tests with (clojure.test/run-tests) provides the intended output, so I'm a bit loss here where the problem might lie.

[gfredericks]

it looks as if the return value from calling the test function is making its way into cider as some kind of emacs data structure, and that cider is expecting a string for some reason.

[gfredericks]

alternatively, instead of being related to the test return value it might be related to something passed to the clojure.test reporting functions. @lackita any guesses?

[gfredericks]

@sktoiva I'm also curious what happens if you used defspec instead.

[sktoiva]

defspec works as intended, or at least so that the *cider-test-report* buffer contains information about failed tests and the specific information about how it failed is printed to *cider-repl*.