Add tests for clojure.core/identity#907
Conversation
|
|
| (deftest test-identity | ||
| (is (NaN? (identity ##NaN))) | ||
| (doseq [value test-vals] | ||
| (is (identical? value (identity value)))))) |
There was a problem hiding this comment.
maybe put the value or the type of value as the docstring for the assertion?
There was a problem hiding this comment.
on second thought, i don't know if that's required...
There was a problem hiding this comment.
Happy to add docstrings, but my hope is that failing assertion will document the value in the terminal output.
| [clojure.core-test.portability #?(:cljs :refer-macros :default :refer) [when-var-exists]])) | ||
|
|
||
| ; Values for tests are declared eagerly, instead of using | ||
| ; clojure.test/are. Some dialects evaluate `are` params more |
There was a problem hiding this comment.
Huh, where exactly closure.test/are behaves oddly like that? Questioning why that wouldn’t be considered a bug to fix in a dialect (ticket could be linked here).
There was a problem hiding this comment.
Actually, it seems all supported dialects fail the same way in these scenarios: jvm, cljs, cljr, bb, lpy, and phel.
; fails with different object references
(deftest atoms-are=
(are [x] (= x x)
(atom nil)))
; fails with 3 instead of 1 (x referenced 3 times in do block)
(deftest increment-atom
(let [counter (atom 0)]
(are [x] (do x x x)
(swap! counter inc))
(is (= 1 @counter))))4 participants
Closes #305.
isvsareAs commented in the code, some dialects seem to evaluate
areparameters more than once, which is problematic for object equality. Test values are declared eagerly and tested usingdoseq/isinstead.##NaN
It turns out that most dialects don't consider
##NaNto be=oridentical?to itself. This case is tested separately withNaN?.