Clojurians Log v2

I noticed that in the mapping namespaces, deftest => defdescribe > it and testing => binding that pushes strings (and the reporting uses that dyn var); however defexpect => defdescribe (with > it in my PR) and expecting => describe which seems to lose information. Then I tried this:

(defdescribe one "two"
  (it "four"
    (describe "three"
      (describe "five"
        (should (= 13 42))))))

(defdescribe one "two"
  (describe "three"
    (describe "four"
      (it "five"
        (should (= 13 42))))))

i didn't base the expectations mapping on deftest but on lazytest

describe creates a suite, a testing context object (plain map) with a :children entry that holds all of the nested suites and test cases

so calling it inside it will return a plain map (with the results of the assertion) and if the assertion throws, only the it will catch it

Okay, that wasn't sufficiently clear to me from the docs (re: appropriate nesting of it / describe). It does highlight that the Expectations mapping is problematic (in two ways): • right now, nothing maps to it so there are no actual test cases being created (`defexpect` => defdescribe, expecting => describe, no existing construct => it) • my PR to slip it into defexpect "works" but effectively negates the behavior of expecting (since that would become a nested describe) If the PR is accepted, test cases will "work", but no have any expecting context -- and expecting should be made to behave like testing with the dyn var etc. If the PR is not accepted, in order to have actual test cases, expect (Expectation) would probably need to be mapped to lt/expect-it somehow?

(in Expectations, expecting maps directly to testing:

(defmacro expecting
  "The Expectations version of `clojure.test/testing`."
  [string & body]
  `(t/testing ~string ~@body))

yeah that's a good description

you're the sole user of the expectations namespace, so you tell me how you want it to work

i'll add thoughts to the documentation detailing when/how to nest the various functions