Set a random seed to ensure repeatability of tests

Drawback to this is that the test doesn't exercise the code as fully
as more random usage would.

Pro is that we don't get false negatives from CI tests as often.

Author: mattiuusitalo

test/malli/parser_test.cljc

@@ -93,7 +93,7 @@
                     parse (m/parser s)
                     unparse (m/parser s)]
                 (if expected-simple
-                  (doseq [g (mg/sample s)]
+                  (doseq [g (mg/sample s {:seed 0})]
                     (testing (pr-str g)
                       (let [p (parse g)]
                         (is (identical? g p))