fix: ensure 'test_policy' works with gymnasium>=0.28.1 (#276)

This commit ensures the `test_policy` utility works with
gymnasium>=0.28.1 (see
https://gymnasium.farama.org/content/migration-guide/).

stable_learning_control/utils/test_policy.py

@@ -285,7 +285,8 @@ def run_policy(
     )
 
     logger = EpochLogger(verbose_fmt="table")
-    o, r, d, ep_ret, ep_len, n = env.reset(), 0, False, 0, 0, 0
+    o, _ = env.reset()
+    r, d, ep_ret, ep_len, n = 0, False, 0, 0, 0
     supports_deterministic = True  # Only supported with gaussian algorithms.
     render_error = False
     while n < num_episodes:
@@ -298,7 +299,7 @@ def run_policy(
                 render_error = True
                 log_to_std_out(
                     (
-                        "WARNING: Nothing was rendered since no render method was "
+                        "Nothing was rendered since no render method was "
                         f"implemented for the '{env.unwrapped.spec.id}' environment."
                     ),
                     type="warning",
@@ -321,13 +322,14 @@ def run_policy(
             a = policy.get_action(o)
 
         # Perform action in the environment and store result.
-        o, r, d, _ = env.step(a)
+        o, r, d, truncated, _ = env.step(a)
         ep_ret += r
         ep_len += 1
-        if d or (ep_len == max_ep_len):
+        if d or truncated:
             logger.store(EpRet=ep_ret, EpLen=ep_len)
             logger.log("Episode %d \t EpRet %.3f \t EpLen %d" % (n, ep_ret, ep_len))
-            o, r, d, ep_ret, ep_len = env.reset(), 0, False, 0, 0
+            o, _ = env.reset()
+            r, d, ep_ret, ep_len = 0, False, 0, 0
             n += 1
 
     print("")