refactor(reforcexy): make optuna sampler dispatch exhaustive

diff --git a/ReforceXY/user_data/freqaimodels/ReforceXY.py b/ReforceXY/user_data/freqaimodels/ReforceXY.py
index 0cf5a38bad7f2d3fc60613ddbe5c167ba9dab15d..13217533ecfc27611db325aa129a3c5bea97b63a 100644 (file)
--- a/ReforceXY/user_data/freqaimodels/ReforceXY.py
+++ b/ReforceXY/user_data/freqaimodels/ReforceXY.py
@@ -20,6 +20,7 @@ from typing import (
     Tuple,
     Type,
     Union,
+    assert_never,
     cast,
 )
 
@@ -1368,35 +1369,39 @@ class ReforceXY(BaseReinforcementLearningModel):
             return False
 
     def create_sampler(self) -> BaseSampler:
-        sampler: SamplerType = self.rl_config_optuna.get(
+        sampler_config = self.rl_config_optuna.get(
             "sampler", ReforceXY._SAMPLER_TYPES[0]
-        )  # "tpe"
-        seed = self.rl_config_optuna.get("seed", 42)
-        # "auto"
-        if sampler == ReforceXY._SAMPLER_TYPES[1]:
-            logger.info(
-                "Hyperopt [global]: using AutoSampler (seed=%d)",
-                seed,
-            )
-            return optunahub.load_module("samplers/auto_sampler").AutoSampler(seed=seed)
-        # "tpe"
-        elif sampler == ReforceXY._SAMPLER_TYPES[0]:
-            logger.info(
-                "Hyperopt [global]: using TPESampler (n_startup_trials=%d, multivariate=True, group=True, seed=%d)",
-                self.optuna_n_startup_trials,
-                seed,
-            )
-            return TPESampler(
-                n_startup_trials=self.optuna_n_startup_trials,
-                multivariate=True,
-                group=True,
-                seed=seed,
-            )
-        else:
+        )
+        if sampler_config not in ReforceXY._SAMPLER_TYPES:
             raise ValueError(
-                f"Hyperopt [global]: unsupported sampler '{sampler}'. "
+                f"Hyperopt [global]: unsupported sampler '{sampler_config}'. "
                 f"Valid: {', '.join(ReforceXY._SAMPLER_TYPES)}"
             )
+        sampler = cast(SamplerType, sampler_config)
+        seed = self.rl_config_optuna.get("seed", 42)
+        match sampler:
+            case "tpe":
+                logger.info(
+                    "Hyperopt [global]: using TPESampler (n_startup_trials=%d, multivariate=True, group=True, seed=%d)",
+                    self.optuna_n_startup_trials,
+                    seed,
+                )
+                return TPESampler(
+                    n_startup_trials=self.optuna_n_startup_trials,
+                    multivariate=True,
+                    group=True,
+                    seed=seed,
+                )
+            case "auto":
+                logger.info(
+                    "Hyperopt [global]: using AutoSampler (seed=%d)",
+                    seed,
+                )
+                return optunahub.load_module("samplers/auto_sampler").AutoSampler(
+                    seed=seed
+                )
+            case _:
+                assert_never(sampler)
 
     @staticmethod
     def create_pruner(

diff --git a/ReforceXY/user_data/tests/test_reforcexy_sampler_dispatch.py b/ReforceXY/user_data/tests/test_reforcexy_sampler_dispatch.py
new file mode 100644 (file)
index 0000000..daae9ba
--- /dev/null
+++ b/ReforceXY/user_data/tests/test_reforcexy_sampler_dispatch.py
@@ -0,0 +1,87 @@
+import importlib
+import inspect
+import sys
+import types
+
+from optuna.samplers import TPESampler
+
+
+sys.path.insert(0, "/freqtrade/user_data")
+reforcexy_module = importlib.import_module("freqaimodels.ReforceXY")
+ReforceXY = reforcexy_module.ReforceXY
+
+
+class AutoSampler:
+    def __init__(self, seed: int) -> None:
+        self.seed = seed
+
+
+def make_model(sampler: str) -> ReforceXY:
+    model = object.__new__(ReforceXY)
+    model.rl_config_optuna = {"sampler": sampler, "seed": 7}
+    model.optuna_n_startup_trials = 3
+    return model
+
+
+def test_create_sampler_returns_tpe_when_sampler_is_tpe() -> None:
+    # Given: a ReforceXY model configured for the public tpe sampler.
+    model = make_model("tpe")
+
+    # When: the sampler factory is invoked.
+    sampler = model.create_sampler()
+
+    # Then: the factory returns Optuna's TPE sampler.
+    assert isinstance(sampler, TPESampler)
+
+
+def test_create_sampler_returns_auto_when_sampler_is_auto() -> None:
+    # Given: a ReforceXY model configured for the public auto sampler.
+    model = make_model("auto")
+    original_load_module = reforcexy_module.optunahub.load_module
+    reforcexy_module.optunahub.load_module = lambda _: types.SimpleNamespace(
+        AutoSampler=AutoSampler
+    )
+
+    try:
+        # When: the sampler factory is invoked.
+        sampler = model.create_sampler()
+    finally:
+        reforcexy_module.optunahub.load_module = original_load_module
+
+    # Then: the factory returns the AutoSampler provided by optunahub.
+    assert isinstance(sampler, AutoSampler)
+
+
+def test_create_sampler_rejects_invalid_public_sampler() -> None:
+    # Given: a ReforceXY model configured with an unsupported public sampler.
+    model = make_model("invalid")
+
+    try:
+        # When: the sampler factory is invoked.
+        model.create_sampler()
+    except ValueError as exc:
+        # Then: public invalid config fails before typed exhaustive dispatch.
+        message = str(exc)
+        assert "Hyperopt [global]: unsupported sampler 'invalid'." in message
+        assert "Valid: tpe, auto" in message
+    else:
+        raise AssertionError("unsupported sampler did not raise ValueError")
+
+
+def test_create_sampler_uses_exhaustive_match_dispatch() -> None:
+    # Given: ReforceXY exposes a finite SamplerType surface.
+    source = inspect.getsource(ReforceXY.create_sampler)
+
+    # When/Then: dispatch is encoded as a match with assert_never for type-level exhaustiveness.
+    assert "match sampler:" in source
+    assert 'case "tpe":' in source
+    assert 'case "auto":' in source
+    assert "assert_never(sampler)" in source
+
+
+if __name__ == "__main__":
+    test_create_sampler_returns_tpe_when_sampler_is_tpe()
+    test_create_sampler_returns_auto_when_sampler_is_auto()
+    test_create_sampler_rejects_invalid_public_sampler()
+    test_create_sampler_uses_exhaustive_match_dispatch()
+    print("sampler dispatch checks passed")