chore: benchmarks

benchmarks.py

@@ -0,0 +1,55 @@
+# ruff: noqa: D100, D103
+
+from functools import reduce
+from typing import Any, Callable
+
+from pytest_benchmark.fixture import BenchmarkFixture
+from stateless import (
+    Ability,
+    Depend,
+    Handler,
+    Need,
+    Success,
+    handle,
+    need,
+    run,
+    supply,
+)
+from stateless.errors import UnhandledAbilityError
+from typing_extensions import Never
+
+
+def never_handler(_: Ability[Any]) -> Never:
+    raise UnhandledAbilityError()
+
+
+dummy_handler: Handler[Never] = handle(never_handler)  # type: ignore
+
+
+def create_effect_chain(chain_length: int, n_yields: int) -> Callable[[], Success[str]]:
+    def base() -> Depend[Need[str], str]:
+        for i in range(n_yields):
+            yield from need(str)
+        return "done"
+
+    def wrap_test_handler(
+        f: Callable[[], Depend[Need[str], str]],
+    ) -> Callable[[], Depend[Need[str], str]]:
+        return dummy_handler(f)
+
+    g = reduce(lambda acc, _: wrap_test_handler(acc), range(chain_length - 1), base)
+
+    return supply("")(g)
+
+
+def test_effect_chain(benchmark: BenchmarkFixture) -> None:
+    """Benchmark a long chain of functions using functools.reduce."""
+
+    effect = create_effect_chain(500, 0)()
+    benchmark(run, effect)
+
+
+def test_handler_chain(benchmark: BenchmarkFixture) -> None:
+    effect = create_effect_chain(chain_length=500, n_yields=500)()
+
+    benchmark(run, effect)

pyproject.toml

@@ -15,12 +15,13 @@ cloudpickle = "^3.0.0"
 mypy = "^1.6.1"
 ipdb = "^0.13.13"
 ipython = "^8.17.2"
-pytest = "^7.4.3"
+pytest = "^8"
 pyright = "^1.1.336"
 pre-commit = "^3.5.0"
 ruff = "^0.14.2"
 coverage = "^7.3.2"
 toml = "^0.10.2"
+pytest-benchmark = "^5.2.1"
 
 
 [tool.mypy]