perf: benchmarks for pint_eval

Author: hgrecco

pint/testsuite/benchmarks/test_01_eval.py

@@ -0,0 +1,23 @@
+from __future__ import annotations
+
+import pytest
+
+from pint.pint_eval import plain_tokenizer, uncertainty_tokenizer
+
+VALUES = [
+    "1",
+    "1 + 2 + 5",
+    "10 m",
+    "10 metros + 5 segundos",
+    "10 metros * (5 segundos)",
+]
+
+
+def _tok(tok, value):
+    return tuple(tok(value))
+
+
+@pytest.mark.parametrize("tokenizer", (plain_tokenizer, uncertainty_tokenizer))
+@pytest.mark.parametrize("value", VALUES)
+def test_pint_eval(benchmark, tokenizer, value):
+    benchmark(_tok, tokenizer, value)