add test_validate_parameterized_size

tests/test_model/test_v0_5.py

@@ -1,5 +1,7 @@
+from copy import deepcopy
 from datetime import datetime
-from typing import Any, Dict, Union
+from types import MappingProxyType
+from typing import Any, Dict, Mapping, Union
 
 import pytest
 from pydantic import RootModel, ValidationError
@@ -222,10 +224,12 @@ def test_input_axis(kwargs: Union[Dict[str, Any], SpaceInputAxis]):
     check_type(InputAxis, kwargs)
 
 
-@pytest.fixture
-def model_data():
+@pytest.fixture(scope="module")
+def model():
+    """reuse model object to avoid expensive model validation,
+    use only when not manipulating the model!"""
     with ValidationContext(perform_io_checks=False):
-        model = ModelDescr(
+        return ModelDescr(
             documentation=UNET2D_ROOT / "README.md",
             license=LicenseId("MIT"),
             git_repo=HttpUrl("https://github.com/bioimage-io/core-bioimage-io-python"),
@@ -288,12 +292,21 @@ def model_data():
             ),
             type="model",
         )
-        data = model.model_dump(mode="json")
-        assert data["documentation"] == str(UNET2D_ROOT / "README.md"), (
-            data["documentation"],
-            str(UNET2D_ROOT / "README.md"),
-        )
-        return data
+
+
+@pytest.fixture(scope="module")
+def const_model_data(model: ModelDescr):
+    data = model.model_dump(mode="json")
+    assert data["documentation"] == str(UNET2D_ROOT / "README.md"), (
+        data["documentation"],
+        str(UNET2D_ROOT / "README.md"),
+    )
+    return MappingProxyType(data)
+
+
+@pytest.fixture
+def model_data(const_model_data: Mapping[str, Any]):
+    return deepcopy(dict(const_model_data))
 
 
 @pytest.mark.parametrize(
@@ -389,18 +402,12 @@ def test_output_fixed_shape_too_small(model_data: Dict[str, Any]):
     assert summary.status == "failed", summary.format()
 
 
-def test_get_axis_sizes_with_surplus_n(model_data: Dict[str, Any]):
-    with ValidationContext(perform_io_checks=False):
-        model = ModelDescr(**model_data)
-
+def test_get_axis_sizes_with_surplus_n(model: ModelDescr):
     key = (model.inputs[0].id, AxisId("y"))
     _ = model.get_axis_sizes(ns={key: 1}, batch_size=1)
 
 
-def test_get_axis_sizes_with_partial_max_size(model_data: Dict[str, Any]):
-    with ValidationContext(perform_io_checks=False):
-        model = ModelDescr(**model_data)
-
+def test_get_axis_sizes_with_partial_max_size(model: ModelDescr):
     key = (model.inputs[0].id, AxisId("y"))
     ns = {key: 100}
     wo_max_shape = model.get_axis_sizes(ns=ns)
@@ -426,7 +433,7 @@ def test_output_ref_shape_mismatch(model_data: Dict[str, Any]):
     model_data["outputs"][0]["axes"][2] = {
         "type": "space",
         "id": "x",
-        "size": {"tensor_id": "input_1", "axis_id": "x"},
+        "size": {"tensor_id": "input_1", "axis_id": "y"},
         "halo": 2,
     }
     summary = validate_format(
@@ -450,7 +457,7 @@ def test_output_ref_shape_too_small(model_data: Dict[str, Any]):
     model_data["outputs"][0]["axes"][2] = {
         "type": "space",
         "id": "x",
-        "size": {"tensor_id": "input_1", "axis_id": "x"},
+        "size": {"tensor_id": "input_1", "axis_id": "y"},
         "halo": 2,
     }
     summary = validate_format(
@@ -553,3 +560,9 @@ def test_empty_axis_data():
 )
 def test_identifier_identity(a: Any, b: Any):
     assert a == b
+
+
+def test_validate_parameterized_size(model: ModelDescr):
+    param_size = model.inputs[0].axes[2].size
+    assert isinstance(param_size, ParameterizedSize)
+    assert param_size.validate_size(512) == 512