fix: use lax validation by default (#159)

* fix: use lax validation

* style: [pre-commit.ci] auto fixes [...]

* fix autogen, fix tests

* xfail round trip

* test: update test

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

src/ome_autogen.py

@@ -1023,6 +1023,30 @@ def fname(self) -> str:
 _url = os.path.join(_this_dir, "ome_types", "ome-2016-06.xsd")
 _target = os.path.join(_this_dir, "ome_types", "model")
 
+AnnotationID_Override = r"""class AnnotationID(LSID):
+    regex = re.compile(
+        r"(urn:lsid:([\w\-\.]+\.[\w\-\.]+)+:Annotation:\S+)|(Annotation:\S+)"
+    )
+
+    @classmethod
+    def __get_validators__(cls):
+        yield cls.validate
+
+    @classmethod
+    def validate(cls, v) -> "AnnotationID":
+        if not cls.regex.match(v):
+            search = cls.regex.search(v)
+            if search:
+                import warnings
+
+                new_v = search.group()
+                warnings.warn(f"Casting invalid AnnotationID {v!r} to {new_v!r}")
+                v = new_v
+        return v
+"""
+
+_SIMPLE_OVERRIDES = {"AnnotationID": AnnotationID_Override}
+
 
 def convert_schema(url: str = _url, target_dir: str = _target) -> None:
     print("Inspecting XML schema ...")
@@ -1052,7 +1076,9 @@ def convert_schema(url: str = _url, target_dir: str = _target) -> None:
         init_imports.append((converter.fname, elem.local_name))
         converter.write(filename=targetfile)
 
-    text = "\n".join([s.format() for s in simples])
+    text = "\n".join(
+        [_SIMPLE_OVERRIDES.get(s.elem.local_name) or s.format() for s in simples]
+    )
     text = black_format(sort_imports(text))
     with open(os.path.join(target_dir, "simple_types.py"), "w", encoding="utf-8") as f:
         f.write(text)

src/ome_types/_xmlschema.py

@@ -224,7 +224,7 @@ def xmlschema2dict(
     xml: str,
     schema: Optional[xmlschema.XMLSchema] = None,
     converter: XMLSchemaConverter = OMEConverter,
-    validate: bool = True,
+    validate: bool = False,
     **kwargs: Any,
 ) -> Dict[str, Any]:
     if isinstance(xml, bytes):

tests/data/invalid_xml_annotation.ome.xml

@@ -0,0 +1,78 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<OME xmlns="http://www.openmicroscopy.org/Schemas/OME/2016-06"
+    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openmicroscopy.org/Schemas/OME/2016-06
+http://www.openmicroscopy.org/Schemas/OME/2016-06/ome.xsd">
+    <Image ID="Image:0" Name="P_t00000,W_s00_0">
+        <Pixels BigEndian="false" DimensionOrder="XYZTC" ID="Pixels:0" Interleaved="false" PhysicalSizeX="0.08850000022125" PhysicalSizeXUnit="µm" PhysicalSizeY="0.08850000022125" PhysicalSizeYUnit="µm" PhysicalSizeZ="1.0" PhysicalSizeZUnit="µm" SignificantBits="16" SizeC="2" SizeT="11" SizeX="171" SizeY="196" SizeZ="5" Type="uint16">
+            <Channel ID="Channel:0:0" SamplesPerPixel="1">
+                <LightPath/>
+            </Channel>
+            <Channel ID="Channel:0:1" SamplesPerPixel="1">
+                <LightPath/>
+            </Channel>
+            <MetadataOnly/>
+        </Pixels>
+    </Image>
+    <StructuredAnnotations>
+        <XMLAnnotation ID="XMLAnnotation:0">
+            <Value>
+                <SpimData version="0.2">
+                    <BasePath type="relative">.</BasePath>
+                    <SequenceDescription>
+                        <ImageLoader format="bdv.hdf5">
+                            <hdf5 type="relative">xyzct_16bit__mitosis.h5</hdf5>
+                        </ImageLoader>
+                        <ViewSetups>
+                            <ViewSetup>
+                                <id>0</id>
+                                <name>channel 1</name>
+                                <size>171 196 5</size>
+                                <voxelSize>
+                                    <unit>µm</unit>
+                                    <size>0.08850000022125 0.08850000022125 1.0</size>
+                                </voxelSize>
+                                <attributes>
+                                    <channel>1</channel>
+                                </attributes>
+                            </ViewSetup>
+                            <ViewSetup>
+                                <id>1</id>
+                                <name>channel 2</name>
+                                <size>171 196 5</size>
+                                <voxelSize>
+                                    <unit>µm</unit>
+                                    <size>0.08850000022125 0.08850000022125 1.0</size>
+                                </voxelSize>
+                                <attributes>
+                                    <channel>2</channel>
+                                </attributes>
+                            </ViewSetup>
+                            <Attributes name="channel">
+                                <Channel>
+                                    <id>1</id>
+                                    <name>1</name>
+                                </Channel>
+                                <Channel>
+                                    <id>2</id>
+                                    <name>2</name>
+                                </Channel>
+                            </Attributes>
+                        </ViewSetups>
+                        <Timepoints type="range">
+                            <first>0</first>
+                            <last>10</last>
+                        </Timepoints>
+                    </SequenceDescription>
+                    <ViewRegistrations>
+                        <ViewRegistration setup="0" timepoint="0">
+                            <ViewTransform type="affine">
+                                <affine>0.08850000022125 0.0 0.0 0.0 0.0 0.08850000022125 0.0 0.0 0.0 0.0 1.0
+              0.0</affine>
+                            </ViewTransform>
+                        </ViewRegistration>
+                    </ViewRegistrations>
+                </SpimData>
+            </Value>
+        </XMLAnnotation>
+    </StructuredAnnotations>
+</OME>

tests/test_invalid_schema.py

@@ -0,0 +1,15 @@
+from pathlib import Path
+
+import pytest
+
+from ome_types import from_xml
+
+DATA = Path(__file__).parent / "data"
+
+
+def test_bad_xml_annotation() -> None:
+    """Test that a bad XML annotation is not imported"""
+    with pytest.warns(match="Casting invalid AnnotationID"):
+        ome = from_xml(DATA / "invalid_xml_annotation.ome.xml")
+    assert len(ome.images) == 1
+    assert ome.structured_annotations[0].id == "Annotation:0"

tests/test_model.py

@@ -18,11 +18,13 @@
     # Some timestamps have negative years which datetime doesn't support.
     "timestampannotation",
 }
+SHOULD_FAIL_VALIDATION = {"invalid_xml_annotation"}
 SHOULD_RAISE_READ = {"bad"}
 SHOULD_FAIL_ROUNDTRIP = {
     # Order of elements in StructuredAnnotations and Union are jumbled.
     "timestampannotation-posix-only",
     "transformations-downgrade",
+    "invalid_xml_annotation",
 }
 SHOULD_FAIL_ROUNDTRIP_LXML = {
     "folders-simple-taxonomy",
@@ -81,7 +83,8 @@ def true_stem(p):
 @pytest.mark.parametrize("validate", validate)
 def test_from_xml(xml, parser: str, validate: bool, benchmark):
 
-    if true_stem(xml) in SHOULD_RAISE_READ:
+    should_raise = SHOULD_RAISE_READ.union(SHOULD_FAIL_VALIDATION if validate else [])
+    if true_stem(xml) in should_raise:
         with pytest.raises(
             (XMLSchemaValidateError, ValidationError, XMLSchemaValidationError)
         ):
@@ -171,6 +174,8 @@ def test_serialization(xml, validate, parser):
     """Test pickle serialization and reserialization."""
     if true_stem(xml) in SHOULD_RAISE_READ:
         pytest.skip("Can't pickle unreadable xml")
+    if validate and true_stem(xml) in SHOULD_FAIL_VALIDATION:
+        pytest.skip("Can't pickle invalid xml with validate=True")
 
     ome = from_xml(xml, parser=parser, validate=validate)
     serialized = pickle.dumps(ome)