Merge pull request #243 from dkpro/bugfix/242-Array-features-are-wrap…

…ped-in-extra-array #242 - Array features are wrapped in extra array
dkpro · Dec 15, 2021 · 2a51991 · 2a51991
2 parents 0753e05 + 4564886
commit 2a51991
Show file tree

Hide file tree

Showing 7 changed files with 199 additions and 7 deletions.
diff --git a/cassis/cas.py b/cassis/cas.py
@@ -8,8 +8,16 @@
 from attr import validators
 from sortedcontainers import SortedKeyList
 
-from cassis.typesystem import TYPE_NAME_SOFA, FeatureStructure, TypeCheckError, TypeSystem, TYPE_NAME_FS_LIST, \
-    TYPE_NAME_FS_ARRAY, FEATURE_BASE_NAME_HEAD, TypeSystemMode
+from cassis.typesystem import (
+    FEATURE_BASE_NAME_HEAD,
+    TYPE_NAME_FS_ARRAY,
+    TYPE_NAME_FS_LIST,
+    TYPE_NAME_SOFA,
+    FeatureStructure,
+    TypeCheckError,
+    TypeSystem,
+    TypeSystemMode,
+)
 
 _validator_optional_string = validators.optional(validators.instance_of(str))
 

diff --git a/cassis/typesystem.py b/cassis/typesystem.py
@@ -189,6 +189,7 @@
 
 _LIST_TYPES = _PRIMITIVE_LIST_TYPES | {TYPE_NAME_FS_LIST}
 
+
 class TypeSystemMode(Enum):
     """How much type system information to include."""
 

diff --git a/cassis/xmi.py b/cassis/xmi.py
@@ -10,7 +10,9 @@
 
 from cassis.cas import Cas, IdGenerator, Sofa, View
 from cassis.typesystem import (
+    _LIST_TYPES,
     _PRIMITIVE_ARRAY_TYPES,
+    _PRIMITIVE_LIST_TYPES,
     FEATURE_BASE_NAME_BEGIN,
     FEATURE_BASE_NAME_END,
     FEATURE_BASE_NAME_HEAD,
@@ -26,6 +28,7 @@
     TYPE_NAME_EMPTY_FLOAT_LIST,
     TYPE_NAME_EMPTY_FS_LIST,
     TYPE_NAME_EMPTY_INTEGER_LIST,
+    TYPE_NAME_EMPTY_STRING_LIST,
     TYPE_NAME_FLOAT,
     TYPE_NAME_FLOAT_ARRAY,
     TYPE_NAME_FLOAT_LIST,
@@ -39,16 +42,17 @@
     TYPE_NAME_NON_EMPTY_FLOAT_LIST,
     TYPE_NAME_NON_EMPTY_FS_LIST,
     TYPE_NAME_NON_EMPTY_INTEGER_LIST,
+    TYPE_NAME_NON_EMPTY_STRING_LIST,
     TYPE_NAME_SHORT,
     TYPE_NAME_SHORT_ARRAY,
     TYPE_NAME_SOFA,
     TYPE_NAME_STRING,
     TYPE_NAME_STRING_ARRAY,
+    TYPE_NAME_STRING_LIST,
     FeatureStructure,
     Type,
     TypeNotFoundError,
-    TypeSystem, TYPE_NAME_STRING_LIST, TYPE_NAME_EMPTY_STRING_LIST, TYPE_NAME_NON_EMPTY_STRING_LIST,
-    _PRIMITIVE_LIST_TYPES, _LIST_TYPES,
+    TypeSystem,
 )
 
 NAN_VALUE = "NaN"
@@ -244,7 +248,9 @@ def deserialize(self, source: Union[IO, str], typesystem: TypeSystem, lenient: b
                         continue
 
                     # Resolve references
-                    if fs.type.name == TYPE_NAME_FS_ARRAY or feature.rangeType.name == TYPE_NAME_FS_ARRAY:
+                    if fs.type.name == TYPE_NAME_FS_ARRAY or (
+                        feature.rangeType.name == TYPE_NAME_FS_ARRAY and not feature.multipleReferencesAllowed
+                    ):
                         # An array of references is a list of integers separated
                         # by single spaces, e.g. <foo:bar elements="1 2 3 42" />
                         targets = []
@@ -627,7 +633,9 @@ def _serialize_feature_structure(self, cas: Cas, root: etree.Element, fs: Featur
                     elem.attrib[feature_name] = " ".join(str(e.xmiID) for e in value.elements)
             elif feature.rangeType.name == TYPE_NAME_FS_LIST and not feature.multipleReferencesAllowed:
                 if value is not None:  # Compare to none to not skip if elements is empty!
-                    elem.attrib[feature_name] = " ".join(str(e.xmiID) for e in self._collect_list_elements(feature.rangeType.name, value))
+                    elem.attrib[feature_name] = " ".join(
+                        str(e.xmiID) for e in self._collect_list_elements(feature.rangeType.name, value)
+                    )
             elif feature_name == FEATURE_BASE_NAME_SOFA:
                 elem.attrib[feature_name] = str(value.xmiID)
             elif feature.rangeType.name == TYPE_NAME_BOOLEAN:
@@ -659,7 +667,7 @@ def _serialize_view(self, root: etree.Element, view: View):
         elem.attrib["sofa"] = str(view.sofa.xmiID)
         elem.attrib["members"] = " ".join(sorted((str(x.xmiID) for x in view.get_all_annotations()), key=int))
 
-    def _collect_list_elements(self, type_name: str,  value) -> List[str]:
+    def _collect_list_elements(self, type_name: str, value) -> List[str]:
         if type_name not in _LIST_TYPES:
             raise ValueError(f"Not a primitive list: {type_name}")
 

diff --git a/tests/fixtures.py b/tests/fixtures.py
@@ -76,6 +76,20 @@ def cas_with_list_features_xmi(cas_with_list_features_path):
         return f.read()
 
 
+# CAS with all kinds of array features
+
+
+@pytest.fixture
+def cas_with_array_features_path():
+    return os.path.join(FIXTURE_DIR, "xmi", "cas_with_array_features.xmi")
+
+
+@pytest.fixture
+def cas_with_array_features_xmi(cas_with_array_features_path):
+    with open(cas_with_array_features_path, "r") as f:
+        return f.read()
+
+
 # CAS with references
 
 
@@ -406,6 +420,20 @@ def typesystem_with_list_features_xml(typesystem_with_list_features_path):
         return f.read()
 
 
+# Type system with all kinds of array features
+
+
+@pytest.fixture
+def typesystem_with_array_features_path():
+    return os.path.join(FIXTURE_DIR, "typesystems", "typesystem_with_array_features.xml")
+
+
+@pytest.fixture
+def typesystem_with_array_features_xml(typesystem_with_array_features_path):
+    with open(typesystem_with_array_features_path, "r") as f:
+        return f.read()
+
+
 # Annotations
 
 

diff --git a/tests/test_files/typesystems/typesystem_with_array_features.xml b/tests/test_files/typesystems/typesystem_with_array_features.xml
@@ -0,0 +1,117 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<typeSystemDescription xmlns="http://uima.apache.org/resourceSpecifier">
+    <types>
+        <typeDescription>
+            <name>uima.tcas.DocumentAnnotation</name>
+            <description/>
+            <supertypeName>uima.tcas.Annotation</supertypeName>
+            <features>
+                <featureDescription>
+                    <name>language</name>
+                    <description/>
+                    <rangeTypeName>uima.cas.String</rangeTypeName>
+                </featureDescription>
+            </features>
+        </typeDescription>
+        <typeDescription>
+            <name>FloatArrayHolder</name>
+            <description/>
+            <supertypeName>uima.cas.TOP</supertypeName>
+            <features>
+                <featureDescription>
+                    <name>floatArray</name>
+                    <description/>
+                    <rangeTypeName>uima.cas.FloatArray</rangeTypeName>
+                </featureDescription>
+            </features>
+        </typeDescription>
+        <typeDescription>
+            <name>FloatArrayHolderMR</name>
+            <description/>
+            <supertypeName>uima.cas.TOP</supertypeName>
+            <features>
+                <featureDescription>
+                    <name>floatArray</name>
+                    <description/>
+                    <rangeTypeName>uima.cas.FloatArray</rangeTypeName>
+                    <multipleReferencesAllowed>true</multipleReferencesAllowed>
+                </featureDescription>
+            </features>
+        </typeDescription>
+        <typeDescription>
+            <name>FsArrayHolder</name>
+            <description/>
+            <supertypeName>uima.cas.TOP</supertypeName>
+            <features>
+                <featureDescription>
+                    <name>fsArray</name>
+                    <description/>
+                    <rangeTypeName>uima.cas.FSArray</rangeTypeName>
+                </featureDescription>
+            </features>
+        </typeDescription>
+        <typeDescription>
+            <name>FsArrayHolderMR</name>
+            <description/>
+            <supertypeName>uima.cas.TOP</supertypeName>
+            <features>
+                <featureDescription>
+                    <name>fsArray</name>
+                    <description/>
+                    <rangeTypeName>uima.cas.FSArray</rangeTypeName>
+                    <multipleReferencesAllowed>true</multipleReferencesAllowed>
+                </featureDescription>
+            </features>
+        </typeDescription>
+        <typeDescription>
+            <name>IntArrayHolder</name>
+            <description/>
+            <supertypeName>uima.cas.TOP</supertypeName>
+            <features>
+                <featureDescription>
+                    <name>intArray</name>
+                    <description/>
+                    <rangeTypeName>uima.cas.IntegerArray</rangeTypeName>
+                </featureDescription>
+            </features>
+        </typeDescription>
+        <typeDescription>
+            <name>IntArrayHolderMR</name>
+            <description/>
+            <supertypeName>uima.cas.TOP</supertypeName>
+            <features>
+                <featureDescription>
+                    <name>intArray</name>
+                    <description/>
+                    <rangeTypeName>uima.cas.IntegerArray</rangeTypeName>
+                    <multipleReferencesAllowed>true</multipleReferencesAllowed>
+                </featureDescription>
+            </features>
+        </typeDescription>
+        <typeDescription>
+            <name>StringArrayHolder</name>
+            <description/>
+            <supertypeName>uima.cas.TOP</supertypeName>
+            <features>
+                <featureDescription>
+                    <name>stringArray</name>
+                    <description/>
+                    <rangeTypeName>uima.cas.StringArray</rangeTypeName>
+                </featureDescription>
+            </features>
+        </typeDescription>
+        <typeDescription>
+            <name>StringArrayHolderMR</name>
+            <description/>
+            <supertypeName>uima.cas.TOP</supertypeName>
+            <features>
+                <featureDescription>
+                    <name>stringArray</name>
+                    <description/>
+                    <rangeTypeName>uima.cas.StringArray</rangeTypeName>
+                    <multipleReferencesAllowed>true</multipleReferencesAllowed>
+                </featureDescription>
+            </features>
+        </typeDescription>
+    </types>
+</typeSystemDescription>
diff --git a/tests/test_files/xmi/cas_with_array_features.xmi b/tests/test_files/xmi/cas_with_array_features.xmi
@@ -0,0 +1,26 @@
+<?xml version="1.0" encoding="UTF-8"?><xmi:XMI xmlns:noNamespace="http:///uima/noNamespace.ecore" xmlns:tcas="http:///uima/tcas.ecore" xmlns:xmi="http://www.omg.org/XMI" xmlns:cas="http:///uima/cas.ecore" xmi:version="2.0">
+    <cas:NULL xmi:id="0"/>
+    <cas:FSArray xmi:id="4" elements="2 3"/>
+    <noNamespace:FsArrayHolder xmi:id="6" fsArray="2 3"/>
+    <noNamespace:FsArrayHolderMR xmi:id="8" fsArray="7"/>
+    <noNamespace:IntArrayHolder xmi:id="10" intArray="1 2"/>
+    <noNamespace:IntArrayHolderMR xmi:id="12" intArray="11"/>
+    <noNamespace:FloatArrayHolder xmi:id="14" floatArray="-1.0 0.0 1.0 NaN -Infinity Infinity 1.4E-45 3.4028235E38"/>
+    <noNamespace:FloatArrayHolderMR xmi:id="16" floatArray="15"/>
+    <noNamespace:StringArrayHolder xmi:id="18">
+        <stringArray>blah</stringArray>
+        <stringArray>blub</stringArray>
+    </noNamespace:StringArrayHolder>
+    <noNamespace:StringArrayHolderMR xmi:id="20" stringArray="19"/>
+    <cas:Sofa xmi:id="1" sofaNum="1" sofaID="_InitialView"/>
+    <tcas:Annotation xmi:id="2" sofa="1" begin="0" end="1"/>
+    <tcas:Annotation xmi:id="3" sofa="1" begin="1" end="2"/>
+    <cas:FSArray xmi:id="7" elements="2 3"/>
+    <cas:IntegerArray xmi:id="11" elements="1 2"/>
+    <cas:FloatArray xmi:id="15" elements="-1.0 0.0 1.0 NaN -Infinity Infinity 1.4E-45 3.4028235E38"/>
+    <cas:StringArray xmi:id="19">
+        <elements>blah</elements>
+        <elements>blub</elements>
+    </cas:StringArray>
+    <cas:View sofa="1" members="4 6 8 10 12 14 16 18 20"/>
+</xmi:XMI>
diff --git a/tests/test_xmi.py b/tests/test_xmi.py
@@ -39,6 +39,10 @@
         pytest.lazy_fixture("cas_with_list_features_xmi"),
         pytest.lazy_fixture("typesystem_with_list_features_xml"),
     ),
+    (
+        pytest.lazy_fixture("cas_with_array_features_xmi"),
+        pytest.lazy_fixture("typesystem_with_array_features_xml"),
+    ),
 ]
Original file line number	Diff line number	Diff line change
Expand Up		@@ -189,6 +189,7 @@

		_LIST_TYPES = _PRIMITIVE_LIST_TYPES \| {TYPE_NAME_FS_LIST}


		class TypeSystemMode(Enum):
		"""How much type system information to include."""

Expand Down