chore: XML validation: resource's iri and ark must be unique, resourc…

…e's label must not be empty (DEV-2154) (#361) In the XML file, it is not allowed anymore to have a resource with - label="" - ark/iri attribute that are not unique in the document
dasch-swiss · May 12, 2023 · 62ec04b · 62ec04b
1 parent a93cb86
commit 62ec04b
Show file tree

Hide file tree

Showing 6 changed files with 123 additions and 2 deletions.
diff --git a/src/dsp_tools/resources/schema/data.xsd b/src/dsp_tools/resources/schema/data.xsd
@@ -407,7 +407,13 @@
                 <xs:element name="boolean-prop" type="booleanprop_type" minOccurs="0" maxOccurs="unbounded"/>
             </xs:choice>
         </xs:sequence>
-        <xs:attribute name="label" type="xs:string" use="required"/>
+        <xs:attribute name="label" use="required">
+            <xs:simpleType>
+                <xs:restriction base="xs:string">
+                    <xs:minLength value="1"/>
+                </xs:restriction>
+            </xs:simpleType>
+        </xs:attribute>
         <xs:attribute name="restype" type="xs:string" use="required"/>
         <xs:attribute name="id" type="xs:ID" use="required"/>
         <xs:attribute name="permissions" type="xs:NCName" use="optional"/>
@@ -538,6 +544,13 @@
             <xs:selector xpath="resource"/>
             <xs:field xpath="@permissions"/>
         </xs:keyref>
+        <xs:unique name="IRI_attribute_of_resource_must_be_unique">
+            <xs:selector xpath=".//*"/>
+            <xs:field xpath="@iri"/>
+        </xs:unique>
+        <xs:unique name="ARK_attribute_of_resource_must_be_unique">
+            <xs:selector xpath=".//*"/>
+            <xs:field xpath="@ark"/>
+        </xs:unique>
     </xs:element>
 </xs:schema>
-
diff --git a/test/unittests/test_shared.py b/test/unittests/test_shared.py
@@ -36,6 +36,24 @@ def test_validate_xml_against_schema(self) -> None:
         ):
             shared.validate_xml_against_schema(input_file="testdata/invalid-testdata/xml-data/utf8-text-with-xml-tags.xml")
 
+        with self.assertRaisesRegex(
+            UserError,
+            "Line 19: Element 'resource': Duplicate key-sequence .+ in unique identity-constraint 'IRI_attribute_of_resource_must_be_unique'"
+        ):
+            shared.validate_xml_against_schema(input_file="testdata/invalid-testdata/xml-data/duplicate-iri.xml")
+
+        with self.assertRaisesRegex(
+            UserError,
+            "Line 19: Element 'resource': Duplicate key-sequence .+ in unique identity-constraint 'ARK_attribute_of_resource_must_be_unique'"
+        ):
+            shared.validate_xml_against_schema(input_file="testdata/invalid-testdata/xml-data/duplicate-ark.xml")
+
+        with self.assertRaisesRegex(
+            UserError,
+            "Line 11: Element 'resource', attribute 'label': .+ The value '' has a length of '0'; this underruns the allowed minimum length of '1'"
+        ):
+            shared.validate_xml_against_schema(input_file="testdata/invalid-testdata/xml-data/empty-label.xml")
+
 
     def test_validate_xml_tags_in_text_properties(self) -> None:
         utf8_texts_with_allowed_html_escapes = [

diff --git a/testdata/invalid-testdata/xml-data/duplicate-ark.xml b/testdata/invalid-testdata/xml-data/duplicate-ark.xml
@@ -0,0 +1,23 @@
+<?xml version='1.0' encoding='utf-8'?>
+
+<knora xmlns="https://dasch.swiss/schema"
+       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+       xsi:schemaLocation="https://dasch.swiss/schema ../../../src/dsp_tools/resources/schema/data.xsd"
+       shortcode="4124"
+       default-ontology="testonto">
+
+    <resource label="TestthingWithArk1"
+              restype=":TestThing2"
+              id="test_thing_with_ark_1"
+              ark="ark:/72163/4123-31ec6eab334-a.2022829">
+        <text-prop name=":hasSimpleText"><text encoding="utf8">Hello</text></text-prop>
+    </resource>
+
+    <resource label="TestthingWithArk2"
+              restype=":TestThing2"
+              id="test_thing_with_ark_2"
+              ark="ark:/72163/4123-31ec6eab334-a.2022829">
+        <text-prop name=":hasSimpleText"><text encoding="utf8">Hello</text></text-prop>
+    </resource>
+
+</knora>
diff --git a/testdata/invalid-testdata/xml-data/duplicate-iri.xml b/testdata/invalid-testdata/xml-data/duplicate-iri.xml
@@ -0,0 +1,23 @@
+<?xml version='1.0' encoding='utf-8'?>
+
+<knora xmlns="https://dasch.swiss/schema"
+       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+       xsi:schemaLocation="https://dasch.swiss/schema ../../../src/dsp_tools/resources/schema/data.xsd"
+       shortcode="4124"
+       default-ontology="testonto">
+
+    <resource label="TestthingWithIri1"
+              restype=":TestThing2"
+              id="test_thing_with_iri_1"
+              iri="http://rdfh.ch/4123/54SYvWF0QUW6a">
+        <text-prop name=":hasSimpleText"><text encoding="utf8">Hello</text></text-prop>
+    </resource>
+
+    <resource label="TestthingWithIri2"
+              restype=":TestThing2"
+              id="test_thing_with_iri_2"
+              iri="http://rdfh.ch/4123/54SYvWF0QUW6a">
+        <text-prop name=":hasSimpleText"><text encoding="utf8">Hello</text></text-prop>
+    </resource>
+
+</knora>
diff --git a/testdata/invalid-testdata/xml-data/empty-label.xml b/testdata/invalid-testdata/xml-data/empty-label.xml
@@ -0,0 +1,15 @@
+<?xml version='1.0' encoding='utf-8'?>
+
+<knora xmlns="https://dasch.swiss/schema"
+       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+       xsi:schemaLocation="https://dasch.swiss/schema ../../../src/dsp_tools/resources/schema/data.xsd"
+       shortcode="4124"
+       default-ontology="testonto">
+
+    <resource label=""
+              restype=":TestThing2"
+              id="test_thing_with_empty_label">
+        <text-prop name=":hasSimpleText"><text encoding="utf8">Hello</text></text-prop>
+    </resource>
+
+</knora>
diff --git a/testdata/xml-data/test-data-systematic.xml b/testdata/xml-data/test-data-systematic.xml
@@ -804,4 +804,33 @@
             <text encoding="utf8">Hello!</text>
         </text-prop>
     </resource>
+
+    <resource label="TestthingWithArk1"
+              restype=":TestThing2"
+              id="test_thing_with_ark_1"
+              ark="ark:/72163/4123-43ec6evb134-a.2022828">
+        <text-prop name=":hasSimpleText"><text encoding="utf8">Hello</text></text-prop>
+    </resource>
+
+    <resource label="TestthingWithArk2"
+              restype=":TestThing2"
+              id="test_thing_with_ark_2"
+              ark="ark:/72163/4123-43xc6ivb931-a.2022829">
+        <text-prop name=":hasSimpleText"><text encoding="utf8">Hello</text></text-prop>
+    </resource>
+
+    <resource label="TestthingWithIri1"
+              restype=":TestThing2"
+              id="test_thing_with_iri_1"
+              iri="http://rdfh.ch/4123/DiAmYQzQSzC7cdTo6OJMYA">
+        <text-prop name=":hasSimpleText"><text encoding="utf8">Hello</text></text-prop>
+    </resource>
+
+    <resource label="TestthingWithIri2"
+              restype=":TestThing2"
+              id="test_thing_with_iri_2"
+              iri="http://rdfh.ch/4123/TqAnYQzrSzC2ctT06OJMYB">
+        <text-prop name=":hasSimpleText"><text encoding="utf8">Hello</text></text-prop>
+    </resource>
+
 </knora>