Merge pull request #110 from MaRDI4NFDI/OpenAire_for_Software

Open aire for software metadata (unittest - reference)

test/data/software/OpenAire-software-reference.xml

@@ -0,0 +1,65 @@
+<resource xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+          xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+          xmlns:dc="http://purl.org/dc/elements/1.1/"
+          xmlns:dcterms="http://purl.org/dc/terms/"
+          xmlns:datacite="http://datacite.org/schema/kernel-4"
+          xmlns:oaire="http://www.openarchives.org/OAI/2.0/oai_dc/"
+          xmlns="http://namespace.openaire.eu/schema/oaire/"
+          xsi:schemaLocation="http://namespace.openaire.eu/schema/oaire/ https://www.openaire.eu/schema/repo-lit/4.0/openaire.xsd">
+   <datacite:identifier IdentifierType="URL">https://zbmath.org/software/2</datacite:identifier>
+   <datacite:creators>
+      <datacite:creator>
+         <datacite:creatorName nameType="Personal">Li, X. S.</datacite:creatorName>
+         <datacite:givenName>X. S.</datacite:givenName>
+         <datacite:familyName>Li</datacite:familyName>
+      </datacite:creator>
+      <datacite:creator>
+         <datacite:creatorName nameType="Personal">Demmel, J. W.</datacite:creatorName>
+         <datacite:givenName>J. W.</datacite:givenName>
+         <datacite:familyName>Demmel</datacite:familyName>
+      </datacite:creator>
+   </datacite:creators>
+   <datacite:titles>
+      <datacite:title>SuperLU-DIST</datacite:title>
+   </datacite:titles>
+   <dc:description xml:lang="en" descriptionType="Abstract">zbMATH Open Web Interface contents unavailable due to conflicting licenses.</dc:description>
+   <datacite:subjects>
+      <datacite:subject subjectScheme="msc2020">35</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">65</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">68</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">74</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">76</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">05</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">15</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">60</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">78</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">80</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">81</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">82</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">85</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">86</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">90</datacite:subject>
+      <datacite:subject subjectScheme="msc2020">92</datacite:subject>
+      <datacite:subject subjectScheme="keyword">orms</datacite:subject>
+   </datacite:subjects>
+   <datacite:relatedIdentifiers>
+      <datacite:relatedIdentifier relatedIdentifierType="URL" relationType="IsSourceOf">http://crd-legacy.lbl.gov/~xiaoye/SuperLU/</datacite:relatedIdentifier>
+      <datacite:relatedIdentifier relatedIdentifierType="URL" relationType="IsCitedBy">https://zbmath.org/2187846</datacite:relatedIdentifier>
+   </datacite:relatedIdentifiers>
+   <resourceType resourceTypeGeneral="Software"/>
+   <datacite:rightsList>
+      <datacite:rights rightsURI="http://purl.org/coar/access_right/c_14cb">metadata only access</datacite:rights>
+      <datacite:rights xml:lang="en"
+                       schemeURI="https://api.zbmath.org/v1/"
+                       rightsIdentifierScheme="zbMATH"
+                       rightsIdentifier="CC-BY-SA 4.0"
+                       rightsURI="https://creativecommons.org/licenses/by-sa/4.0/">
+                    The zbMATH Open OAI-PMH API is subject to the Terms and Conditions for the zbMATH Open API Service of FIZ Karlsruhe – Leibniz-Institut für Informationsinfrastruktur GmbH.
+Content generated by zbMATH Open, such as reviews, classifications, software, or author disambiguation data, are distributed under CC-BY-SA 4.0. This defines the license for the whole dataset, which also contains non-copyrighted bibliographic metadata and reference data derived from I4OC (CC0). Note that the API only provides a subset of the data in the zbMATH Open Web interface. In several cases, third-party information, such as abstracts, cannot be made available under a suitable license through the API. In those cases, we replaced the data with the string
+"zbMATH Open Web Interface contents unavailable due to conflicting licenses."</datacite:rights>
+   </datacite:rightsList>
+   <dc:format>application/xml</dc:format>
+   <dc:language>eng</dc:language>
+   <dc:source>zbMATH Open Web Interface contents unavailable due to conflicting licenses.</dc:source>
+   <oaire:citationTitle>zbMATH Open Web Interface contents unavailable due to conflicting licenses.</oaire:citationTitle>
+</resource>

test/test_metadata-software_OpenAire.py

@@ -0,0 +1,53 @@
+import os
+import re
+import unittest
+
+import lxml.etree as ET
+from xmldiff import main
+from xmldiff.actions import MoveNode
+
+if os.path.basename(os.getcwd()) == 'test':
+    os.chdir(os.path.dirname(os.getcwd()))
+
+
+# Function to split DOI into parts
+def split_doi(doi):
+    # Define regular expression patterns for various DOI formats
+    doi_patterns = [
+        r"(10\.\d{4,}/\S+)",
+        r"(/?\S+)",  # For DOIs without prefix
+    ]
+
+    # Iterate through patterns to find a match
+    for pattern in doi_patterns:
+        match = re.match(pattern, doi)
+        if match:
+            return match.group(1), doi.replace(match.group(1), '', 1).lstrip('/')
+    return None, None
+
+
+class PlainXmlTest(unittest.TestCase):
+    def test_similarity(self):
+        dom = ET.parse('test/data/software/plain.xml')
+        xslt = ET.parse('xslt/software/xslt-software-OpenAire.xslt')
+
+        transform = ET.XSLT(xslt)  # is it a reserved word
+        newdom = transform(dom)
+
+        real_string = ET.tostring(newdom, pretty_print=True, encoding='utf8').decode()
+        # test if result is parsable
+        reference = ET.parse('test/data/software/OpenAire-software-reference.xml')
+
+        expected_string = ET.tostring(reference, pretty_print=True, encoding='utf8').decode()
+        diff = main.diff_texts(expected_string, real_string, {
+            'ratio_mode': 'fast',  # is that for latency
+            'F': 1,
+        })
+
+        essentials = list(filter(lambda e: not isinstance(e, MoveNode), diff))
+
+        self.assertLessEqual(len(essentials), 0, "Found differences between expected and transformed XML")
+
+
+if __name__ == '__main__':
+    unittest.main()