UFAL/Add CCMM 1.1.0 OAI-PMH crosswalk (ccmm-xml metadataPrefix) by milanmajchrak · Pull Request #1262 · dataquest-dev/DSpace

milanmajchrak · 2026-02-27T12:48:50Z

Problem description

Add CCMM 1.1.0 OAI-PMH crosswalk — exports metadata as CCMM 1.1.0 over OAI-PMH (metadataPrefix: ccmm-xml), required for NMD/NRP.

New ccmm.xsl, registered in xoai.xml (Default + openaire_data context), plus unit tests and test data.

Analysis

Run over all 2 628 live LINDAT records and validated against CCMM 1.1.0 with XSD 1.1:

	before	after
records that transform	2 203	2 628
`ListRecords` pages a harvester can fetch	0 / 32	32 / 32
XSD 1.1 valid	2 203	2 628
`access_rights` correct	746 (28 %)	2 628 (100 %)
`publication_year` = `0000`	587	0
distinct IRIs returning HTTP 404	16	0

Transform error — doc:field[@name='value'] is a sequence; the 425 records that carry two language variants of a field threw XPTY0004. OAIException turns into HTTP 500 for the whole page, so every one of the 32 pages failed and nothing was harvestable. Fixed with (…)[1].
Vocabularies — every model.ccmm.cz/vocabulary/… IRI returns 404. Replaced with the registers CCMM names in its own specification (en/dsv.ttl scope notes): COAR resource types and access rights, vocabs.ccmm.cz AgentRole / TimeReference / DescriptionType, EU Publications Office for languages. dc.publisher → AgentRole/Publisher (Distributor does not exist at that level).
Access rights — now read from others/access-status, which DSpace computes in DefaultAccessStatusHelper. The old branches were dead (local.embargo.termslift: 0 records) or inverted (others/@restrictedAccess: 24 records, all of them open).
Dates — dc.date.issued = "0000" plus local.approximateDate.issued are handled the way ClarinDateService does in the UI: the local field replaces dc.date.issued, a range becomes a time_interval, and the original string is kept in date_information. Exactly one Issued reference plus a Created one, as the CCMM profile requires.
Licence — no more invented IRI; falls back to others/cc, otherwise <license/> stays empty and the wording moves to terms_of_use/description.

Problems

CCMM requires at least one subject from the FRASCATI FORD vocabulary. LINDAT has no FORD field, so no record satisfies that yet — needs a submission-form change first.

Not in this PR: distribution from bundles/bitstreams, funding_reference from local.sponsor (616 records vs the 103 currently covered), location from dc.coverage.placeName.

Resolves: ufal#1145

Manual Testing (if applicable)

Added to testing scenarios

Copilot review

Requested review from Copilot

Copilot

Pull request overview

This PR implements a new OAI-PMH crosswalk for exporting DSpace metadata in CCMM (Czech Common Metadata Model) 1.1.0 format, required for NMD/NRP (National Repository of Gray Literature) compliance in the Czech Republic. The implementation follows the standard DSpace OAI-PMH crosswalk pattern used by other metadata formats in the repository.

Changes:

New XSL transformation stylesheet mapping DSpace internal metadata to CCMM 1.1.0 XML schema
Configuration updates to register the new crosswalk in both Default Context and openaire_data Context
Comprehensive test suite with 14 unit tests and dedicated test data file

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 14 comments.

File	Description
dspace/config/crosswalks/oai/xoai.xml	Registers ccmm format (metadataPrefix: ccmm-xml) in Default and openaire_data contexts
dspace/config/crosswalks/oai/metadataFormats/ccmm.xsl	633-line XSLT 2.0 stylesheet implementing DSpace to CCMM metadata transformation with fallback handling for required fields
dspace-oai/src/test/java/org/dspace/xoai/tests/stylesheets/CcmmXslTest.java	Test class with 14 tests validating core CCMM elements (identifiers, creators, subjects, license, etc.)
dspace-oai/src/test/resources/xoai-ccmm-test.xml	Test fixture simulating typical NMD/NRP dataset metadata with Czech academic repository data

Copilot · 2026-02-27T12:59:03Z

+                                </xsl:otherwise>
+                            </xsl:choose>
+                        </xsl:when>
+                        <xsl:otherwise>http://unknown.repository</xsl:otherwise>


The hardcoded string "http://unknown.repository" should be extracted as a constant or defined at the top of the stylesheet for better maintainability.

Copilot · 2026-02-27T12:59:03Z

+                                <xsl:when test="contains($uri, '/handle/')">
+                                    <xsl:value-of select="substring-before($uri, '/handle/')"/>


The logic for extracting the repository base URL by finding '/handle/' in the URI and taking everything before it (lines 114-115) may not work correctly for all Handle server configurations. Some repositories might use different URL patterns (e.g., using a different path segment). Consider adding a comment explaining the assumptions about Handle URL structure, or making this more robust by also checking for 'hdl.handle.net' pattern.

Copilot · 2026-02-27T12:59:06Z

+    @Test
+    public void ccmmCanTransformBasicXoaiInput() throws Exception {
+        // Test with the default xoai-test1.xml (simpler data) to ensure crosswalk
+        // handles missing fields gracefully
+        String result = apply("ccmm.xsl").to(resource("xoai-test1.xml"));
+        assertThat(result, is(ccmm().withXPath("//ccmm:title", equalTo("Test Webpage"))));
+    }


The test suite lacks coverage for the fallback scenarios where required fields are missing. Consider adding tests that verify the behavior when title, identifiers, publication year, time references, and subjects are missing to ensure fallback values (like "Untitled", "9999", etc.) are correctly produced. This is particularly important for required CCMM fields.

- Extract hardcoded fallback strings as XSLT variables at top of stylesheet (Unknown Repository, http://unknown.repository, Untitled, unspecified, 9999) - Add detailed comment explaining Handle URL extraction assumptions - Add comment explaining publisher to Distributor role mapping rationale - Fix time_reference fallback: use dc.date.accessioned instead of hardcoded 9999-01-01 when issued/available dates are missing - Improve FormatDate fallback: validate first 4 chars are digits before using as year, output empty ccmm:date element for invalid input - Improve identifier fallback logic: check if Handle-pattern URIs were actually found rather than just checking for raw element presence - Add 6 new fallback scenario tests (title, publication_year, subject, identifier, repository name, license) with dedicated minimal test fixture - All 21 tests pass (20 CCMM + 1 QDC)

Copilot

Pull request overview

Copilot reviewed 5 out of 5 changed files in this pull request and generated 10 comments.

Copilot · 2026-02-27T14:41:25Z

+                        <ccmm:date><xsl:value-of select="concat($year, '-01-01')"/></ccmm:date>
+                    </xsl:when>
+                    <xsl:otherwise>
+                        <ccmm:date/>


The FormatDate template creates an empty ccmm:date element at line 660 when the date string is invalid. This produces invalid XML output like 'ccmm:date/' or 'ccmm:date</ccmm:date>' which may not validate against the CCMM schema. Consider either omitting the element entirely or providing a fallback value, or documenting that this is intentional if the schema permits empty date elements.

Suggested change

<ccmm:date/>

Copilot · 2026-02-27T14:41:25Z

+        <xsl:variable name="lang" select="doc:metadata/doc:element[@name='dc']/doc:element[@name='language']/doc:element[@name='iso']/doc:element/doc:field[@name='value']"/>
+        <xsl:variable name="langAlt" select="doc:metadata/doc:element[@name='dc']/doc:element[@name='language']/doc:element/doc:field[@name='value']"/>
+        <xsl:choose>
+            <xsl:when test="$lang">
+                <ccmm:primary_language>
+                    <ccmm:iri>
+                        <xsl:value-of select="concat('https://iso639-3.sil.org/code/', $lang)"/>
+                    </ccmm:iri>
+                    <ccmm:label xml:lang="en"><xsl:value-of select="$lang"/></ccmm:label>


The PrimaryLanguage template could be clearer by explicitly selecting only the first language value with '[1]' predicate. Currently, line 440 selects all dc.language.iso values, and XSLT's atomization implicitly uses the first one at line 446. While functionally correct, explicitly selecting '[1]' at line 440 would make the intent clearer and match the pattern used in OtherLanguages at line 468 where position() > 1 explicitly excludes the first.

Copilot · 2026-02-27T14:41:26Z

+                <ccmm:iri>
+                    <xsl:choose>
+                        <xsl:when test="doc:metadata/doc:element[@name='dc']/doc:element[@name='identifier']/doc:element[@name='uri']/doc:element/doc:field[@name='value']">
+                            <xsl:variable name="uri" select="doc:metadata/doc:element[@name='dc']/doc:element[@name='identifier']/doc:element[@name='uri']/doc:element/doc:field[@name='value'][1]"/>
+                            <!--
+                                Extract repository base URL from Handle URI.
+                                Assumes DSpace-style URLs like https://repo.example.org/handle/123/456.
+                                Also handles hdl.handle.net URIs as-is.
+                            -->
+                            <xsl:choose>
+                                <xsl:when test="contains($uri, '/handle/')">
+                                    <xsl:value-of select="substring-before($uri, '/handle/')"/>
+                                </xsl:when>
+                                <xsl:otherwise>
+                                    <xsl:value-of select="$uri"/>
+                                </xsl:otherwise>
+                            </xsl:choose>
+                        </xsl:when>
+                        <xsl:otherwise><xsl:value-of select="$FALLBACK_REPOSITORY_URL"/></xsl:otherwise>
+                    </xsl:choose>


The original_repository IRI extraction logic may produce incorrect results when dc.identifier.uri contains non-Handle URIs. Lines 127-132 will output the full URI (e.g., a DOI URL) as the repository IRI if it doesn't contain '/handle/'. Consider filtering to only Handle URIs (checking for 'hdl.handle.net' or '/handle/'), or falling back to FALLBACK_REPOSITORY_URL for non-Handle URIs, similar to the Identifiers template logic at lines 152-153.

Copilot · 2026-02-27T14:41:26Z

+        <xsl:variable name="dctype" select="doc:metadata/doc:element[@name='dc']/doc:element[@name='type']/doc:element/doc:field[@name='value']"/>
+        <xsl:if test="$dctype">
+            <ccmm:resource_type>
+                <ccmm:iri>
+                    <xsl:choose>
+                        <xsl:when test="$dctype='Dataset' or $dctype='corpus' or $dctype='lexicalConceptualResource' or $dctype='languageDescription'">https://model.ccmm.cz/vocabulary/datacite/resourceTypeGeneral/Dataset</xsl:when>
+                        <xsl:when test="$dctype='Software' or $dctype='toolService'">https://model.ccmm.cz/vocabulary/datacite/resourceTypeGeneral/Software</xsl:when>
+                        <xsl:when test="$dctype='Text' or $dctype='text'">https://model.ccmm.cz/vocabulary/datacite/resourceTypeGeneral/Text</xsl:when>
+                        <xsl:when test="$dctype='Image' or $dctype='image'">https://model.ccmm.cz/vocabulary/datacite/resourceTypeGeneral/Image</xsl:when>
+                        <xsl:when test="$dctype='Collection'">https://model.ccmm.cz/vocabulary/datacite/resourceTypeGeneral/Collection</xsl:when>
+                        <xsl:otherwise>https://model.ccmm.cz/vocabulary/datacite/resourceTypeGeneral/Other</xsl:otherwise>
+                    </xsl:choose>
+                </ccmm:iri>
+                <ccmm:label xml:lang="en"><xsl:value-of select="$dctype"/></ccmm:label>
+            </ccmm:resource_type>


The ResourceType template at line 418 selects all dc.type values into $dctype variable, but only the first value will be used (due to XSLT atomization in the when tests). If multiple dc.type values exist, subsequent values are silently ignored. For clarity, explicitly select only the first value with '[1]' predicate, similar to the pattern used elsewhere in the stylesheet.

Copilot · 2026-02-27T14:41:26Z

+                <ccmm:funding_reference>
+                    <xsl:if test="count($parts) >= 5">
+                        <ccmm:local_identifier><xsl:value-of select="$parts[5]"/></ccmm:local_identifier>
+                    </xsl:if>
+                    <ccmm:funder>
+                        <ccmm:organization>
+                            <ccmm:name>
+                                <xsl:choose>
+                                    <xsl:when test="$parts[3]='EC'">European Commission</xsl:when>
+                                    <xsl:otherwise><xsl:value-of select="$parts[3]"/></xsl:otherwise>
+                                </xsl:choose>
+                            </ccmm:name>
+                        </ccmm:organization>
+                    </ccmm:funder>
+                </ccmm:funding_reference>


The FundingReferences template lacks error handling for malformed grant agreement URIs. If a dc.relation value starts with 'info:' but doesn't have the expected structure (e.g., fewer than 5 parts after tokenizing by '/'), the code attempts to access $parts[3] and potentially $parts[5], which could produce empty values or runtime errors. Consider adding validation to ensure the URI has the expected number of parts before accessing array indices.

Suggested change

<ccmm:funding_reference>

<xsl:if test="count($parts) >= 5">

<ccmm:local_identifier><xsl:value-of select="$parts[5]"/></ccmm:local_identifier>

</xsl:if>

<ccmm:funder>

<ccmm:organization>

<ccmm:name>

<xsl:choose>

<xsl:when test="$parts[3]='EC'">European Commission</xsl:when>

<xsl:otherwise><xsl:value-of select="$parts[3]"/></xsl:otherwise>

</xsl:choose>

</ccmm:name>

</ccmm:organization>

</ccmm:funder>

</ccmm:funding_reference>

<xsl:variable name="parts-count" select="count($parts)"/>

<xsl:if test="$parts-count >= 3">

<ccmm:funding_reference>

<xsl:if test="$parts-count >= 5">

<ccmm:local_identifier><xsl:value-of select="$parts[5]"/></ccmm:local_identifier>

</xsl:if>

<ccmm:funder>

<ccmm:organization>

<ccmm:name>

<xsl:choose>

<xsl:when test="$parts[3]='EC'">European Commission</xsl:when>

<xsl:otherwise><xsl:value-of select="$parts[3]"/></xsl:otherwise>

</xsl:choose>

</ccmm:name>

</ccmm:organization>

</ccmm:funder>

</ccmm:funding_reference>

</xsl:if>

Copilot · 2026-02-27T14:41:26Z

+    <!-- ============================================================ -->
+    <!-- funding_reference (optional, unbounded)                       -->
+    <!-- ============================================================ -->
+    <xsl:template name="FundingReferences">
+        <!-- dc.relation with info:eu-repo/grantAgreement pattern -->
+        <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='relation']/doc:element/doc:field[@name='value']">
+            <xsl:if test="starts-with(., 'info:')">
+                <xsl:variable name="parts" select="tokenize(., '/')"/>
+                <ccmm:funding_reference>
+                    <xsl:if test="count($parts) >= 5">
+                        <ccmm:local_identifier><xsl:value-of select="$parts[5]"/></ccmm:local_identifier>
+                    </xsl:if>
+                    <ccmm:funder>
+                        <ccmm:organization>
+                            <ccmm:name>
+                                <xsl:choose>
+                                    <xsl:when test="$parts[3]='EC'">European Commission</xsl:when>
+                                    <xsl:otherwise><xsl:value-of select="$parts[3]"/></xsl:otherwise>
+                                </xsl:choose>
+                            </ccmm:name>
+                        </ccmm:organization>
+                    </ccmm:funder>
+                </ccmm:funding_reference>
+            </xsl:if>
+        </xsl:for-each>
+    </xsl:template>
+
+    <!-- ============================================================ -->
+    <!-- related_resource (optional, unbounded)                        -->
+    <!-- ============================================================ -->
+    <xsl:template name="RelatedResources">
+        <!-- dc.relation.uri -->
+        <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='relation']/doc:element[@name='uri']/doc:element/doc:field[@name='value']">
+            <ccmm:related_resource>
+                <ccmm:identifier>
+                    <ccmm:value><xsl:value-of select="."/></ccmm:value>
+                    <ccmm:scheme>
+                        <ccmm:iri>https://www.w3.org/ns/iana/uri-schemes</ccmm:iri>
+                        <ccmm:label xml:lang="en">URI</ccmm:label>
+                    </ccmm:scheme>
+                </ccmm:identifier>
+            </ccmm:related_resource>
+        </xsl:for-each>
+        <!-- dc.relation.ispartof -->
+        <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='relation']/doc:element[@name='ispartof']/doc:element/doc:field[@name='value']">
+            <ccmm:related_resource>
+                <ccmm:title><xsl:value-of select="."/></ccmm:title>
+            </ccmm:related_resource>
+        </xsl:for-each>
+    </xsl:template>


The FundingReferences and RelatedResources templates lack test coverage. No tests verify the parsing of grant agreement URIs (info:eu-repo/grantAgreement format) or the handling of dc.relation.uri and dc.relation.ispartof fields. Add test cases with sample funding data and related resource data to ensure these mappings work correctly.

Copilot · 2026-02-27T14:41:27Z

+    <xsl:template name="TermsOfUse">
+        <ccmm:terms_of_use>
+            <!-- access_rights -->
+            <ccmm:access_rights>
+                <xsl:choose>
+                    <xsl:when test="doc:metadata/doc:element[@name='local']/doc:element[@name='embargo']/doc:element[@name='termslift']/doc:element/doc:field[@name='value']">
+                        <ccmm:iri>http://purl.org/coar/access_right/c_f1cf</ccmm:iri>
+                        <ccmm:label xml:lang="en">embargoed access</ccmm:label>
+                    </xsl:when>
+                    <xsl:when test="doc:metadata/doc:element[@name='others']/doc:field[@name='restrictedAccess']/text()='true'">
+                        <ccmm:iri>http://purl.org/coar/access_right/c_16ec</ccmm:iri>
+                        <ccmm:label xml:lang="en">restricted access</ccmm:label>
+                    </xsl:when>
+                    <xsl:otherwise>
+                        <ccmm:iri>http://purl.org/coar/access_right/c_abf2</ccmm:iri>
+                        <ccmm:label xml:lang="en">open access</ccmm:label>
+                    </xsl:otherwise>
+                </xsl:choose>
+            </ccmm:access_rights>


The TermsOfUse template handles three access rights scenarios (embargoed, restricted, open), but only the default "open access" case is tested. Add test cases with local.embargo.termslift and others.restrictedAccess fields to verify embargoed and restricted access mappings produce correct COAR access rights IRIs.

Copilot · 2026-02-27T14:41:27Z

+    <xsl:template name="FormatDate">
+        <xsl:param name="dateStr"/>
+        <xsl:choose>
+            <!-- Full ISO dateTime: 2024-01-15T10:30:00Z -->
+            <xsl:when test="contains($dateStr, 'T')">
+                <ccmm:date_time><xsl:value-of select="$dateStr"/></ccmm:date_time>
+            </xsl:when>
+            <!-- Full date: 2024-01-15 -->
+            <xsl:when test="string-length($dateStr) >= 10 and substring($dateStr, 5, 1) = '-' and substring($dateStr, 8, 1) = '-'">
+                <ccmm:date><xsl:value-of select="substring($dateStr, 1, 10)"/></ccmm:date>
+            </xsl:when>
+            <!-- Year-month: 2024-01 -->
+            <xsl:when test="string-length($dateStr) >= 7 and substring($dateStr, 5, 1) = '-'">
+                <ccmm:date><xsl:value-of select="concat($dateStr, '-01')"/></ccmm:date>
+            </xsl:when>
+            <!-- Year only: 2024 -->
+            <xsl:when test="string-length($dateStr) = 4">
+                <ccmm:date><xsl:value-of select="concat($dateStr, '-01-01')"/></ccmm:date>
+            </xsl:when>
+            <!-- Fallback: attempt to derive a date only if the first 4 chars form a valid year -->
+            <xsl:otherwise>
+                <xsl:variable name="year" select="substring($dateStr, 1, 4)"/>
+                <xsl:choose>
+                    <xsl:when test="string-length($dateStr) &gt;= 4 and not(translate($year, '0123456789', ''))">
+                        <ccmm:date><xsl:value-of select="concat($year, '-01-01')"/></ccmm:date>
+                    </xsl:when>
+                    <xsl:otherwise>
+                        <ccmm:date/>
+                    </xsl:otherwise>
+                </xsl:choose>
+            </xsl:otherwise>
+        </xsl:choose>
+    </xsl:template>


The FormatDate template handles multiple date format variations (datetime with T, full date YYYY-MM-DD, year-month YYYY-MM, year-only YYYY, and malformed dates), but there are no tests covering these different formats. Add tests to verify that dates in different formats are correctly transformed, especially edge cases like year-only or year-month dates that need padding with '-01' or '-01-01'.

Copilot · 2026-02-27T14:41:27Z

+        <Format id="ccmm">
+            <Prefix>ccmm-xml</Prefix>
+            <XSLT>metadataFormats/ccmm.xsl</XSLT>
+            <Namespace>https://schema.ccmm.cz/research-data/1.1</Namespace>
+            <SchemaLocation>https://techlib.github.io/CCMM/dataset/schema.xsd</SchemaLocation>
+        </Format>


Inconsistent indentation for the Format element. The CCMM format definition uses 8-space indentation while the preceding ELG format uses 4-space indentation. For consistency with surrounding Format definitions (lines 217-235), use 4-space indentation for the Format element and 8-space indentation for its child elements.

Suggested change

<Format id="ccmm">

<Prefix>ccmm-xml</Prefix>

<XSLT>metadataFormats/ccmm.xsl</XSLT>

<Namespace>https://schema.ccmm.cz/research-data/1.1</Namespace>

<SchemaLocation>https://techlib.github.io/CCMM/dataset/schema.xsd</SchemaLocation>

</Format>

<Format id="ccmm">

<Prefix>ccmm-xml</Prefix>

<XSLT>metadataFormats/ccmm.xsl</XSLT>

<Namespace>https://schema.ccmm.cz/research-data/1.1</Namespace>

<SchemaLocation>https://techlib.github.io/CCMM/dataset/schema.xsd</SchemaLocation>

</Format>

Copilot · 2026-02-27T14:41:27Z

+    <xsl:template name="TimeReferences">
+        <!-- dc.date.issued -> Issued -->
+        <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='date']/doc:element[@name='issued']/doc:element/doc:field[@name='value']">
+            <ccmm:time_reference>
+                <ccmm:temporal_representation>
+                    <ccmm:time_instant>
+                        <xsl:call-template name="FormatDate">
+                            <xsl:with-param name="dateStr" select="."/>
+                        </xsl:call-template>
+                    </ccmm:time_instant>
+                </ccmm:temporal_representation>
+                <ccmm:date_type>
+                    <ccmm:iri>https://model.ccmm.cz/vocabulary/datacite/dateType/Issued</ccmm:iri>
+                    <ccmm:label xml:lang="en">Issued</ccmm:label>
+                </ccmm:date_type>
+            </ccmm:time_reference>
+        </xsl:for-each>
+        <!-- dc.date.accessioned -> Accepted -->
+        <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='date']/doc:element[@name='accessioned']/doc:element/doc:field[@name='value']">
+            <ccmm:time_reference>
+                <ccmm:temporal_representation>
+                    <ccmm:time_instant>
+                        <xsl:call-template name="FormatDate">
+                            <xsl:with-param name="dateStr" select="."/>
+                        </xsl:call-template>
+                    </ccmm:time_instant>
+                </ccmm:temporal_representation>
+                <ccmm:date_type>
+                    <ccmm:iri>https://model.ccmm.cz/vocabulary/datacite/dateType/Accepted</ccmm:iri>
+                    <ccmm:label xml:lang="en">Accepted</ccmm:label>
+                </ccmm:date_type>
+            </ccmm:time_reference>
+        </xsl:for-each>
+        <!-- dc.date.available -> Available -->
+        <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='date']/doc:element[@name='available']/doc:element/doc:field[@name='value']">
+            <ccmm:time_reference>
+                <ccmm:temporal_representation>
+                    <ccmm:time_instant>
+                        <xsl:call-template name="FormatDate">
+                            <xsl:with-param name="dateStr" select="."/>
+                        </xsl:call-template>
+                    </ccmm:time_instant>
+                </ccmm:temporal_representation>
+                <ccmm:date_type>
+                    <ccmm:iri>https://model.ccmm.cz/vocabulary/datacite/dateType/Available</ccmm:iri>
+                    <ccmm:label xml:lang="en">Available</ccmm:label>
+                </ccmm:date_type>
+            </ccmm:time_reference>
+        </xsl:for-each>
+        <!-- Fallback: if no issued/available dates, create a minimal time_reference from accessioned -->
+        <xsl:if test="not(doc:metadata/doc:element[@name='dc']/doc:element[@name='date']/doc:element[@name='issued']/doc:element/doc:field[@name='value']) and not(doc:metadata/doc:element[@name='dc']/doc:element[@name='date']/doc:element[@name='available']/doc:element/doc:field[@name='value']) and doc:metadata/doc:element[@name='dc']/doc:element[@name='date']/doc:element[@name='accessioned']/doc:element/doc:field[@name='value']">
+            <xsl:variable name="accessionedDate"
+                          select="doc:metadata/doc:element[@name='dc']/doc:element[@name='date']/doc:element[@name='accessioned']/doc:element/doc:field[@name='value'][1]"/>
+            <ccmm:time_reference>
+                <ccmm:temporal_representation>
+                    <ccmm:time_instant>
+                        <xsl:call-template name="FormatDate">
+                            <xsl:with-param name="dateStr" select="$accessionedDate"/>
+                        </xsl:call-template>
+                    </ccmm:time_instant>
+                </ccmm:temporal_representation>
+                <ccmm:date_type>
+                    <ccmm:iri>https://model.ccmm.cz/vocabulary/datacite/dateType/Issued</ccmm:iri>
+                    <ccmm:label xml:lang="en">Issued</ccmm:label>
+                </ccmm:date_type>
+            </ccmm:time_reference>
+        </xsl:if>
+    </xsl:template>


No test coverage for the time_reference element, which is a required CCMM field. The TimeReferences template handles multiple date types (Issued, Accepted, Available) and includes fallback logic, but none of these scenarios are tested. Add tests to verify that different date fields produce correct time_reference elements with appropriate date_type values and temporal_representation formatting.

@name

Verified against all 2628 live LINDAT records; every output validates against CCMM 1.1.0 in XSD 1.1 (0 invalid). - fix XPTY0004: doc:field[@name='value'] returns a sequence, and records with two language variants of a field passed it to concat()/contains(). 425 of 2628 records aborted; the resulting OAIException becomes HTTP 500 for the whole ListRecords page, so all 32 pages failed and nothing was harvestable - replace the model.ccmm.cz IRIs, which all return 404, with the registers CCMM 1.1.0 names in its own specification (en/dsv.ttl scope notes): COAR resource types and access rights, vocabs.ccmm.cz AgentRole / TimeReference / DescriptionType, EU Publications Office for languages - dc.publisher now maps to AgentRole/Publisher; Distributor does not exist at that level of the codelist - derive access_rights from others/access-status, which DSpace computes in DefaultAccessStatusHelper from the actual READ policies. The previous branches were dead (local.embargo.termslift occurs 0 times) or inverted (others/@restrictedaccess is set on 24 records, all of them open access). Correct for 746 of 2628 records before, 2628 of 2628 now - handle the CLARIN "0000" unknown-date convention and local.approximateDate.issued the way ClarinDateService does in the UI: the local field replaces dc.date.issued, a range becomes a time_interval and the original string is kept verbatim in ccmm:date_information - emit exactly one Issued time reference plus a Created one, as the CCMM profile requires; publication_year and the Issued year now always agree - stop emitting an invented licence IRI; fall back to others/cc and otherwise leave ccmm:license empty and move the wording to terms_of_use/description - the minimal test fixture used to produce schema-invalid output (no time_reference at all); it no longer can Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Copilot

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated 9 comments.

Suppressed comments (3)

dspace/config/crosswalks/oai/metadataFormats/ccmm.xsl:174

original_repository is populated from the item identifier, so the common hdl.handle.net/... form makes every record identify its item resolver as the repository. XOAI already supplies the actual repository URL at repository/url (ItemUtils.java:347); use that value instead.

                        <xsl:when test="doc:metadata/doc:element[@name='dc']/doc:element[@name='identifier']/doc:element[@name='uri']/doc:element/doc:field[@name='value']">
                            <xsl:variable name="uri" select="(doc:metadata/doc:element[@name='dc']/doc:element[@name='identifier']/doc:element[@name='uri']/doc:element/doc:field[@name='value'])[1]"/>

dspace-oai/src/test/java/org/dspace/xoai/tests/stylesheets/CcmmXslTest.java:125

This assertion exceeds the enforced 120-character line limit (checkstyle.xml:47-52). Wrap the expected organization name.

            equalTo("Charles University, Faculty of Mathematics and Physics, Institute of Formal and Applied Linguistics"))));

dspace-oai/src/test/java/org/dspace/xoai/tests/stylesheets/CcmmXslTest.java:182

This XPath exceeds the repository's enforced 120-character Java line limit (checkstyle.xml:47-52). Split it into concatenated segments.

            "//ccmm:dataset/ccmm:metadata_identification/ccmm:qualified_relation/ccmm:relation/ccmm:organization/ccmm:name",

+    <xsl:template name="PrimaryTemporalRepresentation">
+        <ccmm:temporal_representation>
+            <xsl:choose>
+                <xsl:when test="count($approxYears) &gt; 0 and $approxMin != $approxMax">


+        <!-- dc.title.alternative mapped to alternate_title -->
+        <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='title']/doc:element[@name='alternative']/doc:element/doc:field[@name='value']">
+            <ccmm:alternate_title>
+                <ccmm:title xml:lang="en">


+                </xsl:choose>
+            </ccmm:license>
+            <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='rights']/doc:element/doc:field[@name='value']">
+                <ccmm:description xml:lang="en"><xsl:value-of select="."/></ccmm:description>


+        <!-- dc.subject -->
+        <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='subject']/doc:element/doc:field[@name='value']">
+            <ccmm:subject>
+                <ccmm:title xml:lang="en"><xsl:value-of select="."/></ccmm:title>


+        <!-- dc.subject.* (nested qualifiers) -->
+        <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='subject']/doc:element/doc:element/doc:field[@name='value']">
+            <ccmm:subject>
+                <ccmm:title xml:lang="en"><xsl:value-of select="."/></ccmm:title>


+        <!-- dc.description (abstract) -->
+        <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='description']/doc:element[@name='abstract']/doc:element/doc:field[@name='value']">
+            <ccmm:description>
+                <ccmm:description_text xml:lang="en"><xsl:value-of select="."/></ccmm:description_text>


+        <!-- dc.description (general, non-qualified) -->
+        <xsl:for-each select="doc:metadata/doc:element[@name='dc']/doc:element[@name='description']/doc:element/doc:field[@name='value']">
+            <ccmm:description>
+                <ccmm:description_text xml:lang="en"><xsl:value-of select="."/></ccmm:description_text>


+        String result = apply("ccmm.xsl").to(resource("xoai-ccmm-test.xml"));
+        assertThat(result, is(ccmm().withXPath("//ccmm:title", equalTo("Czech NLP Dataset v2.0"))));


+    public void ccmmPublisherUsesPublisherRole() throws Exception {
+        String result = apply("ccmm.xsl").to(resource("xoai-ccmm-test.xml"));
+        assertThat(result, is(ccmm().withXPath(
+            "//ccmm:dataset/ccmm:qualified_relation[ccmm:role/ccmm:label='Publisher']/ccmm:relation/ccmm:organization/ccmm:name",


@url

- original_repository/iri named the item itself on every one of the 2628 records: it was derived with substring-before(uri, '/handle/'), which does not match LINDAT's hdl.handle.net URIs, so the otherwise branch emitted the whole item URI. XOAI already carries repository/@url; use it. - xml:lang was hardcoded "en" on subjects, alternate titles, descriptions and the rights wording. XOAI nests every value inside an element named after its language qualifier, so take the tag from there and fall back to "en" only for the unqualified "none" wrapper. Both covered by new tests. Re-checked over all 2628 live LINDAT records: still 2628/2628 valid against CCMM 1.1.0 in XSD 1.1. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Add CCMM 1.1.0 OAI-PMH crosswalk (ccmm-xml metadataPrefix)

6d9f78f

Copilot AI review requested due to automatic review settings February 27, 2026 12:48

Copilot started reviewing on behalf of milanmajchrak February 27, 2026 12:49 View session

Copilot AI reviewed Feb 27, 2026

View reviewed changes

milanmajchrak requested a review from Copilot February 27, 2026 14:33

Copilot started reviewing on behalf of milanmajchrak February 27, 2026 14:33 View session

Copilot AI reviewed Feb 27, 2026

View reviewed changes

milanmajchrak requested a balanced review from Copilot August 28, 2026 10:27

Copilot started reviewing on behalf of milanmajchrak August 28, 2026 10:28 View session

Copilot AI reviewed Aug 28, 2026

View reviewed changes

		<xsl:when test="contains($uri, '/handle/')">
		<xsl:value-of select="substring-before($uri, '/handle/')"/>

		String result = apply("ccmm.xsl").to(resource("xoai-ccmm-test.xml"));
		assertThat(result, is(ccmm().withXPath("//ccmm:title", equalTo("Czech NLP Dataset v2.0"))));

Uh oh!

Conversation

milanmajchrak commented Feb 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Problem description

Analysis

Problems

Manual Testing (if applicable)

Copilot review

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

milanmajchrak commented Feb 27, 2026 •

edited

Loading