CLARIN-DSpace v9/Port #1384 (CMDI ResourceRef machine-actionable download links) to the v9 base

[Kasinhou]

Post-snapshot sync dtq-dev → dtq-dev-9-base

Port of dataquest-dev#1384 (itself a port of ufal#1384 by @kosarko, upstream commit 82e33e817a) to the v9 base.

6114281c15 (dtq-dev #1384) — CMDI ResourceRef machine-actionable download links — PORT

ProcessBitstreams in lindat_cmdi.xsl now emits
${dspace.server.url}/api/core/bitstreams/handle/<handle>/<url-encoded-filename>
instead of the UI-only ${dspace.ui.url}/bitstream/handle/...?sequence=n link, which only resolved through the Angular legacy-URL redirect in a browser and was useless to a machine client. Filenames are URL-encoded (fnx:encode-for-uri), and only ORIGINAL-bundle bitstreams are listed, since the download endpoint serves nothing else.

Clean conflict-free cherry-pick — no v9 adaptations needed. Verified against the v9 base:

• lindat_cmdi.xsl has an identical pre-image here ($dsURL declared at line 28, used in exactly one place — ProcessBitstreams' ResourceRef), and the fnx prefix (http://www.w3.org/2005/xpath-functions) that encode-for-uri needs is already declared and listed in exclude-result-prefixes.
• CmdiDownloadLinkIT.java is a new file and needs no javax→jakarta conversion: its javax.xml.* imports are JDK JAXP, not Java EE.
• Every API the new IT touches exists on this branch: SolrOAIReindexer#reindexItem(Item) + its solrServerResolver field, DSpaceItemRepositoryResolver's itemRepository field, MockSolrServer, XOAICacheService / SolrServerResolver / ItemRepositoryResolver, BitstreamBuilder#createBitstream(ctx, item, is, bundleName), Constants.LICENSE_BUNDLE_NAME / LICENSE_BITSTREAM_NAME, and BitstreamByHandleRestController serving GET /api/core/bitstreams/handle/{prefix}/{suffix}/{filename} for ORIGINAL-bundle bitstreams only.
• spring-boot 3.5.14 still ships @MockBean (deprecated since 3.4); the sibling OAIpmhIT on this branch uses it the same way, so no test-harness change.

Gates

• Ported files are byte-identical to 6114281c15 (git diff 6114281c15 HEAD -- <both paths> is empty).
• Full reactor mvn install -DskipTests — BUILD SUCCESS, 0 Checkstyle violations.
• CmdiDownloadLinkIT run against a testEnvironment.zip rebuilt from this tree (so the new XSL is the one under test): Tests run: 1, Failures: 0, Errors: 0, Skipped: 0. The test harvests the CMDI record via OAI-PMH GetRecord, asserts the ResourceRef equals the by-handle URL with a %20-encoded filename, asserts the legacy /bitstream/handle/ + ?sequence= forms and the LICENSE bitstream are absent, then downloads the file from that URL and compares content.
• No other test asserts on the CMDI crosswalk output — ClarinRefBoxControllerIT only checks the cmdi export-format URL/metadata, not ResourceRef. CI Unit+Integration jobs are the record gate.

Note

The prior v9-base port commits reference CLARIN_V9_POST_SNAPSHOT_SYNC_ACCEPTANCE.md / CLARIN_V9_PARITY_BACKLOG.md, neither of which is tracked on this branch, so this port carries no "Fulfils §5 / card" line. If those planning docs should get a card for 6114281c15, say so and I'll add it.

🤖 Generated with Claude Code

[Copilot]

Pull request overview

This PR ports CLARIN-DSpace change #1384 to the v9 base to ensure CMDI OAI-PMH cmd:ResourceRef values are machine-actionable download links pointing to the REST by-handle bitstream endpoint, instead of UI-only legacy bitstream URLs.

Changes:

• Updated the CMDI XSL crosswalk to emit ${dspace.server.url}/api/core/bitstreams/handle/<handle>/<url-encoded-filename> and to list only ORIGINAL-bundle bitstreams.
• Added an end-to-end integration test which harvests a CMDI record via OAI-PMH, validates the emitted link(s), and downloads the referenced bitstream to verify content.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
dspace/config/crosswalks/oai/metadataFormats/lindat_cmdi.xsl	Switches CMDI ResourceRef output to REST by-handle download URLs with URL-encoded filenames, restricted to ORIGINAL bundle bitstreams.
dspace-server-webapp/src/test/java/org/dspace/app/oai/CmdiDownloadLinkIT.java	New integration test covering CMDI ResourceRef link format, exclusion of legacy links/LICENSE bitstream, and successful download/content verification.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.