Conversation
One qualification dossier exported as a single experiment Dataset with 17 hashed attachments, inline PropertyValues (evidence SHA-256, dossier version, eLabFTW extra fields), keywords as a comma-separated string, and every File listed in the root hasPart alongside the entry. Produced from the public demo dataset; imports into eLabFTW as one experiment.
|
Hey @TM289012, thanks for the example! I really like that you worked towards getting it 1.2+202609 compliant. I just created a PR with the now up-to-date test-suite. Can you give that a try and check if your example succeeds? |
…ned graph Rebuilt with Reformulation Assurance v0.12.4. The metadata descriptor declares RO-Crate 1.2, the root Dataset declares the RO-Crate 1.2 and ELN 1.2+202609 profiles, variableMeasured holds references to top-level PropertyValue nodes, the root has publisher and license, and the author has a local id with an affiliation. Passes tests/run_checks.py from TheELNConsortium#159 at version 1.2+202609 (archive structure, metadata rules, schema, rocrate parse); the validator check was not run locally. README regenerated with tools/eln2md.py.
|
Thanks Steffen. Your suite read the example as 1.1 and skipped the consortium checks, so I moved the export to RO-Crate 1.2 with the 1.2+202609 declaration and rebuilt it; the new commit is on this branch. What changed: the 1.2 context and descriptor, both profiles declared on the root as Profile entities, variableMeasured now pointing at top-level PropertyValue nodes so the graph is fully flattened, a publisher and license on the root, and a local id plus affiliation for the author. With --no-recommend everything passes at 1.2+202609, including the validator; the full run only notes the optional preview and the missing mentions. One side effect other exporters may care about: once variableMeasured is flattened, eLabFTW's importer only picks up the elabftw_metadata PropertyValue if the crate carries eLabFTW's own version marker on the root. I have reported it on the eLabFTW side with a suggested fix: elabftw/elabftw#7501 |
|
I guess it makes sense to pull the new master-branch and remember to update the spec version to …/eln-spec/1.2+202609. |
This branch has not been deployed
2 participants
Adds an example .eln from Reformulation Assurance, an open-source workbench for ingredient-replacement projects (https://github.com/TM289012/reformulation-assurance), and a row in the Known implementations table (export only).
The archive is one qualification dossier as a single experiment Dataset (genre: experiment) with 17 hashed attachments, inline PropertyValues for the evidence SHA-256, the dossier version and eLabFTW extra fields, keywords as a comma-separated string, and every File listed in the root hasPart next to the entry, per the 1.2+202609 wording. The crate stays on RO-Crate 1.1 for now because the validator check in tests/ targets the 1.1 profile; happy to switch when that moves.
Produced from the public demo dataset (no real formulas). It imports into the eLabFTW demo instance as one experiment with all 17 attachments and the extra-fields panel populated. Archive structure, ro-crate-py, metadata-rules and schema checks from tests/ pass locally; I could not run the roc-validator check offline, so please shout if CI disagrees.
README_template.md and README.md were generated with tools/eln2md.py.