Skip to content

Verify re-publishing an unpublished dataset end to end on nm099999 #2202

Open

Description

@neuromechanist

Summary

nemar admin repo unpublish (#2198, PR #2199, ADR 0125) leaves every DOI, the tag protection ruleset and the S3 object lock in place, and never calls EZID.
Nothing yet proves that a dataset unpublished this way can be published again end to end.
Verify it on nm099999 (the standing end-to-end fixture; never nm000103-nm000107).

To verify

  • Publish nm099999, run nemar admin repo unpublish nm099999 --execute --reason ..., then file and approve a new publication request.
  • create_tag: the version tag already exists from the first publication; confirm the step treats an existing tag as done (or recreates it safely) and does not fail on the protected tag ruleset.
  • version_doi: the second run re-mints a version DOI against an existing dataset_versions row that already holds one; confirm it reuses or reconciles it and does not create a duplicate or error on the unique constraint.
  • S3 re-lock: confirm the object lock and the public carve-out are applied again, and the policy ends public for the dataset.
  • Repo spec: confirm the branch ruleset is re-applied on re-publication and the owner can push in between.
  • publish status and the admin list show the new request as published and the earlier one as published (unpublished <date>).

Fix whatever the walk-through finds on its own branch, with a test for each.

Refs #2198, #2199.

This issue was drafted with AI assistance.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions