Skip to content

[DOI] proxy URN-colon carrier + future allocation triggers (deferred v1 scope) #174

Description

@azaharizaman

Follow-up to #173 (DOI capability, MILESTONE row 11). Three items were deliberately deferred during v1; only the first is actionable now.

1. Proxy URN-colon carrier (actionable)

The DOI proxy understands colon-for-initial-slash substitution (Handbook §6.3.1 NOTE): https://doi.org/urn:doi:10.123:456 resolves 10.123/456, with multi-slash suffixes hex-encoded (10.123/456ABC/zyz → …:456ABC%2Fzyz). v1 recognizes bare urn:doi: but never maps the colon, so these read MISSING (pinned by test_urn_colon_form_missing in tests/capabilities/doi/test_grammar.py).

Acceptance: new carrier group (colon→slash mapping + %2F rule) resolving https://doi.org/urn:doi:10.123:456 → SUCCESS 10.123/456 (prefix must still satisfy the 4–9 registrant bound, so 10.123:456 stays MISSING); distinct grammar semantics id per the affinity rule (research §4.5); natural fit for an extra_grammars community extension rather than the shipped grammar.

Refs: research §13 row 11 (docs/development/research/2026-09-18-doi-canonicalization.md), plan Task 3 (docs/development/plans/2026-09-18-doi-capability.md), Handbook local conversion docs/development/research/DOIHandbook_2025.pdf.md:2298-2307.

2. Future triggers (not actionable — revisit only if the Foundation acts)

  • Non-10. directory indicators (Handbook §4.3.2 reserves them): v1's 10. gate reads them MISSING by design.
  • ≥10-digit registrants (Handbook: no defined length limit): v1's {4,9} bound encodes current allocation practice.
  • Either allocation change requires a recognition_revision bump and a bound/gate revisit (research §13 row 4). No action until observed in the wild.

Activity

  1. added
    v1.0.0v1.0.0 release
    followupFollow-up from merged PR — deferred or skipped review concern
    on Sep 30, 2026
  2. azaharizaman commented on Sep 30, 2026

    @azaharizaman
    CollaboratorAuthor

    Triage 2026-09-30: keeping open, labeled v1.0.0 + followup (follow-up to PR #173). Plain English: one rare DOI web-address shape (colon where a slash should be) is currently declined; this teaches an opt-in add-on to read it, with no change to default behavior. The other two items in this issue are explicitly future triggers, not work. No consolidation needed.

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

    enhancementNew feature or requestfollowupFollow-up from merged PR — deferred or skipped review concernv1.0.0v1.0.0 release

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions