biblatex icon indicating copy to clipboard operation
biblatex copied to clipboard

eprinttype ark

Open StevanWhite opened this issue 4 years ago • 3 comments

Archival material is often accessed via an ark (Archival Resource Key), for instance, ark keys can be seen in the reference data on this page: https://archive.org/details/elementarytreati00pgta/page/n1/mode/2up

I suggest that you add support for this kind of document identifier.

These are resolved using an ARK resolver. There is more than one; I use the one at http://n2t.net/

Note: in references, I always see "ark" in lowercase letters.

These declarations suffice for me:

\DeclareFieldFormat{eprint:ark}{%
  ark\addcolon\space
  \ifhyperref
    {\href{https://n2t.net/ark:#1}{\nolinkurl{#1}}}
    {\nolinkurl{#1}}}
\DeclareFieldAlias{eprint:ARK}{eprint:ark}

This is used in a bibliography entry like so

  eprint =      {/13960/t0zp3zt39},
  eprinttype =  {ark},

to produce a URL like this https://n2t.net/ark:/13960/t9379hs7x

See:

ARK NAANs and systems https://arks.org/about/ark-naans-and-systems/

The ARK Identifier Scheme https://www.ietf.org/id/draft-kunze-ark-28.html

StevanWhite avatar Oct 01 '21 18:10 StevanWhite

See also https://github.com/plk/biblatex/issues/38 and https://github.com/plk/biblatex/issues/1155.

If unlike for URNs it is guaranteed that (the/every) resolver resolves all valid ARKs, we can look into adding this as a default eprint declaration.

What I'm a bit worried about is that you say N2T is just one resolver. Why should we force that particular resolver upon users as a default choice? Are there others? Would users want to select other resolvers? Is N2T likely to remain stable?

moewew avatar Oct 02 '21 07:10 moewew

I'm afraid that such questions are for you to decide: I don't understand the issues well enough.

The fact that there is more than one resolver, I noticed in the link I provided: https://arks.org/about/ark-naans-and-systems/. Maybe that page explains better how it works --- I only skimmed through it.

StevanWhite avatar Oct 02 '21 08:10 StevanWhite

@moewew - N2T seems to be the main "meta-resolver" and seems rather useful. For example, it resolves DOIs, I just tried:

https://n2t.net/DOI:/10.17605/OSF.IO/9P7CB

which works fine

PubMED IDs:

https://n2t.net/PMID:/36890230

ISSNs and ISBNs:

https://n2t.net/ISSN:/615-682536 https://n2t.net/ISBN:/0593420411

I wonder if we shouldn't add an option to use n2t for every resolvable thing we can, which will cover all of these proliferating eprint types?

plk avatar Mar 12 '23 16:03 plk