Open vincerubinetti opened 1 year ago
Can you provide the identifiers as well for a wosid
, eid
, and source-work-id
(particularly the records without a DOI)? I'd like to learn more about those works and what's the best way to cite them.
More generally, we rely on Bioregistry for assigning prefixes. In an ideal world, ORCID would align their prefixes with more standard Bioregistry prefixes whenever possible. But in the meantime, perhaps we can define a mapping from ORCID-style prefix to Bioregistry-style prefix which would enable Manubot citation support.
I think wosid
is Web of Science ID. Bioregistry would likely accept registration of the wosid
prefix.
I don't think the ORCID eid
prefix is a synonym for civic.eid
which refers to Clinical Interpretation of Variants in Cancer.
Here are a few IDs as examples:
eid:2-s2.0-85012144154
wosuid:WOS:000386326200051
source-work-id:0417180825983-87
And here is the full (JSON, not XML) API response for the ~10% of Casey's papers that don't have a doi
(in the top-level external-ids
), just in case the extra info gives some clues:
It seems like possibly a big lift to make a mapping between all of these ORCID id types, and there might be no mapping for some of them. Maybe we could somehow determine what the most common non-doi types are and support those. And if Casey's ORCID is any indication, eid and wosuid seem to be it... 🤷♀️
Okay so eid
refers to Scopus. For example, eid:2-s2.0-85012144154
refers to the study "Computational approaches to study microbes and microbiomes" with a URL of http://www.scopus.com/inward/record.url?eid=2-s2.0-85012144154&partnerID=MN8TOARS
. Based on this blog post, we can also reformat the URL to be https://www.scopus.com/record/display.uri?eid=2-s2.0-85012144154&origin=resultslist
. Both of these resolve in my browser, but I can't get them to work programtically with curl
or manubot cite
. See also Wikidata P1154. Too bad the ORCID data doesn't include other more accessible identifiers like pmc:PMC4832978
, pubmed:26776218
, or doi:10.1142/9789814749411_0051
(despite the scopus record contained some of these identifiers).
Possible solutions:
Looks like you can get a URL for the Web of Science records like https://www.webofscience.com/wos/woscc/full-record/WOS:000411479100034
(see Wikidata P8372. This does work programatically, but WoS doesn't set complete citation metadata in the HTML and/or Zotero translator needs tweaks.
Possible solutions:
Web of Science Nextgen.js
and the other WoS translators.if Casey's ORCID is any indication, eid and wosuid seem to be it
Casey has probably done Scopus to ORCID and this.
ORCID has APIs that can return paper identifiers of the following types: https://pub.orcid.org/v3.0/identifiers
For example, about 10% Casey Greene's papers returned from ORCID don't have a
doi
, but instead have aneid
orwosuid
or others. It'd be nice if Manubot could support citing such papers.ORCID unfortunately seems to provide very little info about what these identifiers are or where they come from. For example, is
eid
the same ascivic.eid
here...Not sure if this is doable with upstream packages/services.