changes to make to the entity.identifier.present, entity.identifier.resolvable, and resource.distribution.present checks

NCEAS / metadig-checks

MetaDIG suites and checks for data and metadata improvement and guidance.

Apache License 2.0

8 stars 9 forks source link

changes to make to the entity.identifier.present, entity.identifier.resolvable, and resource.distribution.present checks #374

Closed jeanetteclark closed 4 years ago

jeanetteclark commented 4 years ago

Description

Improvements for these checks and reducing redundancy

Procedure

entity.identifier.present: should check for presence either in [entity]/@id or [entity]/physical/distribution/url
entity.identifier.resolves: should check for resolution in[entity]/physical/distribution/url
resource.distribution.present: should only check in dataset/distribution/url and not in the entities (removes redundancy, and resource is metadata not data)

gothub commented 4 years ago

@jeanetteclark thanks for these check corrections, they will be in the next FAIR suite release. However, entity.identifier.resolves will be removed, as described here

jeanetteclark commented 4 years ago

I think on the Arctica call we changed our minds. @mbjones can you confirm?

mbjones commented 4 years ago

I can confirm that I think Jeanette's proposal makes sense to me, and should replace our prior discussion.

jeanetteclark commented 4 years ago

as an addendum:

resource.landingPage.present: should only check in dataset/distribution/url and not in the entities (resource is metadata not data)

gothub commented 4 years ago

@jeanetteclark how does the Arctica call affect this issue (from the comments above)?

Also, for entity.identifier.present, are you saying that for the path [entity]/physical/distribution/url, the last portion of the path has to be an identifier? How could someone using the metadata know which portion is the identifier?

jeanetteclark commented 4 years ago

My original description of the issue matches what we decided on the Arctica call

I believe that the identifier is meant to be represented by the entire URI, @mbjones can you confirm?

mbjones commented 4 years ago

That matches my take on it too.

gothub commented 4 years ago

Thx for the clarification.

The following updates have been make:

entity.identifier.present updated in commit 49e445c5e0558144498a7a58ea6f482b0bff7350
- include xpath for entity./physical/distribution/url
entity.identifier.resolvable removed from FAIR suite in commit 3a8d4a647fc9afb244f3e26d96c0e9ff328c8295
resource.distribution.present updated in commit b3844645399978e9a23046295697cea0d1a6208a
- removed entity ./physical/distribution/url xpath
resource.landingPage.present updated in commit 7f07670f91ac2322eaf36234d5a124cb90cba066
- removed entity ./physical/distribution/url xpath

jeanetteclark commented 4 years ago

@gothub I think we got some wires crossed on entity.identifier.resolvable

From above:

entity.identifier.resolveable: should check for resolution in[entity]/physical/distribution/url

gothub commented 4 years ago

@jeanetteclark this check was removed as a result of a conversation in slack. It was my understanding that this check should be removed entirely:

https://github.com/NCEAS/metadig-checks/issues/373

gothub commented 4 years ago

OK - instead of removing entity.identifier.resolvable, rename it to entity.distributionUrl.resolvable and check the entity distribution url for resolvability. For EML, this is /eml/dataset/[entity]/physical/distribution/online/url.

What is the current thinking on requiring the <url function="download"> attribute? Should this be included or not?

gothub commented 4 years ago

See https://github.com/NCEAS/metadig-checks/issues/379 for a resolution on entity.identifier.resolvable