search

w3c / rdf-star

RDF-star specification
https://w3c.github.io/rdf-star/
Other
119 stars 23 forks source link

Turtle* syntax tests #53

Closed afs closed 3 years ago

afs commented 3 years ago

This issue is for Turtle* syntax tests.

A first set is provided PR #52.

lisp commented 3 years ago

have the two modes been ratified - with a distinction in interpretation anchored in the syntax?

afs commented 3 years ago

@lisp - they are syntax tests. As per PR, description:

They test whether a parser accepts or rejects a file. There is no implication as what triples are generated which is left for a separate set of tests.

lisp commented 3 years ago

is the "annotation" syntax being considered for both modes?

afs commented 3 years ago

I put annotation examples in for discussion.

I would have thought annotation applied to both SA and PG modes and I see nothing in the discussion to suggest it is anything other than a convenience syntax for :s :p :o. <<:s :p :o>> (it's output is all the SA triples) and neatly collecting the triple term assertions together (the second example shows that for the modelling case of multiple provenance). But no decision.

Indeed, IMO it is one way to resolve the SA/PG choice.

hartig commented 3 years ago

is the "annotation" syntax being considered for both modes?

No, the way I see it, adding the annotation gives us a way to cover both modes, where the annotation syntax is for PG mode.

By the way, there is a separate issue for the annotation syntax: #9

hartig commented 3 years ago

have the two modes been ratified - with a distinction in interpretation anchored in the syntax?

That would be my desired outcome. However, nothing has been ratified yet.

lisp commented 3 years ago

then, the question becomes, is the test suite projective or prescriptive?

pchampin commented 3 years ago

the, the question becomes, is are the test suite projective or prescriptive?

The way as see it, nothing is prescriptive at the moment.

pchampin commented 3 years ago

@afs should we close this issue? This is not to mean that the corresponding test suite is complete -- in my view, it will always be possible to add new tests. But that's "normal process" and should not require an outstanding issue.

afs commented 3 years ago

I agree - the primary version of content is the documents and new issues can be raised for documents. Closing an issue is not ending discussion.