bruju / prec Goto Github PK

This snippet is never tested in a standard npm run test run, and probably buggy.

The term entity is confusing and not really needed here.

Replace prec:edgeIs and prec:entityIs with prec:selfIs

• Having two different predicates for basically the same thing is confusing
• selfIs is more intuitive: "I'm removing pvar:self, what plays the role of pvar:self in my template graph ?"

Replace pvar:entity with pvar:holder

• When using pvar:entity, the question of "which entity am I talking about?" rises
• A property is on something, held by something

Because of the changes in the JSON alternate model, the tinkerpop tests doesn't pass anymore (mocha test\with_property_graph\tinkerpop.js)

Note that the node gremlin-to-rdf.js still produces the expected output.

^ Title

Why: WIP proof of PRSC reversability mentions pvar:self instead of distinct values for pvar:node and pvar:edge.

Corner cases that are not yet tested and are probably bugged:

Property Graph

[ ] -(s: "A Subject")-> [  ]

Context

[] a prec:PropertyModel ;
  prec:relationshipIRI rdf:subject ;
  prec:modelAs prec:DirectTriples .

prec:Relationships prec:modelAs prec:RDFStarUnique .

Expected result

:node1 :edge :node2 .
<< :node1 :edge :node2 >> rdf:subject "A Subject" .

Probable bugged result

:node1 :edge :node2 .
"A Subject" :edge :node2