The shape topology to Mermaid

[pietercolpaert]

It would be useful to be able to export the shape topology as a mermaid description.

I.e., the shape topology

Shape {
   atLeastOneLists: [ Shape {
       requiredPaths: ["gsp:geo asWKT|asGML", "^foaf:member"],
       optionalPaths: ["foaf:knows"]
   },
   Shape {
    //...
   }]
} 

Could be translated to this in mermaid:

flowchart TD
    S(( ))-->X{ }
    X-->S1(( ))
    X-->S2(( ))
    S1-->|gsp:geo|S3[ ]-->|"asWKT | asGML"|S4[ ]
    S5[ ]-->|^foaf:member|S1
    S1-.->|foaf:knows|S6[ ]

Visualizing into:

flowchart TD
    S(( ))-->X{ }
    X-->S1(( ))
    X-->S2(( ))
    S1-->|gsp:geo|S3[ ]-->|"asWKT | asGML"|S4[ ]
    S5[ ]-->|^foaf:member|S1
    S1-.->|foaf:knows|S6[ ]

Loading

[pheyvaer]

@pietercolpaert Can you explain how it fits in the extractor? What is the function description and input parameters?

[pietercolpaert]

import {CBDShapeExtractor} from "extract-cbd-shape";

This will create the shape topology already. the shapesGraphStore is an rdf-stores RDFStore containing all quads of the shape

let extractor = new CBDShapeExtractor(shapesGraphStore);

This now has something of type ShapesGraph in it, in which I’d like to add a toMermaid function that takes the term of the nodeshape as a parameter

extractor.shapesGraph.toMermaid(df.namedNode('nodeshapeIRI'));

[pheyvaer]

@pietercolpaert How would you convert something like this shape to Mermaid?

ex:Shape a sh:NodeShape ;
    sh:property [
        sh:path ex:subject ;
        sh:minCount 1 ;
        sh:node [
            sh:property [
                sh:path rdfs:label ;
                sh:minCount 1
            ]
        ]
    ] .

When it's converted to a Shape topology it has both requiredPaths and nodeLinks, and both mention ex:subject:

{
  "requiredPaths": [
    {
      "predicate": {
        "termType": "NamedNode",
        "value": "http://example.org/subject"
      }
    }
  ],
  "nodeLinks": [
    {
      "pathPattern": {
        "predicate": {
          "termType": "NamedNode",
          "value": "http://example.org/subject"
        }
      },
      "link": {
        "termType": "BlankNode",
        "value": "n3-39"
      }
    }
  ],
  "atLeastOneLists": [],
  "optionalPaths": [],
  "closed": false
}

[pietercolpaert]

A nodelink path is also going to be repeated in either the optional or in the required path indeed. So I’d say if a path is already processed as part of a nodelink, let’s ignore it in the optional/required, unless we will also visualize required vs. optional in a different way, i.e. by using a dashed line instead of a full line?

I.e., for optional node links, this then would become:

flowchart TD
    S(( ))---X{OR}
    X---S1(( ))
    X---S2(( ))
    S1-->|gsp:geo|S3[ ]-->|"asWKT | asGML"|S4[ ]
    S5(( ))-.->|^foaf:member|S1
    S1-->|foaf:knows|S6[ ]
    S5-->|rdfs:label|S7[ ]

Imagine the foaf:member’s subject is an organization, that if given, must have an rdfs:label, we can show it as follows:

flowchart TD
    S(( ))---X{OR}
    X---S1(( ))
    X---S2(( ))
    S1-->|gsp:geo|S3[ ]-->|"asWKT | asGML"|S4[ ]
    S5(( ))-.->|^foaf:member|S1
    S1-->|foaf:knows|S6[ ]
    S5-->|rdfs:label|S7[ ]

Loading

[pheyvaer]

Ok! I included that.

Another question: how do we deal with double inverse paths?

ex:DoubleInversePathShape a sh:NodeShape ;
    sh:closed true;
    sh:property [
        sh:path ([sh:inversePath [sh:inversePath ex:p1 ] ]) ;
        sh:minCount 1
    ] .

Do we just invert the first inverse path and leave the second as is (which is what happens now) or do we keep inverting them until we can't anymore?

P.S. I don't fully understand why someone would use a double inverse path in the first place 😄

[pietercolpaert]

I believe we should turn it around again indeed: double inverse becomes forward

[pheyvaer]

Closed by #30