transmute-industries / jsonld-to-cypher Goto Github PK

View Code? Open in Web Editor NEW

4.0 6.0 2.0 4.14 MB

Compile JSON-LD (Decentralized Identifiers and Verifiable Credentials) to Cypher.

Home Page: https://github.com/transmute-industries/jsonld-to-cypher

JavaScript 99.16% Shell 0.84%

cypher decentralized-identifiers jsonld neo4j verifiable-credentials

jsonld-to-cypher's Introduction

JSON-LD to Cypher

Compile JSON-LD (Decentralized Identifiers and Verifiable Credentials) to Cypher.

🚧 Warning Experimental 🔥

Usage

Command Line

npm i -g @transmute/jsonld-to-cypher@latest

jsonld-to-cypher convert ./examples/minimal-vc/minimal-vc.json
jsonld-to-cypher convert ./examples/v1-vc-jwt/v1-vc-jwt.jwt -- --type jwt

Library

npm i @transmute/jsonld-to-cypher@latest --save

Transforming Verifiable Credentials

See minimal-vc.

const { cypher, graph } = await Cypher.fromDocument({
  '@context': ['https://www.w3.org/2018/credentials/v1'],
  type: ['VerifiableCredential'],
  issuer: 'https://example.edu/issuers/565049',
  issuanceDate: '2010-01-01T19:23:24Z',
  credentialSubject: {
    id: 'did:example:ebfeb1f712ebc6f1c276e12ec21',
  },
})

Action

name: Demo
on: [workflow_dispatch]
jobs:
  merge:
    runs-on: ubuntu-latest
    name: Merge
    steps:
      - name: Merge Query from Document
        uses: transmute-industries/[email protected]
        id: merge
        with:
          neo4j-uri: ${{ secrets.NEO4J_URI }}
          neo4j-user: ${{ secrets.NEO4J_USERNAME }}
          neo4j-password: ${{ secrets.NEO4J_PASSWORD }}
          operation: merge
          document: |
            {
              "@context": ["https://www.w3.org/2018/credentials/v1"],
              "type": ["VerifiableCredential"],
              "issuer": "https://example.edu/issuers/565049",
              "issuanceDate": "2010-01-01T19:23:24Z",
              "credentialSubject": {
                "id": "did:example:ebfeb1f712ebc6f1c276e12ec21"
              }
            }

Development

npm i
npm t

Testing CLI

npm run jsonld-to-cypher convert ./examples/minimal-vc/minimal-vc.json

jq -r '.jws' ./examples/v1-vc-jwt/v1-vc-jwt.json > ./examples/v1-vc-jwt/v1-vc-jwt.jwt

npm run jsonld-to-cypher convert ./examples/v1-vc-jwt/v1-vc-jwt.jwt -- --type jwt

Testing GitHub Action with Act

act -j neo4j-merge-documents-local --secret-file .env

Installing Neo4j Desktop

See https://neo4j.com.

jsonld-to-cypher's People

Contributors

Stargazers

Watchers

Forkers

nissimsan bnationsdev

jsonld-to-cypher's Issues

Failing node merge

I'm attempting to import the same resource twice, as part of two separately generated/imported cypher scripts.

I'm getting Node(45) already exists with label Resourceand propertyuri = 'did:key:z6Mkf8r9cUvN48fPhmi13FMgFDg6ccVqmt1spny4f2QoymDP' which seems like unintended behavior. If the URI is the same, then the nodes should merge. But it seems like MERGE doesn't do the trick for some reason.

Predicate relationships improvement

MERGE ( n0 : DecentralizedIdentifier { id: "did:key:z6MktiSzqF9kqwdU8VkdBKx56EYzXfpgnNPUAGznpicNiWfn" ,  description: "Melting and Pouring Steel", name: "Slovenian Steel" } )
MERGE ( n3 : UniformResourceLocator { id: "https://schema.org/description"  } )
MERGE (n3)-[e1: Predicate { id : "urn:uuid:a8e3cd1f-8e13-4630-b0f9-6dbaf9112471", predicate: "https://vocabulary.transmute.industries/ns/dynamic/#disclosedPredicate" } ]->(n0)

I think the idea of this e1 "Predicate" edge is to include the term definition of n0's description property. However, there isn't a direct relationship to the property (only its node), so in this can you couldn't tell whether "https://schema.org/description" defines description or name.

It also creates a quite cluttered picture with all those nodes and predicate definitions.

A simpler model would to include the term definitions alongside , instead of the above three statements, just:

MERGE ( n0 : DecentralizedIdentifier { id: "did:key:z6MktiSzqF9kqwdU8VkdBKx56EYzXfpgnNPUAGznpicNiWfn" ,  description: "Melting and Pouring Steel", descriptionTerm: "https://schema.org/description", name: "Slovenian Steel", nameTerm: "https://schema.org/name" } )

Introduce `term` keyword to simplify data model

Continuing from #4...

I propose that the following JSON-LD:

{
  "@context": [
    "https://www.w3.org/2018/credentials/v1",
    "https://w3id.org/traceability/v1"
  ],
  "id": "urn:uuid:a8e3cd1f-8e13-4630-b0f9-6dbaf9112471",
  "type": [
    "VerifiableCredential",
    "CommercialInvoiceCredential"
  ],
  "name": "Commercial Invoice Certificate",
  "issuanceDate": "2022-10-08T12:30:00Z",
  "issuer": {
    "type": [
      "Organization"
    ],
    "id": "did:key:z6MktiSzqF9kqwdU8VkdBKx56EYzXfpgnNPUAGznpicNiWfn",
    "name": "Slovenian Steel",
    "description": "Melting and Pouring Steel"
  }
}

Could be both simpler and more precisely represented like this:

MERGE ( n0 : Organization { id: "did:key:z6MktiSzqF9kqwdU8VkdBKx56EYzXfpgnNPUAGznpicNiWfn", description: "Melting and Pouring Steel", descriptionTerm: "https://schema.org/description", name: "Slovenian Steel", nameTerm: "https://schema.org/name" } )
MERGE ( n2 : OrganizationTerm { id: "https://schema.org/Organization"  } )
MERGE ( n5 : CommercialInvoiceCredentialTerm { id: "https://w3id.org/traceability#CommercialInvoiceCredential"  } )
MERGE ( n6 : VerifiableCredentialTerm { id: "https://www.w3.org/2018/credentials#VerifiableCredential"  } )
MERGE ( n9 : VerifiableCredential : CommercialInvoiceCredential { id: "urn:uuid:a8e3cd1f-8e13-4630-b0f9-6dbaf9112471" ,  name: "Commercial Invoice Certificate", nameTerm: "https://schema.org/name", issuanceDate: datetime("2022-10-08T12:30:00Z"), issuanceDateTerm: "https://www.w3.org/2018/credentials#issuanceDate" } )
MERGE (n0)-[e0: Type { id : "urn:uuid:a8e3cd1f-8e13-4630-b0f9-6dbaf9112471", predicate: "http://www.w3.org/1999/02/22-rdf-syntax-ns#type" } ]->(n2)
MERGE (n9)-[e5: Type { id : "urn:uuid:a8e3cd1f-8e13-4630-b0f9-6dbaf9112471", predicate: "http://www.w3.org/1999/02/22-rdf-syntax-ns#type" } ]->(n5)
MERGE (n9)-[e6: Type { id : "urn:uuid:a8e3cd1f-8e13-4630-b0f9-6dbaf9112471", predicate: "http://www.w3.org/1999/02/22-rdf-syntax-ns#type" } ]->(n6)
MERGE (n9)-[e7: Issuer { id : "urn:uuid:a8e3cd1f-8e13-4630-b0f9-6dbaf9112471", predicate: "https://www.w3.org/2018/credentials#issuer" } ]->(n0)
RETURN n0,n2,n5,n6,n9

Specifically, I am introducing a term keyword which is used:

For property definitions, for example for name and nameTerm
For class definitions, which the type edges can point at, for example OrganizationTerm.

Github action example on readme

The README's "Usage" should also include a section on how to use the Github action.