Getting Started

Mapping a transparency graph requires a Mapper Registry and a Graph Builder. This package provides a default registry called CoreRegistry that will handle mapping the base UNTP Types (think Digital Product Passport, Digital Traceability Event, and so fourth) to the underlying n3 dataset.

import { CoreRegistry, GraphBuilder } from "@trustprovenance/graph-builder";

// Use the `CoreRegistry` as a standard set of mappers
const registry = new CoreRegistry();

// Instantiate the graph builder
const builder = new GraphBuilder({
  registry,
});

// Fetch the Digital Product Passport, as an example via a HTTP GET request.
const digitalProductPassport = await fetch("GET", {
  url: "https://acme.com/dpp.json",
});

// Add to the builder
builder.addNode(DPP);

// Repeat until Transparency Graph is traversed.

// Once complete, you can get the `n3` store from the builder.
const n3Store = builder.getGraph();

The Foundations

The Graph Builder has two fundamental components for operation:

• A Mapper Registry
• A Graph Builder

Mapper Registry

A Mapper Registry is just a simple Map datatype that will map a payload type to a mapper function. The type is equivalent to a value found in the type array of UNTP data types (i.e. DigitalProductPassport, VerifiableCredential, and so fourth).

The mapper function return will take this JSON payload, and a n3 DataStore and add to the dataset the relevant subject, predicate, and object's that should be extracted.

// The `CoreRegistry` is a "out-of-the-box" registry for handling the basic UNTP data types.
import { CoreRegistry } from "@trustprovenance/graph-builder";
import { DataStore } from "n3";

const registry = new CoreRegistry();
const store = new DataStore();

// Example JSON payload of a Digital Product Passport
const example = { type: ["DigitalProductPassport", "VerifiableCredential"] };

// We can see the JSON payload is a Digital Product Passort, but in practice the `GraphBuilder` handles this for us.
const mapper = CoreRegistry.get("DigitalProductPassport");

// Add the relevant RDF triples to the tore
mapper.map(example, store);

The above code is something you would ideally not have to write as the Graph Builder handles this for you.

Extending the Mapper Registry

Since the UNTP supports extensions, so does @trustprovenance/graph-builder. Your industry map have an extension beyond the core UNTP types (such as the Australian Agriculture Transparency Protocol). You can build your own Mapper and add it to a MapperRegistry which is used to instantiate a GraphBuilder.

Let's take a DigitalLivestockPassport from the AATP as an example. This extends a DigitalProductPassport with additional attributes we may want to capture.

{
    "type": ["DigitalLivestockPassport", "DigitalProductPassport", "VerifiableCredential"],
    ...TODO
}

We can then create a mapper that handles this.

// ./digital-livestock-passport.ts
import { Mapper } from "@trustprovenance/graph-builder";
import { DigitalLivestockPassport } from "some type definition you have";

export class DigitalLivestockPassportMapper
  implements Mapper<DigitalLivestockPassport>
{
  map(data: DigitalLivestockPassport, store: Store): void {
    // Add to `n3` store as required! Up to you!
  }
}

and create a custom registry,

import { CoreRegistry, GraphBuilder } from "@trustprovenance/graph-builder";
import { DigitalLivestockPassportMapper } from "./digital-livestock-passport";

// Let's extend the base UNTP registry
const registry = new CoreRegistry();

// And add our own.
registry.register("DigitalLivestockPassport", DigitalLivestockPassportMapper);

We can then use this Mapper Registry for Graph building.

import { GraphBuilder } from "@trustprovenance/graph-builder";

const builder = new GraphBuilder({ registry });