Skip to main content

Schema

The Hypergraph schema allows you to define the data model for your application. It is based on the GRC-20 specification and allows you to define Types with properties and relations to other Types.

Mapping​

The public knowledge graph is based on property IDs. In order to integrate with the public knowledge graph you need to map your own schema to IDs from the public graph's schema. Until you interact with the public knowledge graph or before publishing your schema to the public knowledge graph you can use any valid UUID as a property ID.

Hypergraph Schema Browser​

Building your app using a schema that is already actively used in Hypergraph's knowledge graph unlocks composability between your dataset and other datasets used in Hypergraph's knowledge graph.

You can search for schemas used within the Hypergraph knowledge graph using the Hypergraph Schema Browser. You can also use the TypeSync tool to generate a mapping for your schema.

Warning: TypeSync is not supported in the latest version of the hypergraph package. We are working on a new version of TypeSync.

Example​

Here is an example of a schema for an Event entity with the properties name and description.

import { Entity, Type, Id } from '@graphprotocol/hypergraph';

export const Event = Entity.Schema(
  {
    name: Type.String,
    description: Type.String,
  },
  {
    types: [Id('event-type-id')],
    properties: {
      name: Id('name-property-id'),
      description: Id('description-property-id'),
    },
  },
);

Relations​

In order to define relations between Types, you can use the Type.Relation type.

import { Entity, Type, Id } from '@graphprotocol/hypergraph';

export const Company = Entity.Schema(
  { name: Type.String },
  {
    types: [Id('company-type-id')],
    properties: {
      name: Id('name-property-id'),
    },
  },
);

export const Event = Entity.Schema(
  {
    name: Type.String,
    description: Type.String,
  },
  {
    types: [Id('event-type-id')],
    properties: {
      name: Id('name-property-id'),
      description: Id('description-property-id'),
      sponsors: Id('sponsors-property-id'),
    },
  },
);

Available Types​

  • Type.String (string)
  • Type.Number (number)
  • Type.Date (date)
  • Type.Boolean (boolean)
  • Type.Point (serialized to a string with a comma separated list of numbers)
  • Type.Relation (relation to another Type)

Example:

import { Entity, Type, Id } from '@graphprotocol/hypergraph';

export const Company = Entity.Schema(
  {
    name: Type.String,
    employees: Type.Number,
    founded: Type.Date,
    active: Type.Boolean,
    location: Type.Point,
  },
  {
    types: [Id('company-type-id')],
    properties: {
      name: Id('name-property-id'),
      employees: Id('employees-property-id'),
      founded: Id('founded-property-id'),
      active: Id('active-property-id'),
      location: Id('location-property-id'),
    },
  },
);

Optional Fields​

You can make a field optional by wrapping it in Type.optional.

import { Entity, Type, Id } from '@graphprotocol/hypergraph';

export const Company = Entity.Schema(
  {
  name: Type.String,
  description: Type.optional(Type.String),
  founded: Type.optional(Type.Date),
  },
  {
    types: [Id('company-type-id')],
    properties: {
      name: Id('name-property-id'),
      description: Id('description-property-id'),
      founded: Id('founded-property-id'),
    },
  },
);

Creating and publishing new Properties and Types​

We created a script to create and publish new properties and types to the public knowledge graph. You can find it at https://github.com/geobrowser/create-types-and-properties.