GraphQL Community Graph

A few examples of using the GraphQL Community Graph from Neo4j.

For managing additional Neo4j instances with Neo4j-GraphQL installed, you can provide an additional binding configuration to the bindings parameter. If the Neo4j-GraphQL extension is not available, then you can use the neo4j-graphql-js package to directly translate your GraphQL requests to Cypher requests.

With Multiple Bindings

In this example, we build a binding for your local server using typeDefs and for the GraphQL Community Graph remote endpoint using twitterTypeDefs, from twitter schema supported by the API generated by the Neo4j-Graphql extension on the remote endpoint.

Local Schema

const typeDefs = `
type Technology @model {
name: String! @unique
integration: [Technology] @relation(
name: "HAPPINESS",
direction: OUT
)
}
`;

The following is a modified version of the twitter schema from the GraphQL Community Graph. Each type has received a @model directive. There are no further changes made because the endpoint is read-only and thus does not allow the schema or data to be modified.

Remote Schema

const twitterTypeDefs = `
type Link @model {
url: ID!
}
type TwitterUser @model {
id: ID!
screen_name: String!
name: String
location: String
followers: Int
following: Int
statuses: Int
profile_image_url: String
posted: [Tweet] @relation(name:"POSTED", direction:"OUT")
}
type Tweet @model {
id: ID!
text: String
created: Int
favorites: Int
postedBy: TwitterUser @relation(name:"POSTED", direction:"IN")
mentioned: [TwitterUser] @relation(name:"MENTIONED", direction:"OUT")
reply: Tweet @relation(name:"REPLIED_TO", direction:"OUT")
retweeted: Tweet @relation(name:"RETWEETED", direction:"OUT")
links: [Link] @relation(name:"LINKED", direction:"OUT")
tags: [Tag] @relation(name:"TAGGED", direction:"OUT")
}
type Tag @model {
name: ID!
tagged: [Tweet] @relation(name:"TAGGED", direction:"IN")
}
`;

Now we can setup the server, using typeDefs for our local Neo4j instance and twitterTypeDefs for the remote GraphQL Community Graph.

Server Setup

import { v1 as neo4j } from 'neo4j-driver';
import { Neo4jGraphQLServer } from 'neo4j-graphql-server';
const driver = neo4j.driver(
process.env.NEO4J_URI || "bolt://localhost:7687",
neo4j.auth.basic(
process.env.NEO4J_USER || "neo4j",
process.env.NEO4J_PASSWORD || "neo4j"
)
);
const GraphQLCommunityGraphDriver = neo4j.driver(
"bolt://107.170.69.23:7687",
neo4j.auth.basic("graphql", "graphql")
);
const server = Neo4jGraphQLServer({
typeDefs: typeDefs,
driver: driver,
log: true,
bindings: {
twitter: {
typeDefs: twitterTypeDefs,
driver: GraphQLCommunityGraphDriver,
readOnly: true
}
}
});
server.listen().then( ({ url }) => {
console.log(`🚀 Server ready at ${url}`);
});

Notice the readOnly parameter set to true in the configuration object for the twitter binding because the GraphQL Community Graph server only provides read access.

We can now use the same auto-generated query types produced by the Neo4j-GraphQL extension to read data from the remote GraphQL Community Graph while also using a binding to manage a local Neo4j instance.

Combined API in GraphQL Playground

Remote Query

This query obtains the first 5 most recently created Tweet nodes that contain exactly the text "GRAND". A generated Tweet resolver uses the twitter binding to delegate processing of the Tweet query type to the GraphQL Community Graph endpoint.

Request

query {
Tweet(
first: 5,
orderBy: created_desc,
filter: { text_contains: "GRAND" }
) {
id
text
}
}

Response

{
"data": {
"Tweet": [
{
"id": "1013844932883304450",
"text": "Mortality - Tool for querying US mortality data,
modeled in Neo4J with React/GraphQL interface
(GRAND stack). https://t.co/chMrWP9SqC #neo4j"
},
{
"id": "1012693204834177024",
"text": "RT @sfrechette: Just found out about this today ;-)
GRANDstack - Build full stack graph applications with
ease. https://t.co/Mjprm3ZmHo #gr…"
},
{
"id": "1012690404418846720",
"text": "RT @sfrechette: Just found out about this today ;-)
GRANDstack - Build full stack graph applications with
ease. https://t.co/Mjprm3ZmHo #gr…"
},
{
"id": "1012399297369985024",
"text": "RT @jsmonthlylondon: Want to find out more about the
#GRAND stack and @neo4j check the talk from @lyonwj
in our meetup, back in November…"
},
{
"id": "1012377131911909376",
"text": "Just found out about this today ;-) GRANDstack -
Build full stack graph applications with ease.…
https://t.co/Q002zqbbSf"
}
]
}
}

With neo4j-graphql-js

It may be that a Neo4j instance does not have the Neo4j-GraphQL endpoint available. Or perhaps the generated API offered by the extension does not cover everything in the database you want to access. In such cases, the neo4j-graphql-js package can be used to directly translate GraphQL requests to Cypher requests, as shown in the following example.

Server Setup

In this scenario, you can use the unmodified version of the twitter schema and resolvers that all use neo4jgraphqljs.

import { v1 as neo4j } from 'neo4j-driver';
import { Neo4jGraphQLServer } from 'neo4j-graphql-server';
import { neo4jgraphql } from 'neo4j-graphql-js';
const driver = neo4j.driver(
process.env.NEO4J_URI || "bolt://localhost:7687",
neo4j.auth.basic(
process.env.NEO4J_USER || "neo4j",
process.env.NEO4J_PASSWORD || "neo4j"
)
);
const GraphQLCommunityGraphDriver = neo4j.driver(
"bolt://107.170.69.23:7687",
neo4j.auth.basic("graphql", "graphql")
);
const typeDefs = `
type Link {
url: ID!
}
type TwitterUser {
id: ID!
screen_name: String!
name: String
location: String
followers: Int
following: Int
statuses: Int
profile_image_url: String
posted(first: Int = 10, offset: Int = 0): [Tweet] @relation(name:"POSTED", direction:"OUT")
}
type Tweet {
id: ID!
text: String
created: Int
favorites: Int
postedBy: TwitterUser @relation(name:"POSTED", direction:"IN")
mentioned: [TwitterUser] @relation(name:"MENTIONED", direction:"OUT")
reply: Tweet @relation(name:"REPLIED_TO", direction:"OUT")
retweeted: Tweet @relation(name:"RETWEETED", direction:"OUT")
links: [Link] @relation(name:"LINKED", direction:"OUT")
tags: [Tag] @relation(name:"TAGGED", direction:"OUT")
}
type Tag {
name: ID!
tagged: [Tweet] @relation(name:"TAGGED", direction:"IN")
}
type Query {
users(id: ID, name: String, first: Int = 10, offset: Int = 0): [TwitterUser]
tweets(id: ID, text: String, first: Int = 10, offset: Int = 0): [Tweet]
tag(name: ID!): Tag
}
`;
const resolvers = {
Query: {
users: neo4jgraphql,
tweets: neo4jgraphql,
tag: neo4jgraphql
}
};
const server = Neo4jGraphQLServer({
typeDefs: typeDefs,
resolvers: resolvers,
driver: driver,
context: {
driver: GraphQLCommunityGraphDriver
},
log: true
});
server.listen().then( ({ url }) => {
console.log(`🚀 Server ready at ${url}`);
});