Neode

Neode is a Neo4j OGM for Node JS designed to take care of the CRUD boilerplate involved with setting up a neo4j project with Node. Just install, set up your models and go.

• Getting Started
• Reading from the Graph
• Writing to the Graph
• Query Builder
• Schema

Getting Started

Installation

npm install --save neode

Usage

// index.js
import Neode from 'neode';

const instance = new Neode('bolt://localhost:7687', 'username', 'password');

Enterprise Mode

To initiate Neode in enterprise mode and enable enterprise features, provide a true variable as the fourth parameter.

// index.js
import Neode from 'neode';

const instance = new Neode('bolt://localhost:7687', 'username', 'password', true);

Usage with .env variables

npm i --save dotenv

// .env
NEO4J_PROTOCOL=neo4j
NEO4J_HOST=localhost
NEO4J_USERNAME=neo4j
NEO4J_PASSWORD=neo4j
NEO4J_PORT=7687
NEO4J_DATABASE=neo4j
NEO4J_ENCRYPTION=ENCRYPTION_OFF

// index.js
import Neode from 'neode';

const instance = Neode.fromEnv();

Additional Driver Config

Additional driver configuration can be passed as the fifth parameter in the constructor, or defined in .env:

NEO4J_ENCRYPTED=ENCRYPTION_ON                   # ENCRYPTION_ON or ENCRYPTION_OFF
NEO4J_TRUST=TRUST_SIGNED_CERTIFICATES           # TRUST_ALL_CERTIFICATES, TRUST_ON_FIRST_USE, TRUST_SIGNED_CERTIFICATES, TRUST_CUSTOM_CA_SIGNED_CERTIFICATES, TRUST_SYSTEM_CA_SIGNED_CERTIFICATES
NEO4J_TRUSTED_CERTIFICATES=/path/to/cert.pem
NEO4J_KNOWN_HOSTS=127.0.0.1
NEO4J_MAX_CONNECTION_POOLSIZE=100
NEO4J_MAX_TRANSACTION_RETRY_TIME=5000
NEO4J_LOAD_BALANCING_STRATEGY=least_connected   # least_connected or round_robin
NEO4J_MAX_CONNECTION_LIFETIME=36000
NEO4J_CONNECTION_TIMEOUT=36000
NEO4J_DISABLE_LOSSLESS_INTEGERS=false

Loading with Models

You can use the with() method to load multiple models at once.

const neode = require('neode')
    .fromEnv()
    .with({
        Movie: require('./models/Movie'),
        Person: require('./models/Person')
    });

Load from Directory

You can load a directory of models by calling the withDirectory() method.

// models/Person.js
module.exports = {
  id: {
    type: 'uuid',
    primary: true
  },
  name: 'string'
}

// index.js
instance.withDirectory(__dirname+'/models');

Defining a Node Definition

Neode revolves around the notion of node definitions, or Models. To interact with the graph, you will need to define a node, identified by a name and with a schema of properties.

instance.model(name, schema);

Schema Object

instance.model('Person', {
    person_id: {
        primary: true,
        type: 'uuid',
        required: true, // Creates an Exists Constraint in Enterprise mode
    },
    payroll: {
        type: 'number',
        unique: 'true', // Creates a Unique Constraint
    },
    name: {
        type: 'name',
        index: true, // Creates an Index
    },
    age: 'number' // Simple schema definition of property : type
});

Property Types

The following property types are supported:

• string
• number
• int
• integer
• float
• uuid
• node
• nodes
• relationship
• relationships
• Temporal
  • date
  • time
  • datetime
  • localtime
  • localdatetime
  • duration
• Spatial
  • point
  • distance

Validation

Validation is provided by the Joi library. Certain data types (float, integer, boolean) will also be type cast during the data cleansing process. For more information on the full range of validation options, read the Joi API documentation.

All Types

Boolean

Date, Time, DateTime, LocalDateTime, LocalTime

Numbers (number, int, integer, float)

Strings

Defining Relationships

Relationships can be created in the schema or defined retrospectively.

instance.model(label).relationship(type, relationship, direction, target, schema, eager, cascade, node_alias);

instance.model('Person').relationship('knows', 'relationship', 'KNOWS', 'out', 'Person', {
    since: {
        type: 'number',
        required: true,
    },
    defaulted: {
        type: 'string',
        default: 'default'
    }
});

Eager Loading

You can eager load relationships in a findAll() call by setting the eager property inside the relationship schema to true.

{
    acts_in: {
        type: "relationship",
        target: "Movie",
        relationship: "ACTS_IN",
        direction: "out",
        properties: {
            name: "string"
        },
        eager: true // <-- eager load this relationship
    }
}

Eager loaded relationships can be retrieved by using the get() method. A Collection instance will be returned.

const person = person.find({name: "Tom Hanks"})
const movies = person.get('acts_in');
const first = movies.first();

Extending a Schema definition

You can inherit the schema of a class and extend by calling the extend method.

instance.extend(original, new, schema)

instance.extend('Person', 'Actor', {
    acts_in: {
        type: "relationship",
        target: "Movie",
        relationship: "ACTS_IN",
        direction: "out",
        properties: {
            name: "string"
        }
    }
})

Reading

Running a Cypher Query

instance.cypher(query, params)

instance.cypher('MATCH (p:Person {name: $name}) RETURN p', {name: "Adam"})
    .then(res => {
        console.log(res.records.length);
    })

Running a Batch

Batch queries run within their own transaction. Transactions can be sent as either a string or an object containing query and param properties.

instance.batch(queries)

instance.batch([
    {query: 'CREATE (p:Person {name: $name}) RETURN p', params: {name: "Adam"}},
    {query: 'CREATE (p:Person {name: $name}) RETURN p', params: {name: "Joe"}},
    {query: 'MATCH (first:Person {name: $first_name}), (second:Person {name: $second_name}) CREATE (first)-[:KNOWS]->(second)', params: {name: "Joe"}}
])
    .then(res => {
        console.log(res.records.length);
    })

Get all Nodes

instance.all(label, properties)
instance.model(label).all(properties)

instance.all('Person', {name: 'Adam'}, {name: 'ASC', id: 'DESC'}, 1, 0)
    .then(collection => {
        console.log(collection.length); // 1
        console.log(collection.get(0).get('name')); // 'Adam'
    })

Get Node by Internal Node ID

instance.findById(label, id)
instance.model(label).findById(id)

instance.findById('Person', 1)
    .then(person => {
        console.log(person.id()); // 1
    });

Get Node by Primary Key

Neode will work out the model's primary key and query based on the supplied value.

instance.find(label, id)
instance.model(label).find(id)

instance.find('Person', '1234')
    .then(res => {...});

First by Properties

Using a key and value

instance.first(label, key, value)
instance.first(label).first(key, value)

instance.first('Person', 'name', 'Adam')
    .then(adam => {...})

Using multiple properties

instance.first(label, properties)
instance.first(label).first(properties)

instance.first('Person', {name: 'Adam', age: 29})
    .then(adam => {...})

Writing

Creating a Node

instance.create(label, properties);
instance.model(label).create(properties);

instance.create('Person', {
    name: 'Adam'
})
.then(adam => {
    console.log(adam.get('name')); // 'Adam'
});

Merging a Node

Nodes are merged based on the indexes and constraints.

instance.merge(label, properties);
instance.model(label).merge(properties);

instance.merge('Person', {
    person_id: 1234,
    name: 'Adam',
});

Merge On Specific Properties

If you know the properties that you would like to merge on, you can use the mergeOn method.

instance.mergeOn(label, match, set);
instance.model(label).mergeOn(match, set);

instance.mergeOn('Person', {person_id: 1234}, {name: 'Adam'});

Updating a Node

You can update a Node instance directly by calling the update() method.

instance.create('Person', {name: 'Adam'})
    .then(adam => adam.update({age: 29}));

Creating a Relationships

You can relate two nodes together by calling the relateTo() method.

model.relateTo(other, type, properties)

Promise.all([
    instance.create('Person', {name: 'Adam'}),
    instance.create('Person', {name: 'Joe'})
])
.then(([adam, joe]) => {
    adam.relateTo(joe, 'knows', {since: 2010})
        .then(res => {
            console.log(res.startNode().get('name'), ' has known ', res.endNode().get('name'), 'since', res.get('since'));  // Adam has known Joe since 2010
        });
});

Note: when creating a relationship defined as in (DIRECTION_IN), from from() and to() properties will be inversed regardless of which model the relationship is created by.

Detaching two nodes

You can detach two nodes by calling the detachFrom() method.

model.detachFrom(other)

Promise.all([
    instance.create('Person', {name: 'Adam'}),
    instance.create('Person', {name: 'Joe'})
])
.then(([adam, joe]) => {
    adam.detachFrom(joe) // Adam does not know Joe
});

### Deleting a node
You can delete a Node instance directly by calling the `delete()` method.

```javascript
instance.create('Person', {name: 'Adam'})
  .then(adam => adam.delete());

Cascade Deletion

While deleting a Node with the delete() method, you can delete any dependant nodes or relationships. For example, when deleting a Movie you may also want to remove any reviews but keep the actors.

You cna do this by setting the cascade property of a relationship to "delete" or "detach". "delete" will remove the node and relationship by performing a DETACH DELETE, while "detach" will simply remove the relationship, leaving the node in the graph.

// Movie.js
module.exports = {
  // ...
  ratings: {
    type: 'relationship',
    'relationship': 'RATED',
    direction: 'IN',
    target: 'User',
    'cascade': 'delete'
  },
  actors: {
    type: 'relationship',
    'relationship': 'ACTS_IN',
    direction: 'IN',
    target: 'Actor',
    'cascade': 'detach'
  }
};

Note: Attempting to delete a Node without first removing any relationships will result in an error.

Deleting a set of nodes

TODO

instance.delete(label, where)

instance.delete('Person', {living: false});

Deleting all nodes of a given type

instance.deleteAll('Person');
  .then(() => console.log('Everyone has been deleted'));

Query Builder

Neode comes bundled with a query builder. You can create a Query Builder instance by calling the query() method on the Neode instance.

const builder = instance.query();

Once you have a Builder instance, you can start to defining the query using the fluent API.

builder.match('p', 'Person')
    .where('p.name', 'Adam')
    .return('p');

For query examples, check out the Query Builder Test suite.

Building Cypher

You can get the generated cypher query by calling the build() method. This method will return an object containing the cypher query string and an object of params.

const {query, params} = builder.build();

instance.query(query, params)
    .then(res => {
        console.log(res.records.length);
    });

Executing a Query

You can execute a query by calling the execute() method on the query builder.

builder.match('this', 'Node')
    .whereId('this', 1)
    .return('this')
    .execute()
    .then(res => {
        console.log(res.records.length);
    });

Schema

Neode will install the schema created by the constraints defined in your Node definitions.

Installing the Schema

instance.schema.install()
    .then(() => console.log('Schema installed!'))

Note: exists constraints will only be created when running in enterprise mode. Attempting to create an exists constraint on Community edition will cause a Neo.DatabaseError.Schema.ConstraintCreationFailed to be thrown.

Dropping the schema

Dropping the schema will remove all indexes and constraints created by Neode. All other indexes and constraints will be left intact.

instance.schema.drop()
    .then(() => console.log('Schema dropped!'))