node package manager
We need your input. Help make JavaScript better: Take the 2017 JavaScript Ecosystem survey »


Node2Neo Schema

Schema support for Node2Neo models

This is intended to be used with the node2neo modules but it is generic to be used for any purposes

NOTE: Neo4j 2.0 is required.


npm install node2neo-schema


Define a schema

var Schema = require('node2neo-schema');
var rawUserSchema = {
  first: {type: String, required: true, trim: true, match: /name/},
  birthday: Date,
  likes: Array,
  alive: {type: Boolean, default: true},
  age: {type: Number, min:5, max:90}
  gender: {type:String, enum:['male', 'female']}
var options = {label: 'User'};
var userSchema = new Schema(rawUserSchema, options);

Schemas support the following data types:

  1. String
  2. Number
  3. Date
  4. Array
  5. Boolean

The following options are available for each schema item:

  1. required
  2. default (can be a function or a value)
  3. validate (pass in a custom validator function)
  4. match (supply a regular expression)
  5. String: lowercase, uppercase, trim, enum
  6. Number: min, max

Schemas can be defined to be strict or not, by default schemas are strict. If you pass in a strict:false value in the option the schema will not be strict. Strict means that only the fields defined in the schema can be saved. Any new fields will fail validation.

var options = {
  label: 'User',
  strict: false,
  unique: true  // sets the schema._unique flag. Can be used to CREATE UNIQUE nodes 

Static Methods

You can add static methods to a schema.

var schema = new Schema({
  name: String
}, {label: 'Blue'});
schema.static('turnBlue', function(obj){ = 'blue';
  return obj;
//using node2neo models 
var m = Model.model('Blue', schema);
var sample = {name: 'Green'};
sample = m.turnBlue(sample); // sample.namenow equals blue; 

Schema Wide Preparers

Sometimes you want to perform some calculations on your data - e.g. update a field based on the values in other fields. You can add schema level preparers to a schema.

Preparers are functions which take the object being created as an argument. This object must be returned after all of the changes have been made.

var prepSchema = new Schema({
  year: Number,
  month: Number,
  day: Number,
  timestamp: Date,
  approximate: Boolean
}, {label: 'prepSchema'});
  if(!obj.year || !obj.month || !{
    obj.approximate = true;
  else {
    obj.timestamp = new Date(obj.year, obj.month,;
  return obj;

Schema Wide Validators

Sometimes you need to perform validations on multiple fields (e.g. it is only valid if two out of three fields are present, but any two).

Validators are functions which take the object being created as an argument. Validators must return either true or false.

var valSchema = new Schema({
  year: Number,
  month: Number,
  day: Number
}, {label: 'valSchema'});
  var keys = Object.keys(obj);
  if(keys.length < 2){
    return false;
  else {
    return true;


The main point of defining a schema is for hassle free validation.

``js var newUser = { first: 'Name', age: 30 }

userSchema.validate(newUser, function(err, user){ // user object will be different if transform methods used e.g. cast to type, trim etc. })

The following validations/mainpulations are performed:
1. Cast to Type
Changes input into the desired object type. Will fail if an invalid type is supplied (e.g. if a value is supoed to be a date and you pass 'blue' into the field)
2. String manipulation (trim, ,uppercase, lowercase)
3. Number validation (min, max)
4. Enum validation
5. Regular expression validation
6. Strict schema validation
Will error if a field is attempted to be save and is not defined in the schema
The validation will return on the first error. I'm open to change on this so let me know if you would prefer an array of all errors.
#### Indexes
You can supply indexes to a schema. This module is database independent so they will not be applied but will be available to be created.
var rawSchema = {
  first: {type: String, index: true},
  last: {type: String, unique: true},
  email: {type: String, index: {unique: true}},
var options = {label: 'User'}
userSchema = new Schema(rawSchema, options)

The structure of the userSchema response is as follows:

  _strict = true || false;
  _fields = undefined || [];
  //validation options 
  _defaults = {
    alive: true
  _required = ['first', 'email'];
  _types = {
    first: 'string'
  _enum = {
    gender: ['male', 'female']
  _match = {
    email: /email/
  _number = {
    age: {
      min: 5,
      max: 90
  _string = {
    first: ['trim', 'lowercase']
  // to store transactions for indexes/constraints 
  _indexes = ['email'];
  _constraints = ['securityNumber'];
  _appliedConstraints = []; // this is used as Neo4j errors if you attempt to re-apply a constraint 
  _appliedIndexes = [];