1.1.0 • Published 6 years ago
sequelize-json-serializer v1.1.0
Sequelize-Json-Serializer
A simple and fast convention based library for serializing sequelize models to JSON.
Basic Usage
conse Sequelize = require("sequelize");
const Serialzer = require("sequelize-json-serializer");
const User = sequelize.define("users",{
id: { type: Sequelize.INTEGER , primaryKey: true , autoIncrement: true },
username: { type: Sequelize.STRING(128) },
email: { type: Sequelize.STRING(256) },
phone: { type: Sequelize.STRING(128) },
country: { type: Sequelize.STRING }
});
const UserSchema = {
fields: [
'username',
'email',
'phone',
'country'
]
};
Serializer.defineSchema( User , UserSchema ); Somewhere in your application:
router.get('/user/:id',function(req,res,next) {
User.findById(req.params.id).then(user => {
const payload = Serializer.serialize( user,User);
/*
Response Body...
{
"id" : "2",
"username": "user360",
"email": "test@live.com",
"phone": "02XXXXXXXX",
"country": "Ghana"
}
*/
res.status(200).json( payload );
});
}); Another Basic Example
const UserSchema = {
fields: '*' // Include all fields
};
...
const payload = Serializer.serialize( user,User);
/*
Response Body...
{
"id" : "2",
"username": "user360",
"email": "test@live.com",
"phone": "02XXXXXXXX",
"country": "Ghana"
}
*/ The default convention ensures the primary key attribute is included by default. To prevent this behavior, set excludePK option in the schema to true .
const UserSchema = {
fields: '*', // Include all fields
options: {
excludePK: true // ignores all primary keys
}
};
...
const payload = Serializer.serialize( user,User);
/*
Response Body...
{
"username": "user360",
"email": "test@live.com",
"phone": "02XXXXXXXX",
"country": "Ghana"
}
*/ The serialize method works for both arrays and sequelize models
const UserSchema = {
fields: [
'username',
'email',
'phone',
'country'
],
options: {
excludePK: true // Exclude primary key 'id'
}
}
...
/*
Response Body...
{
"username": "user360",
"email": "test@live.com",
"phone": "02XXXXXXXX",
"country": "Ghana"
}
*/ Formatting fields
Formatting can be applied to defined fields in schema to alter the output of values. See how
const UserSchema = {
fields: [
'username',
'email',
'phone',
'country'
],
options: {
formatter: (key,value, model) => {
return value.startsWith('$') ? null : value;
}
}
}
Renaming Fields
Fields can be renamed by setting the fields to an object following the syntax
[raw_field] : [new_name] Rename Field Example
const UserSchema = {
fields: {
...
'username': 'user_name',
'phone': 'phone_number',
'email': '' // no renaming done here
}
};
User.findById(2).then(user => {
// serialize model to json object
const result = Serializer.serialize(user,User);
/*
result ...
{
"user_name:" "user260",
"phone_number" : "02XXXXXXXXX"
"email": "test@email.com"
}
*/
}); Serializer Options
You can control how models are serialized by passing an options argument to the serializer
|Option | Description |
|-|--|
| serializer | Indicates the type of serializer to use. Possible values include 'model' and 'raw'. Model is only supported at the moment |
| emptyPolicy | Defines the policy that is used when a field is not available or found during serialization. Possible values include SET_NULL, SET_UNDEFINED and IGNORE (default). |
|includeAll| Controls whether to automatically serialize all include properties. Default is false
Set options
const Serializer = require("sequelize-json-serializer");
Serializer.setOptions({
serializer: "model",
includeAll: true, // Forces all included properties to be serialied
emptyPolicy: Serializer.NullPolicy.IGNORE // Will ignore any field that cannot be found in the model beign serialized
}) Including Other Defined Schemas
You can mix a schema with others by specifying them in the include section.
Example
const Sequelize = require("sequelize");
const sequelize = /*setup db connection*/
const Serializer = require("sequelize-json-serializer");
const Teacher = sequelize.define("teacher",{
id: { type: Sequelize.INTEGER , primaryKey: true , autoIncrement: true },
school_id: { type: Sequelize.INTEGER }, // make this a foreign key
name: Sequelize.STRING
});
const TeacherSchema = {
fields: ['name']
};
const School = sequelize.define("school", {
id: { type: Sequelize.INTEGER , primaryKey: true , autoIncrement: true },
name: Sequelize.STRING,
location: Sequelize.STRING,
courses: Sequelize.STRING
});
const SchoolSchema = {
fields: [
'name',
'location',
'courses'
],
include: {
'teacher': Teacher // Simplified syntax
},
options: {
// Define formatting for single properties. Only works on fields...
propertyFormat: {
name: (value) => {
return value.trim(); // always trim school name
},
location: value => {
return value.replace(/,/g,'.');
},
courses: (value) => {
return value ? value.split(',') : []
}
}
}
}
// Define User schema
const UserSchema = {
fields: [
'username',
'email'
'phone',
'country'
],
include: {
// define property 'school' on output model
'school': {
model: School, // the sequelize model.
field: 'schools', // the raw field name in users model
association: 'many' // [single] or [many]
}
}
};
// Define school schema
Serializer.defineSchema( School , SchoolSchema );
// Define teacher schema
Serializer.defineSchema( Teacher, TeacherSchema );
// Define user schema
Serializer.defineSchema( User, UserSchema );
// Usage
--------------------------------------------
// ### Single Entity
User.findById(100, {
include: {
model: School,
include: Teacher
}).then(user => {
//
const result = Serializer.serialize(user,User , {
include: { all: true } // include all properties defined in include section.
});
/*
{
"id" : 2
"username": "user360" ,
"email": "user@email.com",
"phone": "02XXXXXXXX",
"country": "Ghana",
"school": {
"id": 1,
"name": "Havard",
"location" : "US"
"courses": [ "Quantum Computing" , "Electrical Engineering" ],
"teacher": {
"id": 20,
"name": "Albert Massouli"
}
}
}
*/
}); Schema Tags
Schema tags allow defining multiple schema for a single model. It can be useful in scenarios where responses are different for other entities (eg. user account types). Below is a typical example
// #Example
const Request = sequelize.define("request", {
id : { type: Sequelize.INTEGER , primaryKey: true, autoIncrement: true },
type: { type: Sequelize.INTEGER },
name: { type: Sequelize.STRING },
email: { type: Sequelize.STRING }
},{
timestamps: true,
updatedAt: "last_updated",
createdAt: "date_created"
})
const User = sequelize.define("user",{
id : { type: Sequelize.INTEGER , primaryKey: true, autoIncrement: true },
fullname: { type: Sequelize.STRING },
email: { type: Sequelize.STRING },
user_type: { type: Sequelize.STRING }, // Distinguishes users. Can be 'regular'or 'admin'
},{
timestamps: true,
updatedAt: "last_updated",
createdAt: "date_created"
});
Defining schema for different user accounts (or roles)
Admin Schema
Here we define the response structure for administrative calls
const AdminRequestSchema = {
fields: {
'type': 'request_type',
'date_created': '', // no renaming..
'last_updated': ''
},
include: {
'sender': {
model: User,
field: "user",
// association: "single" --- Default is always single
}
}
}
// Define schema
Serializer.defineSchema( AdminRequestSchema , Request , "admin" ); // don't forget to set the last parameter to 'admin'
Regular User Schema
Define response structure of regular user calls
const RegularRequestSchema = {
fields: {
'type': 'request_type',
'name': 'sender_name'
}
}
// Define schema for regular users
Serializer.defineSchema( RegularRequestScheam , Request ,"regular"); Testing schema
We've now defined two schemas for Request model. We can serialize based on user account and prevent eavesdropping on other detailed properties of the sender's user account to a regular user call.
router.get('/:id', /**authentication middleware here*/, function(req,res,next){
const user = req.user;
Request.findById(req.params.id,{
include: {
model: User
}
}).then(request => {
const payload = Serializer.serialize( request , Request , {
include: { all: true },
tags: [ user.user_type ] // set the tags for the schema to use
});
/*
user_type = regular
--------------------------------------
{
"id": 2,
"request_type": 5,
"sender_name": "Test User"
}
user_type = admin
--------------------------------------
{
"id": 2,
"request_type": 5,
"sender": {
"id" : 1,
"fullname": "Test User",
"email": "test@email.com",
"user_type" : "admin",
"date_created": "2018-07-12T10:55:28.081Z",
"last_updated": "2018-07-12T10:55:28.081Z"
}
}
*/
})
});