1.2.0 • Published 2 months ago

typescript-flat-serializer v1.2.0

Weekly downloads
-
License
MIT
Repository
github
Last release
2 months ago

typescript-flat-serializer

Build status Coverage Status Known Vulnerabilities npm bundle size GitHub npm

A typescript library to serialize/deserialize classes to/from string in a flat format. Supports inheritance, circular reference and more

Summary

  1. Why
  2. Installation
  3. Usage
  4. API
  5. Important notes
  6. Inspired by

Why

When I was developing a electron app I needed a serializer that fill some requirements to use with IPC:

  • After serialize I wanted that the object keep the methods (right prototype)
  • I needed inheritance
  • Supports circular reference is always good

Installation

npm install typescript-flat-serializer --save

Usage

With decorators

Preparation

If you want to use our decorators you also need to set experimentalDecorators and emitDecoratorMetadata to true into the tsconfig.json file.

For example:

{
  "compilerOptions": {
    ...
    "emitDecoratorMetadata": true,
    "experimentalDecorators": true,
    ...
  }
}

Usage

First let's create some models

import { TSFlatCollection, TSFlatObject } from "typescript-flat-serializer";

@TSFlatObject()
export abstract class Animal {
  protected constructor(public name: string) {
  }
}

// This decorator NEEDS to be placed on every class that you want to serialize. 
// Without this decorator the behavior will be like stringify/parse from JSON.
@TSFlatObject()
export class Dog extends Animal {
  // This decorator will take care of serialize/deserialize our collection
  @TSFlatCollection({collectionType: "set"})
  favoriteFoods: Set<Food>;

  constructor(name: string, public beautiful: boolean, favoriteFoods: Set<Food>) {
    super(name);
    this.favoriteFoods = favoriteFoods;
  }
}

@TSFlatObject()
export class Food extends Animal {
  constructor(name: string) {
    super(name);
  }
}

Now we only need to serialize/deserialize the animal.

import { parse, stringify } from "typescript-flat-serializer";


const foods = new Set([new Food('all')])
const animal: Animal = new Dog('Luffy', true, foods);

// Let's go serialize our animal
const str = stringify(animal)
// value of str: [{"name":"Luffy","beautiful":true,"favoriteFoods":"#_1_#","__class__":"Dog"},["#_2_#"],{"name":"all","__class__":"Food"}]

// And now we can deserialize the animal
const parsedAnimal = parse<Animal>(str);

You can find another examples of utilisation on tests.

Without decorators

First let's create some models

import { registerTSFlat } from "typescript-flat-serializer";

export abstract class Animal {
  protected constructor(public name: string) {
    registerTSFlat(
      {
        target: Animal,
      }
    );
  }
}

export class Dog extends Animal {
  favoriteFoods: Set<Food>;

  constructor(name: string, public beautiful: boolean, favoriteFoods: Set<Food>) {
    super(name);
    registerTSFlat(
      {
        target: Dog,
      },
      { collectionName: 'favoriteFoods', options: {collectionType: 'set'} }
    )
    this.favoriteFoods = favoriteFoods;
  }
}

export class Food extends Animal {
  constructor(name: string) {
    super(name);
    registerTSFlat(
      {
        target: Food,
      },
    )
  }
}

Now we only need to serialize/deserialize the animal.

import {parse, stringify} from "typescript-flat-serializer";


const foods = new Set([new Food('all')])
const animal: Animal = new Dog('Luffy', true, foods);

// Let's go serialize our animal
const str = stringify(animal)
// value of str: [{"name":"Luffy","beautiful":true,"favoriteFoods":"#_1_#","__class__":"Dog"},["#_2_#"],{"name":"all","__class__":"Food"}]

// And now we can deserialize the animal
const parsedAnimal = parse<Animal>(str);

API

Decorators

@TSFlatObject

Used to make a class serializable.

Example
@TSFlatObject()
export class Dog {
}

@TSFlatCollection

Used do make a Array|Set|Map|Dictionary

Example
@TSFlatObject()
export class Dog {
  @TSFlatCollection({collectionType: "map"})
  placesVisited: Map<string, boolean>
}
Parameters

collectionType Type: CollectionTypeString
Optional: false
Description: The way to specify the type of collection

@TSFlatProperty

Used modify the serialization/deserialization of a property.

Example
@TSFlatObject()
export class Dog {
  @TSFlatCollection({collectionType: "map"})
  @TSFlatProperty({
    beforeStringify: (m) => {
      m.set('Brazil', true);
    }
  })
  placesVisited: Map<string, boolean>
}
Parameters

options Type: TSFlatPropertyOptions
Optional: true
Description: The option to customize the serialization/deserialization of the target property.

Methods

registerTSFlat

Used do register an object and its items.

registerTSFlat<T>(objectRegistration: FlatObjectRegistration<T>, ...items: FlatItem<T>[])
Parameters

objectRegistration

Type: FlatObjectRegistration<T>
Optional: false
Description: Information about the class that we want to make serializable.

Type: FlatItem<T>[]
Optional: true
Description: Items that this class have that we want to make serializable.


stringify

Used to serialize a object.

stringify(obj: any, options?: StringifyOptions): string
Parameters

obj

Type: any
Optional: false
Description: The object that will be serialized

options

Type: StringifyOptions
Optional: true
Description: Custom options to the serialization

Return

string


parse

Used to deserialize a string into a object.

parse<T>(str: string): T
Return

T

Definitions

Types

FlatObjectRegistration
export type FlatObjectRegistration<T> = {
  target: Type<T>,
  options?: TSFlatObjectProperties
}
FlatItem\
export type FlatItem<T> = FlatPropertyRegistration<T> | FlatCollectionRegistration<T>;
FlatPropertyRegistration\
export type FlatPropertyRegistration<T> = {
  propertyName: (keyof T), options?: TSFlatPropertyOptions,
}
FlatCollectionRegistration\
export type FlatCollectionRegistration<T> = {
  collectionName: (keyof T), options: TSFlatCollectionOptions
}
CollectionTypeString
export type CollectionTypeString = 'array' | 'dictionary' | 'map' | 'set';
TSFlatPropertyOptions
export type PropertyTransformer = (property: any) => any;

export interface TSFlatPropertyMetadata {
    beforeStringify?: PropertyTransformer;
    afterParse?: PropertyTransformer;
}

export interface TSFlatPropertyOptions extends TSFlatPropertyMetadata {
}

stringifyOptions

export type StringifyOptions = {
  rFDCOptions?: RFDCOptions
}

export type CustomWayOfCloningObjectMap = Map<Type<any>, (obj: any) => any>;

export type RFDCOptions = {
  customWayOfCloningObject?: CustomWayOfCloningObjectMap
}

Important notes

Cloning

This library before the serialization makes a clone of the object. By default the cloning supports the types:

  • Object
  • Array
  • Number
  • String
  • null
  • Date
  • undefined
  • Buffer
  • TypedArray
  • Map
  • Set
  • Function
  • AsyncFunction
  • GeneratorFunction
  • arguments

To support other type, like DateTime of Luxon, you should do something like that:

const customWayOfCloningObject: CustomWayOfCloningObjectMap = new Map();
const rFDCOptions: RFDCOptions = {
  customWayOfCloningObject
}

customWayOfCloningObject.set(DateTime, (obj) => DateTime.fromMillis(obj.toMillis()));

const str = stringify(obj, {rFDCOptions});

Inspired by

1.2.0

2 months ago

1.1.0

6 months ago

1.0.0

11 months ago

0.0.9

11 months ago

0.0.8

11 months ago

0.0.7

11 months ago

0.0.6

11 months ago

0.0.5

11 months ago

0.0.4

11 months ago

0.0.3

11 months ago

0.0.2

11 months ago

0.0.1

11 months ago