All Projects → jfet97 → Omniclone

jfet97 / Omniclone

Licence: mit
An isomorphic and configurable javascript utility for objects deep cloning that supports circular references.

Programming Languages

javascript
184084 projects - #8 most used programming language

Labels

Projects that are alternatives of or similar to Omniclone

Deepj
A deep learning model for style-specific music generation.
Stars: ✭ 681 (+270.11%)
Mutual labels:  deep
Deep Nlp
Tensorflow Tutorial files and Implementations of various Deep NLP and CV Models.
Stars: ✭ 51 (-72.28%)
Mutual labels:  deep
Dsh tensorflow
implemement of DEEP SUPERVISED HASHING FOR FAST IMAGE RETRIEVAL_CVPR2016
Stars: ✭ 97 (-47.28%)
Mutual labels:  deep
Py Style Transfer
🎨 Artistic neural style transfer with tweaks (pytorch).
Stars: ✭ 23 (-87.5%)
Mutual labels:  deep
Satellite imagery analysis
Implementation of different techniques to find insights from the satellite data using Python.
Stars: ✭ 31 (-83.15%)
Mutual labels:  deep
Mixin Deep
Deeply mix the properties of objects into the first object, while also mixing-in child objects.
Stars: ✭ 72 (-60.87%)
Mutual labels:  deep
Deep Object Diff
Deep diffs two objects, including nested structures of arrays and objects, and returns the difference. ❄️
Stars: ✭ 515 (+179.89%)
Mutual labels:  deep
Siamese Networks
Few Shot Learning by Siamese Networks, using Keras.
Stars: ✭ 146 (-20.65%)
Mutual labels:  deep
Ludwig
Data-centric declarative deep learning framework
Stars: ✭ 8,018 (+4257.61%)
Mutual labels:  deep
Neural Api
CAI NEURAL API - Pascal based neural network API optimized for AVX, AVX2 and AVX512 instruction sets plus OpenCL capable devices including AMD, Intel and NVIDIA.
Stars: ✭ 94 (-48.91%)
Mutual labels:  deep
Pytorch Forecasting
Time series forecasting with PyTorch
Stars: ✭ 849 (+361.41%)
Mutual labels:  deep
Model Free Episodic Control
This is the implementation of paper Model Free Episodic Control
Stars: ✭ 31 (-83.15%)
Mutual labels:  deep
Deeplearning Mindmap
A mindmap summarising Deep Learning concepts.
Stars: ✭ 1,251 (+579.89%)
Mutual labels:  deep
Variational Autoencoder
Variational autoencoder implemented in tensorflow and pytorch (including inverse autoregressive flow)
Stars: ✭ 807 (+338.59%)
Mutual labels:  deep
Djl Demo
Demo applications showcasing DJL
Stars: ✭ 126 (-31.52%)
Mutual labels:  deep
Amazon Sagemaker Examples
Example 📓 Jupyter notebooks that demonstrate how to build, train, and deploy machine learning models using 🧠 Amazon SageMaker.
Stars: ✭ 6,346 (+3348.91%)
Mutual labels:  deep
Deepc
Suite of Deep compositing tools for Foundry's Nuke.
Stars: ✭ 56 (-69.57%)
Mutual labels:  deep
Unsuperviseddeephomographyral2018
Unsupervised Deep Homography: A Fast and Robust Homography Estimation Model
Stars: ✭ 161 (-12.5%)
Mutual labels:  deep
Keras Rl2
Reinforcement learning with tensorflow 2 keras
Stars: ✭ 134 (-27.17%)
Mutual labels:  deep
Merge Deep
Recursively merge values in a JavaScript object.
Stars: ✭ 90 (-51.09%)
Mutual labels:  deep

NPM version license

omniclone

An isomorphic and configurable javascript function for object deep cloning.

omniclone(source [, config, [, visitor]]);

Example:

const obj = { foo: { bar: 'baz' } };
const obj2 = omniclone(obj);

obj2; // { foo: { bar: 'baz' } };
obj === obj2; // false
obj.foo === obj2.foo; // false

installation

$ npm install --save omniclone

usage

const omniclone = require('omniclone');
import omniclone from 'omniclone';

test

Test it in no time with RunKit!

strengths

  1. automatically invoke object constructors before copying properties (customizable behavior)
  2. let you to share the [[Prototype]] object between source and the resulting object (customizable behavior)
  3. let you to clone objects with circular references (customizable behavior)
  4. let you to copy getters and setters, non enumerables properties and also symbols (customizable behavior)
  5. correct handling of String, Boolean, Number, Error, Promise, Map, Set, WeakMap, WeakSet, ArrayBuffer, TypedArray and DataView objects
  6. similar references are not duplicated
  7. correct cloning of Array objects
  8. correct cloning of RegExp and Date objects
  9. let you define custom cloning logic

config

invokeConstructors (default true)

If you need to invoke the objects constructors for each object prop set the invokeConstructors flag to true:

const res = omniclone(source, {
  invokeConstructors: true
});

This option will correctly set up the new object, because constructors are invoked to create it. The resulting object and each of its object property therefore will have the [[Prototype]] and the constructor props correctly setted up, corresponding to the source object and its object properties for everyone.

class Test {
  constructor() {
    console.log('constructor invoked');
  }
};

const t = new Test(); // 'constructor invoked'
t.foo = new Test(); // 'constructor invoked'
t; // Test { t: Test {} }

const res = omniclone(t, {
  invokeConstructors: true
}); // 'constructor invoked' 'constructor invoked'

res; // Test { t: Test {} }
res instanceof Test; // true
res.foo instanceof Test; // true

It is actually a default enabled setting, but you can disable it (Array, Map and Set objects will be properly cloned anyway).
It is highly discouraged to disable this flag, do it only if you know what you are doing.

If the invokeConstructors flag is set to false, a plain new object will be created for each object prop and for the resulting object as well. So the constructor prop will be set to the Object function, and the [[Prototype]] prop will be Object.prototype.
Unless you use the setPrototype flag.

setPrototype (default false)

If the invokeConstructors flag is setted to false we could anyway share the [[Prototype]] object between the source object and the resulting object thanks to the setPrototype flag, without calling the constructors.
(Array, Map and Set objects will be properly cloned anyway because for them the constructor will be always called).
This means that the constructor prop will be shared as well because it is related to the [[Prototype]] prop.
This flag affects all the object properties as weel, like the previous flag.
If the invokeConstructors flag is setted to true, the setPrototype flag will be is ignored.

const res = omniclone(source, {
  invokeConstructors: false,
  setPrototype: true
});

The resulting object therefore will have the [[Prototype]] and the constructor props correctly setted up, but the constructors are not invoked.

class Test {
  constructor() {
    console.log('constructor invoked');
  }
};

const t = new Test(); // 'constructor invoked'
t.foo = new Test(); // 'constructor invoked'
t; // Test { t: Test {} }

const res = omniclone(t, {
  invokeConstructors: false,
  setPrototype: true
});

res; // Test { t: Test {} }
res instanceof Test; // true
res.foo instanceof Test; // true
const prot = { foo:'bar' };

const obj1 = Object.create(prot);
Object.getPrototypeOf(obj1) === prot; // true

const res = omniclone(obj1, {
  invokeConstructors: false,
  setPrototype: true
});
Object.getPrototypeOf(res) === prot; // true

copyNonEnumerables (default false)

Enable it to deep copy also non enumerables properties.
Disable it to ignore them.

const res = omniclone(source, {
  copyNonEnumerables: true
});

copySymbols (default false)

Enable it to deep copy also symbol properties.
Disable it to ignore them.
Symbols are shallow copied;

const res = omniclone(source, {
  copySymbols: true
});

copyGettersSetters (default false)

Enable it to copy also getters and setters.
Disable it to ignore them.

const res = omniclone(source, {
  copyGettersSetters: true
});

Odds are that to properly copy gets&setts you have also to enable the copyNonEnumerables flag.

allowCircularReferences (default true)

Enable it to allow circular references.
Disable it to throw an error if one is met. Know that omniclone is more performing with this flag enabled, so disable it only if you really need.

const res = omniclone(source, {
  allowCircularReferences: true
});

discardErrorObjects (default true)

Enable it to discard Error objects props.
Know that omnicopy will return null if one is passed as source.
Disable it to throw an error if one is met.

const res = omniclone(source, {
  discardErrorObjects: true
});

default config

The default config is the following:

omniclone(source, {
    invokeConstructors : true,
    setPrototype : false,
    copyNonEnumerables : false,
    copySymbols : false,
    copyGettersSetters : false,
    allowCircularReferences: true,
    discardErrorObjects: true,
});

what about String, Boolean, Number, Error, Promise, Map, Set, WeakMap, WeakSet, ArrayBuffer, TypedArray and DataView objects?

String, Boolean and Number objects passed to omniclone as sources will produce null.
Error objects passed to omniclone as sources will produce null if the discardErrorObjects is set to true (as default).
Error objects passed to omniclone as sources will produce a TypeError if the discardErrorObjects is set to false (not the predefined behaviour).

String, Boolean and Number objects props will be unwrapped.
Error objects props will be discarded if the discardErrorObjects is set to true (as default).
Error objects props will produce a TypeError if the discardErrorObjects is set to false (not the predefined behaviour).

Promise, WeakMap and WeakSet objects will be returned if passed to omniclone as sources.
Promise, WeakMap and WeakSet objects props will be copied by reference.

Map objects (keys/values) will be always deeply cloned, but any properties added to the map object itself will not be copied.
Set objects will be always deeply cloned, but any properties added to the set object itself will not be copied.
ArrayBuffer and TypedArray objects will be always deeply cloned, but any properties added to the array objects themselves will not be copied. DataView objects are copied by reference.

what about the 6th strength?

To understand it, let's compare the function omniclone with the well-know JSON.parse(JSON.stringify(source)):

const obj = { foo: 'bar'};
const source = {
  a: obj,
  b: obj,
};

JSON.stringify(source); // '{"a":{"foo":"bar"},"b":{"foo":"bar"}}'

When you will use JSON.parse(), an {"foo":"bar"} object will be created for the a prop and a {"foo":"bar"} distinct object will be created for the b prop. But this is not the initial situation where source.a == source.b; // true.

how to define custom cloning logic?

You can define a callback function that will be called on each node which copy can be customized:

function visitor(node, config) {
  // do stuff
}

The function will receive the node and a copy of the config object. If the function returns something different than undefined, that returned value will become the cloned value. On the contrary if the function returns undefined the default cloning algorithm will be used.

You cannot overwrite the default algorithm logic for String, Boolean, Number, Error, Promise, Map, Set, WeakMap, WeakSet, Date and RegExp objects. You can overwrite the default algorithm logic for Array, ArrayBuffer, TypedArray, DataView, plain js and custom objects.

Let's see an example where we add custom logic to properly clone a Node.js Buffer object:

const buffer = Buffer.from([1, 2, 3, 4]);

const resBuffer = omniclone(buffer, {}, node => {
  if (node instanceof Buffer) { // it affects only Buffer objects
    return Buffer.from(node);
  }
  return undefined; // all the other nodes will be cloned as usual
});

Thanks to the instanceof check, only Buffer objects will be affected by the intervention of the visitor callback.

warnings and limitations

  1. promises and methods are always copied by reference
  2. super is statically bound to a class heirarchy, remember it
  3. Error objects cannot be properly copied because of js limitations
  4. currently there is no isomorphic way to detect if an object is a Proxy nor is possible to access the handler object. Because of transparent virtualization, omniclone will copy each properties, the constructor and the [[Prototype]] directly from the proxed object.

support

Let me know that you like this project: star it! Thanks :)

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].