README.md

json-enhance

Extend JSON.stringify() & JSON.parse() to handle more values:

• Date
• undefined
• Function
• RegExp
• Getter
• Setter

And both stringify() and parse() are safe:

• No modification on original data.

• No special markup on special values. Auxiliary data structure is used to mark special values.

  For example, if we mark Date instance with prefix __DATE__, it may conflicts with real values in some cases.

• Identical before stringify() and after parse().

How to use

npm i --save json-enhance

import { stringify, parse } from 'json-enhance'

const obj = {
    a: 1,
    b: true,
    c: [
        '123',
        new Date(2018, 0, 1),
        { x: '123' },
        (a, b) => a - b,
    ],
    d: undefined,
    e: /\/(.*?)\/([gimy])?$/,
    f: (a, b) => a + b,
}
Object.defineProperty(obj, 'x', {
    get() {
        return this.mX
    },
    set(v) {
        this.mX = v
    },
})
const str = stringify(obj)
parse(str) // equals obj

Why

Mainly used for state recoverage of web app. See my article here.

Pitfalls

toString()

Functions, RegExp, Getter, Setter are compared by calling toString() and then direct string matching.

Closure

If functions access variables in the upper closure, calling them after parse() may cause error. Keep your functions pure.

Getter/Setter

There are several ways to write getter/setter, however, one of them is unapplicable to parse():

✅

Object.defineProperty(obj, 'x', {
    get() {
        return this.mX
    },
    set(v) {
        this.mX = v
    },
})

✅

obj.__defineGetter__('x', function () {
    return this.mX
})
obj.__defineSetter__('x', function (v) {
    this.mX = v
})

❌

const obj = {
    get x() {
        return this.mX
    }
    set x(v) {
        this.mX = v
    }
}