内容简介:Encode and decode modern JavaScript, e.g. Map, Set, NaN, Infinity, typed Arrays for JSON.stringify and JSON.parse.Here is an example that covers many special JavaScript cases and classes as well as a custom class:
js-codex
Encode and decode modern JavaScript, e.g. Map, Set, NaN, Infinity, typed Arrays for JSON.stringify and JSON.parse.
JSON.stringify
and JSON.parse
are of high utility for serializing JSON data and restoring it for later use. However,
they were both designed prior to the introduction of a large number of JavaScript objects that do not serialize and
subsequently restore well, i.e. Set
, Map
, all of the typed arrays like Int8Array
. Additionally, JSON.stringify
loses semantic information unless toJSON
methods are implemented for each class. The js-codex
library solves this
problem and supports serializaton preparation for all native JavaScript classes and learns custom classes without semantic loss.
Usage
Here is an example that covers many special JavaScript cases and classes as well as a custom class:
<script type="module"> import {Codex} from "../js-codex.js"; const codex = new Codex(), Person = function(config={}) { Object.assign(this,config); // create hidden metadata property Object.defineProperty(this,"^",{value:Object.assign({},config["^"])}); // add a unique id, if one does not exist if(!this["#"]) { this["#"] = `Person@${Math.random()}`; // good enough for demo } // set the createdAt metadata, if it does not exist if(!this["^"].createdAt) { this["^"].createdAt = new Date(); } }, data = { anUndefined: undefined, aBoolean: true, aNumber: 1, aString: "a string", aDate: new Date(), aBigInt: BigInt("9007199254740991"), aCustomObject: new Person({name:"joe"}), aSet: [1,NaN,Infinity,undefined] .reduce((accum,value) => accum.add(value),new Set()), aMap: [["a",1],["b",NaN],["c",Infinity],["d",undefined],[{name:"test"},{name:"test"}]] .reduce((accum,[key,value]) => accum.set(key,value),new Map()), anInt8Array: Int8Array.from([1,2,3]), aBigInt64Array: BigInt64Array.from([9007199254740991n,"9007199254740991"]) }; navigator.geolocation.getCurrentPosition(async (location) => { const references = {}; data.aGeolocation = location; const encoded = codex.encode(data,{idProperty:"#",references,hiddenProperties:["^"]}); console.log(encoded); console.log(await codex.decode(JSON.parse(JSON.stringify(encoded)),{references})); }); </script>
API
The API is presented in order of likely use.
declare({ctor,name=ctor.name,decode,create,encode})
Extends codex to support the named class. The core encoder will automatically learn new classes based on the instances it is passed. If
the same Codex instance is used for decoding, declare
will not usually be needed. However, if you are decoding on a remote device,
then you will nned to provide the functions for the Codex to use.
ctor
- The class constructor.
name
- Classname for which to support encoding and decoding.
decode
- Optional. Function to decode class. Has the same signature as decode
below. If not provided an attempt with by made to create
an instance using a create
function passed all the decoded data as an Object
. If no create function can be found, then Object.create
is used with the class prototype and the data is assigned to the instance.
create
- Optional. Factory function to create an instance based on a single config object containing all decoded data. May be asynchronous.
If not provided and needed a search for a factory will be conducted by looking at the class to see if it has a static method create
.
encode
- Optional. Function to encode class instance. Has the same signature as encode
below. If not provided, a search will be made
by first looking for an encode
method on the data passed to encode
and next for a static method named encode
on the class of the
data passed to encode
.
encode(data,{idProperty,hiddenProperties=[],references}={})
Encodes the data so it can be serialized using JSON.stringify
. Supports circular references so long as the objects referenced have an idProperty
.
data
- The data to encode. Can be anything including undefined
.
idProperty
- Optional. The key in which unique object identifiers are stored, e.g. _id
or #
.
hiddenProperties
- Optional. An array of hidden, i.e. non-enumerable, property names to include in the encoded data.
references
- An object, the entries of which will be unique object ids and objects when encode
returns.
async decode(data,{isReference,references}={})
Decodes data. It is asynchronous because decoding data will frequently require asynchronous retrieval of referenced objects form a database based on their ids.
data
- The data to decode. Can be anything.
isReference(value)
- A function that returns truthy if the value passed is an object reference, i.e. an id pulled from the idProperty
field
specified with encode
.
references
- Either a function or an object. If a function, when passed a unique object id it should return the object. It may be asynchronous.
Typically, this will be a database getter. Or, an object, the keys of which are unique object ids and the values are objects to substitute
for the ids, e.g. the references
object populated by encode
.
License
MIT
Release History (reverse chronologicla order)
2020-02-16 v0.0.2a ALPHA Corrected typo in declare function signature.
2020-02-16 v0.0.1a ALPHA Initial public release
以上就是本文的全部内容,希望本文的内容对大家的学习或者工作能带来一定的帮助,也希望大家多多支持 码农网
猜你喜欢:本站部分资源来源于网络,本站转载出于传递更多信息之目的,版权归原作者或者来源机构所有,如转载稿涉及版权问题,请联系我们。
算法分析-有效的学习方法(影印版)
Jeffrey J.McConnell / 高等教育出版社 / 2003-03-01 / 28.0
本书主要目标是提高读者关于算法对程序效率的影响等问题的认知水平,并培养读者分析程序中的算法所必需的技巧。各章材料以激发读者有效的、协同的学习方法的形式讲述。通过全面的论述和完整的数学推导,本书帮助读者最大限度地理解基本概念。 本书内容包括促使学生参与其中的大量程序设计课题。书中所有算法以伪码形式给出,使得具备条件表达式、循环与递归方面知识的读者均易于理解。本书以简洁的写作风格向读者介绍了兼具......一起来看看 《算法分析-有效的学习方法(影印版)》 这本书的介绍吧!