Accelatrix 1.6.2

Accelatrix

A parallel functional programming framework for in-browser processing of enumerations of business entities. v1.6.2 is compatible with ECMAScript 5, ES6, TypeScript, React, Angular and Vue.

If you would like to have a typed C#-like runtime in the browser instead of just at designtime with TypeScript, capable of type introspection, you reached the right place.

If you are a fan of LINQ for Objects and enumerations, you definitely reached the right place.

If you are looking for a low-friction way to tackle multithreading in the browser, this is your ticket in.

Accelatrix is free for non-commercial use, evaluation purposes, or commercial use without a user login wall. Detailed license at https://github.com/accelatrix/accelatrix/blob/main/LICENSE.md

Putting Accelatrix to use

You can host the minified file yourself or include the latest version hosted by the author:

<script src="https://ferreira-family.org/accelatrix/accelatrix.min.js"></script>

You can then import Accelatrix in your TS file:

import Accelatrix from "accelatrix/accelatrix";

Angular users need to unpack the import:

import * as AccelatrixFramework from 'accelatrix/accelatrix';
const Accelatrix: typeof AccelatrixFramework.Accelatrix = <any>AccelatrixFramework;

The CommonJS syntax is also supported:

const Accelatrix = require("accelatrix/accelatrix");

Node

To include Accelatrix in Node, first install with npm.

npm install accelatrix

A type introspection system

The type system of JavaScript is enhanced to include the four fundamental operations:

- GetHashCode()
- GetType()
- Equals()
- ToString()

You can now deal with classes in JavaScript at runtime as you would in C#, e.g.:

var myDog = new Bio.Mammal(8);  
var myCat = new Bio.Feline(8, 9);

var timeIsSame = (new Date()).Equals(new Date());                //true           
var areEqual = myDog.Equals(myCat);                              // false           
var myCatType = myCat.GetType();                                 // Bio.Feline           
var myCatBaseType = myCat.GetType().BaseType;                    // Bio.Mammal           
var isAnimal = myCat.GetType().IsAssignableFrom(Bio.Animal);     // true           
var enums = Bio.TypesOfLocomotion.GetType();                     // Accelatrix.EnumType 

// sample classes:

export namespace Bio
{           
    export enum TypesOfLocomotion           
    {           
        Crawl,           
        Swim,           
        Walk,           
        Fly,           
    }           

    abstract class LivingBeing
    {
        public isExtinct = false;
    }

    export abstract class Eukaryotes extends LivingBeing
    {
        private locomotion: TypesOfLocomotion = null;

        public get Locomotion(): TypesOfLocomotion
        {
            return this.locomotion;
        }
        public set Locomotion(value: TypesOfLocomotion)
        {
            this.locomotion = value;
        }
    }

    export class Animal extends Eukaryotes
    {
        public isAnimal = true;

        public constructor()
        {
            super();
        }
    }
    
    export class Mammal extends Animal
    {
        private readonly numberOfTits: number;

        public constructor(numberOfTits: number)
        {
            super();
            this.numberOfTits = numberOfTits;
        }

        public get NumberOfTits(): number
        {
            return this.numberOfTits;
        }

        public SayHello(): string
        {
            return "Hello";
        }
    }

    export class Feline extends Mammal
    {
        private readonly numberOfLives: number;

        public constructor(numberOfTits: number, numberOfLives: number)
        {
            super(numberOfTits);
            this.numberOfLives = numberOfLives == null ? 9 : numberOfLives;
            this.Locomotion = TypesOfLocomotion.Walk;
        }

        public get NumberOfLives(): number
        {
            return this.numberOfLives;
        }
    }
}

Enumerations and Functional Programming

You can now use your favourite LINQ operator functions operating on enumerations, not collections, and arrays are now enumerations as well, e.g.:

  var myEnumeration = Accelatrix.Collections.Enumerable.Range(0, 10000000)
                                            .Select(z => z % 2 == 0
                                                         ? new Bio.Feline(z % 10, 9)
                                                         : new Bio.Mammal(z % 10))
                                            .OfType(Bio.Mammal)
                                            .Where(z => z.NumberOfTits != 1)
                                            .GroupBy(z => z.NumberOfTits)

  var myResult = myEnumeration.Skip(2)
                              .Take(4)
                              .ToList()
                              .OrderBy(z => z.NumberOfTits);

Async Enumerations and Functional Programming

You can now use your favourite LINQ operator functions operating on enumerations where members are calculated async and are (cancellable) promises e.g.:

  var myEnumeration = new Accelatrix.Collections.AsyncEnumerable(Accelatrix.Collections.Enumerable.Range(0, 10000000))
                                                .Select(z => Accelatrix.Async.AsPromise(z)) // creates self-resolving promise
                                                .Skip(2)
                                                .Take(4)
                                            
  myEnumeration.ToList().then(z => console.log(z));

Typed JSON deserialization

In order to control the serialization process, the type-centric JSON serializer makes several decorators available:

- @KnownType
- @DataMember
- @OnSerializing
- @OnSerialized    
- @OnDeserializing
- @OnDeserialized

allowing for TypeScript properties to be stringified, but not the underlying members.

The deserialization also respects types and deserializes to classes instead of plain object, and can cope with functions as well.

var x = Accelatrix.Serialization.ToJSON(SerializationTests.GetClassInstance())

// '{"$type":"SerializableClass","NameProp":"Test Sat Jul 13 2024 09:12:03 GMT+0200 (Central European Summer Time)","Time":"2024-07-13T09:12:03.527"}'


var y = Accelatrix.Serialization.FromJSON(x)
console.log(y.GetType())  // SerializableClass

Even Enumerations with their functions can be serialised and deserialised:

var myEnumerable = Accelatrix.Collections.Enumerable
                             .Range(0, 10)
                             .Select(z => new Bio.Canine(z, 2))
                             .OfType(Bio.Mammal)
                             .Where(z => z.NumberOfTits % 2 == 0);  // enumeration not executed

var serialised = Accelatrix.Serialization.ToJSON(myEnumerable);     // enumeration not executed

var newEnumeration = Accelatrix.Serialization.FromJSON(serialised); // enumeration not executed

console.log(newEnumeration);
console.log(newEnumeration.ToList());  // enumeration executed

Try it yourself with these classes:

export module SerializationTests
{
    export function GetClassInstance()
    {
        return new SerializableClass("Test " + (new Date()).toString());
    }

    class BaseSerializableClass
    {
        @Accelatrix.Serialization.DataMember(false)
        private baseTime: Date = new Date();

        /** Field will be serialized and when deserialised the value will be retained. */
        @Accelatrix.Serialization.DataMember()
        public get Time(): Date
        {
            return this.baseTime;
        }
        public set Time(value: Date)
        {
            this.baseTime = value;
        }        
    }

    @Accelatrix.Serialization.KnownType
    class SerializableClass extends BaseSerializableClass
    {
        /** Field will NOT be serialized. */
        @Accelatrix.Serialization.DataMember(false)        
        private name: string;

        public constructor(name: string)
        {
            super();
            this.name = name;
        }

        /** Property will be serialized. */
        @Accelatrix.Serialization.DataMember()
        public get NameProp(): string
        {
            return this.name;
        }
        public set NameProp(value: string)
        {
            this.name = value;
        }

        @Accelatrix.Serialization.OnSerializing()        
        private OnSerializing()
        {
            console.log("About to serialize");
        }

        @Accelatrix.Serialization.OnSerialized()
        private OnSerialized()
        {
            console.log("Serialized complete.");
        }

        @Accelatrix.Serialization.OnDeserializing()
        private OnDeserializing()
        {
            console.log("About to deserialize");
        }

        @Accelatrix.Serialization.OnDeserialized()
        private OnDeserialized()
        {
            console.log("Deserialization complete.");
        }            
    }
}

Parallel execution with multithreading

a Task system with Web Workers

Ever wanted to cater for parallel execution in the browser, but find the Web Workers specification too low-level and cumbersome to be of any use?!

Do you appreciate the ellegance of C#'s Tasks and would like to have something similar in the browser? Always type-centric?!

// one-time init with the Scripts made available to the Workers (no DOM stuff)
// Accelatrix already includes itself and you do not need to worry about it
Accelatrix.Tasks.Config.Scripts.push( // ....... your scripts here

// Example 1
var myTask = new Accelatrix.Tasks.Task(z => "Hello " + z.toString(), "John Doe");
var cancellablePromise = myTask.Start();

cancellablePromise.Then(result => console.log(result))
                  .Catch(ex => console.error(ex))
                  .Finally(task => console.log(task));

// Example 2
var myData = [ new Bio.Canine.Dog(), new Bio.Canine.Wolf(), new Bio.Feline(8, 9) ]

Accelatrix.Tasks.Task.StartNew(data => data.OfType(Bio.Canine).Distinct().ToList(), myData)
                     .GetAwaiter()
                     .Then(result => console.log(result))
                     .Catch(ex => console.error(ex))
                     .Finally(task => console.log(task));

// Example 3: you can even pass enumerations and have them execute in the Web Worker
var myData = Accelatrix.Collections.Enumerable
                                   .Range(0, 100000)
                                   .Select(z => new Bio.Feline(z % 3 == 0, 9));  // nothing executed

Accelatrix.Tasks.Task.StartNew(data => data.Distinct().ToList(), myData)
                     .GetAwaiter()
                     .Then(result => console.log(result))
                     .Catch(ex => console.error(ex))
                     .Finally(task => console.log(task));


// Example 4: Stress-load with 100 parallel requests
Accelatrix.Collections.Enumerable.Range(0, 100)
                     .ForEach(z =>
                     {
                            Accelatrix.Tasks.Task.StartNew(data => data.Distinct().ToList(), myData)
                                                 .GetAwaiter()
                                                 .Finally(task => console.log("Task: " + z.toString()));
                     });


// Example 5: Combine tasks into a single resultset
Accelatrix.Tasks.CombinedTask.StartNew([
                                            new Accelatrix.Tasks.Task((a, b) => Accelatrix.Collections.Enumerable.Range(a, b).ToList(), 0, 20),
                                            new Accelatrix.Tasks.Task(() => Accelatrix.Collections.Enumerable.Range(20, 20).ToList()),
                                            () => Accelatrix.Collections.Enumerable.Range(40, 20).ToList(),
                                       ])
                              .GetAwaiter()
                              .Then(result => console.log(result))
                              .Catch(ex => console.error(ex))
                              .Finally(task => console.log(task));                   


// Example 6: Share state between parallel activies (with a cost!)
// This example will produce a single result from the task that runs first
var shared = Accelatrix.Tasks.StatefulActivity();

Accelatrix.Tasks.CombinedTask.StartNew([
					   new Accelatrix.Tasks.ActivitySet([
										z => z.Take(1),
										shared.PushAndEvaluate(z => 1,
                                                               (accumulated, mine) => accumulated.Where(z => z != null).Any()
                                                                                      ? z => z.Take(0)
                                                                                      : z => z ),
										z => z.ToList()
									  ],
									  [[0, 1, 2, 3, 4, 5]]),
					   new Accelatrix.Tasks.ActivitySet([
										z => z.Take(3),
										shared.PushAndEvaluate(z => 3,
                                                               (accumulated, mine) => accumulated.Where(z => z != null).Any()
                                                                                      ? z => z.Take(0)
                                                                                      : z => z.Take(1) ),
										z => z.ToList()
									  ],
									  [[6, 7, 8, 9,10, 11]])
					])
			       .GetAwaiter()
			       .Then(z => console.log(z))
			       .Catch(ex => console.error(ex))
			       .Finally(t => shared.Dispose())

Parallel Enumerations

Parallel execution of enumerations with a .AsParallel() that parallelises execution across different threads is possible with the .AsParallel() function, e.g.:

Accelatrix.Collections.Enumerable
                      .Range(0, 100)
                      .AsParallel() // sends everything to threads                      
                      .Select(z => "Item " + z.toString())
                      .Skip(2)
                      .Take(10)                      
                      .ToList()
                          .Catch(ex => console.error(ex))
                          .Then(z => console.log(z))                          

The location of .AsParallel() in the chain gives you control at which point operations are moved to parallel execution as DOM operations need to remain in the UI thread.