flexloaders
v1.1.0
Published
Helper class that improves the the loading process of multiple endpoints on multiple processes # Development Guide ## About - This library is meant to be a node packaged module that optimize concurrent api requests and simplify as much as possible , th
Downloads
5
Readme
flexloaders
This Library has moved and its no more supported , Please get the supported library here
DISCLAIMER: This library is no more supported please use the jsWorkerutils library instead
- The whole paradigm of JavaScript is based on the idea of single-threaded environment but here come Web Workers which remove (partially) this limitation.Web Workers allow developers to put long-running and computationally intensive tasks on the background What’s more, no
- To incorporate this library at this stage to your existing apps you might need to tweak some parts of the library
- Helper class that improves the the loading process of endpoints on multiple processes
- Runs multiple processes in parallel.
- Uses webworkers and targets es6, es2015, es5, Typescript
- Primarily focuses on running multiple api request in parallel
- On the client side
- This library is meant to be a node packaged module that optimize concurrent api requests and simplify as much as possible , the library should be able integrate with other javascript projects indepent of frameworks.
Collaborrative Guide
- Collaborators are welcome to fork this project please read the DevelopersREADME.md form guides
- Use Webworkers to optimize request time for multiple requests
- For web applications that requires loading multiple resources this library will help optimize calls
Use Case
- import the library from conncurrent request from flexloaders
import { concurrentRequests,layersSt} from 'flexloaders'
concurrentRequest(< array of values >)
or common js
const concurrentRequests = require('flexloaders');
const layersSt = require('flexloaders');
All requests must be labelled with a layername and a url as declared below
const req = {layername: <string>, url : <string>}
Pass all requests to an array and call as follows
concurrentRequests([reqArray])
You can get the values of your request by subscibing to an observer which you imported above as
layerSt
layersSt.subscribe((item: any) =>
Check for values in your stream before accessing them as follows
- You must also define a means to check if values are different before passing them as observable might stream same values over repeated calls
if (Object.keys(item).length > 0 { item becomes your return value }
All requests uses web workers hence cutting request time by a fraction of number of cores available on the client machine