A Job Queue.
Add your jobs there and the .onJob
will be called whenever is possible.
Only one job will be in progress each time.
In the root folder of you app run:
npm install --save dyna-job-queue
interface IDynaJobQueueConfig {
parallels?: number; // default 1, the number of the parallel jobs
}
Although the signature looks quite complex, is the easiest method of the job queue ever.
It converts an already existed function/method that returns Promise
, to a job that will be added to the queue.
Once you wrapped it the only you have is to call is as you did before. Nothing is changed.
For typescript writers, there is no need even to define the TResolve
, as explicitly comes from method's definition.
example:
class NewsFeeder {
private readonly feeds: number[] = [];
private queue = new DynaJobQueue();
constructor() {
this.addFeed = this.queue.jobFactory(this.addFeed.bind(this)); // That's all
}
public addFeed(feed: number, afterDelay: number): Promise<number> {
return new Promise((resolve: Function) => {
setTimeout(() => {
this.feeds.push(feed);
resolve(feed);
}, afterDelay);
});
}
}
Adds a job and will be executed when all other jobs will be executed (FIFO) according also the priority (where is optional).
The command
is a string that will help you to understand what is this job.
The data
can be anything, let's say, the parameters for this job.
The priority
is optional. Default value is 1. Smaller numbers have priority.
example:
queue.addJob('loadConfig', {endPoint: 'http://example.com/awesomeCondig'});
queue.addJob('loadImage', {endPoint: 'http://example.com/awesomeCondig'}, 2); // <-- priotity 2
This is another way to add a job. You don't define command
and data
but directly the callback function you want to call. The callback will be called with only the done: Function
as argument.
example:
// implement an anonymous function
queue.addJobCallback((done: Function) => {
// so something special here
done();
});
// as above, define also the priority
queue.addJobCallback((done: Function) => {
// so something special here
done();
}, 2); // <-- priority 2!
// use an already implemented function
queue.addJobCallback(this.processMyJob, 2); // <-- priority 2!
addJobPromise(callback: (resolve: (data?: TResolve) => void, reject: (error?: any) => void) => void, priority: number = 1): Promise
This method adds a job with callback and returns a Promise. It create a new Promise. The callback provides two functions, the resolve
and the reject
when will fulfill the Promise. In resolve
pass the output of the Promise.
The difference with the callback of other methods is that you have to call the resolve
or reject
instead of done
; that's all!
So this method is a Promise generator. The benefit is that you can get the Promise that will be fulfilled on the proper time.
example:
queue.addJobPromise((resolve: Function, reject: Function) => {
try{
// do some work here
resolve(data);
} catch (err) {
reject(err);
}
}, 2) // <-- this 2 is the priority
.then((data: any) => {
// our resloved data are here
})
.catch((err: any)) => {
// our exception is dropped here
});
This method adds a job from a Promise.
Since the Promise by it's nature is executed instantly, you have to pass a callback that will execute the promise at the proper time.
example:
queue.addJobPromises((resolve: Function, reject: Function) => {
return fetch('http://api.example.com/customer-info?:id=4853847343');
}, 2) // <-- this 2 is the priority
.then((data: any) => {
// our resloved data are here
})
.catch((err: any)) => {
// our exception is dropped here
});
Promise that is resolved when the queue becomes empty.
jobs
is the number of the jobs that pending
running
the number of parallel running jobs
Note: it is possible to have jobs
but not running
in the rare case of switching the jobs.
First stable version
Export web
and node
versions.
You should import from dyna-job-queue/web
or dyna-job-queue/node
according your running environment.
For universal apps you should import with lazy load.
New method allDone()
.