npm install mini-queue
Job queue
If you have different needs regarding the functionality, please add a feature request.
``sh`
npm install --save mini-queue
`
QueueJob State Diagram (methods of JobQueue object)
======================
|
createJob |
+------V-----+
| new |
| |
+---+--+--+--+
| | | _rejectJob +------------+
| | +--------------------------------> reject |
| | | |
_startJob | | _queueJob +------------+
| +------------+
| |
| +-----v------+ _cancelJob
| | queue +-----------------+
| | | |
| +---+----^---+ |
| _dequeueJob | | |
| | | |
| | | _queueJob |
| +---v----+---+ +-V----------+
| | dequeue +---------------> cancel |
| | | _cancelJob | |
| +-----+------+ +------------+
| _startJob |
| |
| +----------+
| |
+---v----v---+ _terminateJob +------------+
| process +----------------------> terminate |
| | |(not implem)|
+-----+------+ +------------+
|
|
|
+-----v------+
| complete |
| |
+------------+
`
job.journalEntry for each state (queue, dequeue, process, complete, reject, cancel) stores the time when transition to state occured. If several changes has occured, only the last time is stored(id is job identifier job.id):
`json
{ id: 4,
new: 2017-11-10T08:37:17.428Z,
queue: 2017-11-10T08:37:17.428Z,
dequeue: 2017-11-10T08:37:17.941Z,
process: 2017-11-10T08:37:17.941Z,
complete: 2017-11-10T08:37:18.943Z },
`
For each group and name as provided in option for createJob(), journalEntries are kept ar array in queue.journal (newest is the first, oldest is the last).
Example (group and name not set, default value is used):
`json
journal: { group:
{ name:
[ { id: 5,
new: 2017-11-10T08:37:17.929Z,
reject: 2017-11-10T08:37:17.929Z },
. . .
{ id: 2,
new: 2017-11-10T08:37:16.426Z,
queue: 2017-11-10T08:37:16.427Z,
dequeue: 2017-11-10T08:37:16.937Z,
process: 2017-11-10T08:37:16.937Z,
complete: 2017-11-10T08:37:17.941Z } ] } } +2s
`
Up to maxJournalLength option for createJob() records are kept.
You may find this example in demo subdirectory of the package.
`js
"use strict";
process.env.DEBUG = 'queue,app';// + (process.env.DEBUG || '');
var util = require('util');
var debug = require('debug')('app');
//var Queue = require('express-queue');
var Queue = require('../');
var queue = new Queue({ activeLimit: 1, queuedLimit: 1, maxJournalLength: 4 });
// create jobs
var maxCount = 5,
count = 0;
var interval = setInterval(function() {
var jobData = {};
// Create new job for the queue
// If number of active job is less than activeLimit, the job will be started on Node's next tick.
// Otherwise it will be queued.
var job = queue.createJob(
jobData, // we may pass some data to job when calling queue.createJob() function
{ group: 'group', name: 'name' } // group/name to be used for journal
);
if (++count >= maxCount) {
clearInterval(interval);
setTimeout(()=> { // after last job has finished
debug('journal:', util.inspect(queue.journal, {depth:3}));
}, 1500);
}
}, 500);
// execute jobs
queue.on('process', function(job, jobDone) {
debug(queue.on('process'): [${job.id}]: status: ${job.status}, journalEntry: ${JSON.stringify(job.journalEntry)});
// Here the job starts
//
// It is also possible to do the processing inside job.on('process'), just be careful
// to call jobDone() callback once and only once.
//
// Value of job.data is set to value passed to queue.createJob()
//
// Imitate job processing which takes 1 second to be finished
setTimeout(function() {
// Call the callback to signal to the queue that the job has finished
// and the next one may be started
jobDone();
// Now on Node's next tick the next job (if any) will be started
}, 1000);
});
// Signal about jobs rejected due to queueLimit
queue.on('reject', function(job) {
debug(queue.on('reject'): [${job.id}]: status: ${job.status}, journalEntry: ${JSON.stringify(job.journalEntry)});``
});
github.com npmjs.com travis-ci.org coveralls.io inch-ci.org
MIT