Controllable and resonable job queue manager
npm install jobman
``
npm install --save jobman
The example code:
`javascript
const jobman = require('jobman')
const man = jobman({ // create new jobman object
max: 3, // max concurrent jobs
jobStart: (job, man)=>{ // when A job will start
if(job.prop==5) return false // return FALSE will cancel the job
},
jobEnd: (job, man)=>{ // when A job did end
console.log('result: ', job.prop, job.error || 'ok')
},
jobTimeout: (job, man)=>{ // when A job timeout
console.log('timeout: ', job.prop)
},
allStart: (info, man)=>{ // all jobs will start
console.log('all jobs will start')
},
allEnd: (info, man)=>{ // all jobs done
console.log('all end', man.isDone)
// also clear all pending jobs, cancel running callbacks
man.clear(true)
},
autoStart: true // start monitor when create
})
for(let i=0;i<10;i++){
man.add(cb=>{ // add A job
task(i, cb)
}, i) // job.prop = i
}
man.add(0, cb=>{ // insert A job into 0 index
task(99, cb)
}, {id: 'insert to first!', timeout: 500}) // the job timeout is 500ms
function task(i, cb){ // dummy task function
setTimeout(()=>{
if(i==3) cb('bad') // cb with error
else cb() // cb with ok
}, 1000) // async happen at 1 second
}
`
The result:
`
all jobs will start
timeout: { id: 'insert to first!', timeout: 500 }
result: 0 ok
result: 1 ok
result: 2 ok
result: 3 bad
result: 4 ok
result: 6 ok
result: 7 ok
result: 8 ok
result: 9 ok
all end true
#### jobman(config) -> manObject
- config
- config.max int
> max number of concurrence jobs.
- config.timeout int
> ms to wait for --A job--, trigger jobTimeout event after the time
- config.interval int
> ms to pass into setInterval check internally
- config.jobTimeout fn(job, man)->boolean
> callback function after --A job-- timeout, return false will book another timeout
- config.jobStart fn(job, man)->boolean
> callback function before --A job-- will start, return false will cancel this job
- config.jobRun fn(job, man)->void
> callback function after --A job-- did start, job.state become run
- config.jobEnd fn(job, man)->void
> callback function after --A job-- callback invoked, with man.lastError set to result error
- config.allStart fn(info, man)->void
> callback function before --ALL job-- start run, useful for init, info is user passed with man.start(info)
- config.allEnd fn(info, man)->void
> callback function when --ALL job-- finished run, info is user passed with man.end(info)
- manObject
- man.add([position:int], jobFn:cb=>{}, jobProp:any)
> jobFn is with callback for one job, cb(err) have to be called for each job. jobProp will become job.prop
- man.start([info])
> function to start current jobman, use man.stop to stop it, trigger man.allStart with first arg set to info
- man.stop()
> function to stop current jobman, use man.start to start again
- man.end([info], [cancelPending])
> trigger man.allEnd with first arg set to info, cancelPending will cancel pending jobs
- man.clear(cancelRunningJobs)
> Remove all pending jobs, cancel running jobs if cancelRunningJobs is true
- man.config object
> the config object passed into jobman
- man.jobs array
> the jobs array internally, query for it for state, length etc.
- man.isRunning boolean
> prop to get if the job monitor is is running (not stopped)
- man.isDone boolean
> prop to get if all job ended
- man.pending [jobObject, ...]
> prop to get pending jobs in queue
- man.running [jobObject, ...]
> prop to get running jobs in process
- man.slot int
> prop to get current available job runner slot
- man.startCount int
> jobman start counter, from 0, after first start() will be 1, after next start() plus 1
- jobObject
- job.fn function
> the jobFn function passed into man.add()
- job.error any
> when job callback called with error, set to it with the error
- job.prop any
> the prop to passed into man.add()
- job.prop.timeout int
> set timeout for individual job
- job.state any
> internal state for each job, value is run/done/error/timeout/cancel