How to implement the lodash _.throttle in vanilla javascript? [duplicate] - javascript

I am looking for a simple throttle in JavaScript. I know libraries like lodash and underscore have it, but only for one function it will be overkill to include any of those libraries.
I was also checking if jQuery has a similar function - could not find.
I have found one working throttle, and here is the code:
function throttle(fn, threshhold, scope) {
threshhold || (threshhold = 250);
var last,
deferTimer;
return function () {
var context = scope || this;
var now = +new Date,
args = arguments;
if (last && now < last + threshhold) {
// hold on to it
clearTimeout(deferTimer);
deferTimer = setTimeout(function () {
last = now;
fn.apply(context, args);
}, threshhold);
} else {
last = now;
fn.apply(context, args);
}
};
}
The problem with this is: it fires the function once more after the throttle time is complete. So let's assume I made a throttle that fires every 10 seconds on keypress - if I do keypress 2 times, it will still fire the second keypress when 10 seconds are completed. I do not want this behavior.

I would use the underscore.js or lodash source code to find a well tested version of this function.
Here is the slightly modified version of the underscore code to remove all references to underscore.js itself:
// Returns a function, that, when invoked, will only be triggered at most once
// during a given window of time. Normally, the throttled function will run
// as much as it can, without ever going more than once per `wait` duration;
// but if you'd like to disable the execution on the leading edge, pass
// `{leading: false}`. To disable execution on the trailing edge, ditto.
function throttle(func, wait, options) {
var context, args, result;
var timeout = null;
var previous = 0;
if (!options) options = {};
var later = function() {
previous = options.leading === false ? 0 : Date.now();
timeout = null;
result = func.apply(context, args);
if (!timeout) context = args = null;
};
return function() {
var now = Date.now();
if (!previous && options.leading === false) previous = now;
var remaining = wait - (now - previous);
context = this;
args = arguments;
if (remaining <= 0 || remaining > wait) {
if (timeout) {
clearTimeout(timeout);
timeout = null;
}
previous = now;
result = func.apply(context, args);
if (!timeout) context = args = null;
} else if (!timeout && options.trailing !== false) {
timeout = setTimeout(later, remaining);
}
return result;
};
};
Please note that this code can be simplified if you don't need all the options that underscore support.
Please find below a very simple and non-configurable version of this function:
function throttle (callback, limit) {
var waiting = false; // Initially, we're not waiting
return function () { // We return a throttled function
if (!waiting) { // If we're not waiting
callback.apply(this, arguments); // Execute users function
waiting = true; // Prevent future invocations
setTimeout(function () { // After a period of time
waiting = false; // And allow future invocations
}, limit);
}
}
}
Edit 1: Removed another reference to underscore, thx to #Zettam 's comment
Edit 2: Added suggestion about lodash and possible code simplification, thx to #lolzery #wowzery 's comment
Edit 3: Due to popular requests, I added a very simple, non-configurable version of the function, adapted from #vsync 's comment

What about this?
function throttle(func, timeFrame) {
var lastTime = 0;
return function () {
var now = Date.now();
if (now - lastTime >= timeFrame) {
func();
lastTime = now;
}
};
}
Simple.
You may be interested in having a look at the source.

callback: takes the function that should be called
limit: number of times that function should be called within the time limit
time: time span to reset the limit count
functionality and usage: Suppose you have an API that allows user to call it 10 times in 1 minute
function throttling(callback, limit, time) {
/// monitor the count
var calledCount = 0;
/// refresh the `calledCount` varialbe after the `time` has been passed
setInterval(function(){ calledCount = 0 }, time);
/// creating a closure that will be called
return function(){
/// checking the limit (if limit is exceeded then do not call the passed function
if (limit > calledCount) {
/// increase the count
calledCount++;
callback(); /// call the function
}
else console.log('not calling because the limit has exceeded');
};
}
////////////////////////////////////////////////////////////
// how to use
/// creating a function to pass in the throttling function
function cb(){
console.log("called");
}
/// calling the closure function in every 100 milliseconds
setInterval(throttling(cb, 3, 1000), 100);

Adding to the discussion here (and for more recent visitors), if the reason for not using the almost de facto throttle from lodash is to have a smaller sized package or bundle, then it's possible to include only throttle in your bundle instead of the entire lodash library. For example in ES6, it would be something like:
import throttle from 'lodash/throttle';
Also, there is a throttle only package from lodash called lodash.throttle which can be used with a simple import in ES6 or require in ES5.

I've just needed a throttle/debounce function for window resize event, and being curious, I also wanted to know what these are and how they work.
I've read multiple blog posts and QAs on SO, but they all seem to overcomplicate this, suggest libraries, or just provide descriptions and not simple plain JS implementations.
I won't provide a description since it's plentiful. So here's my implementation:
function throttle(callback, delay) {
var timeoutHandler = null;
return function () {
if (timeoutHandler == null) {
timeoutHandler = setTimeout(function () {
callback();
timeoutHandler = null;
}, delay);
}
}
}
function debounce(callback, delay) {
var timeoutHandler = null;
return function () {
clearTimeout(timeoutHandler);
timeoutHandler = setTimeout(function () {
callback();
}, delay);
}
}
These might need tweaks (e.g., initially the callback isn't called immediately).
See the difference in action (try resizing the window):
function throttle(callback, delay) {
var timeoutHandler = null;
return function () {
if (timeoutHandler == null) {
timeoutHandler = setTimeout(function () {
callback();
timeoutHandler = null;
}, delay);
}
}
}
function debounce(callback, delay) {
var timeoutHandler = null;
return function () {
clearTimeout(timeoutHandler);
timeoutHandler = setTimeout(function () {
callback();
}, delay);
}
}
var cellDefault = document.querySelector("#cellDefault div");
var cellThrottle = document.querySelector("#cellThrottle div");
var cellDebounce = document.querySelector("#cellDebounce div");
window.addEventListener("resize", function () {
var span = document.createElement("span");
span.innerText = window.innerWidth;
cellDefault.appendChild(span);
cellDefault.scrollTop = cellDefault.scrollHeight;
});
window.addEventListener("resize", throttle(function () {
var span = document.createElement("span");
span.innerText = window.innerWidth;
cellThrottle.appendChild(span);
cellThrottle.scrollTop = cellThrottle.scrollHeight;
}, 500));
window.addEventListener("resize", debounce(function () {
var span = document.createElement("span");
span.innerText = window.innerWidth;
cellDebounce.appendChild(span);
cellDebounce.scrollTop = cellDebounce.scrollHeight;
}, 500));
table {
border-collapse: collapse;
margin: 10px;
}
table td {
border: 1px solid silver;
padding: 5px;
}
table tr:last-child td div {
width: 60px;
height: 200px;
overflow: auto;
}
table tr:last-child td span {
display: block;
}
<table>
<tr>
<td>default</td>
<td>throttle</td>
<td>debounce</td>
</tr>
<tr>
<td id="cellDefault">
<div></div>
</td>
<td id="cellThrottle">
<div></div>
</td>
<td id="cellDebounce">
<div></div>
</td>
</tr>
</table>
JSFiddle

Here's how I implemented throttle function in ES6 in 9LOC, hope it helps
function throttle(func, delay) {
let timeout = null
return function(...args) {
if (!timeout) {
timeout = setTimeout(() => {
func.call(this, ...args)
timeout = null
}, delay)
}
}
}
Click on this link to see how it works.

I've seen a lot of answers here that are way too complex for "a simple throttle in js".
Almost all of the simpler answers just ignore calls made "in throttle" instead of delaying execution to the next interval.
Here's a simple implementation that also handles calls "in throttle":
const throttle = (func, limit) => {
let lastFunc;
let lastRan = Date.now() - (limit + 1); //enforces a negative value on first run
return function(...args) {
const context = this;
clearTimeout(lastFunc);
lastFunc = setTimeout(() => {
func.apply(context, args);
lastRan = Date.now();
}, limit - (Date.now() - lastRan)); //negative values execute immediately
}
}
This is almost the exact same implementation for a simple debounce. It just adds a calculation for the timeout delay which requires tracking when the function was last ran. See below:
const debounce = (func, limit) => {
let lastFunc;
return function(...args) {
const context = this;
clearTimeout(lastFunc);
lastFunc = setTimeout(() => {
func.apply(context, args)
}, limit); //no calc here, just use limit
}
}

Simple solution in ES6. Codepen Demo
const handleOnClick = () => {
console.log("hello")
}
const throttle = (func, delay) => {
let timeout = null;
return function (...args) {
if (timeout === null) {
func.apply(this, args);
timeout = setTimeout(() => {
timeout = null;
}, delay)
}
}
}
document.querySelector("#button").addEventListener("click", throttle(handleOnClick, 500))
<button type="button" id="button">Click me</button>

Here's my own version of Vikas post:
throttle: function (callback, limit, time) {
var calledCount = 0;
var timeout = null;
return function () {
if (limit > calledCount) {
calledCount++;
callback();
}
if (!timeout) {
timeout = setTimeout(function () {
calledCount = 0
timeout = null;
}, time);
}
};
}
I find that using setInterval is not a good idea.

With leading and trailing invocations:
const throttle = (fn, ms) => {
let locked = false
return function () {
if (!locked) {
locked = true
fn.apply(this, arguments)
setTimeout(() => {
fn.apply(this, arguments)
locked = false
}, ms)
}
}
}
Test case:
function log({ gender, address }) {
console.log({
name: this.name,
gender,
address,
})
}
const jack = {
name: 'Jack',
log: throttle(log, 3000),
}
Array.from({ length: 5 }, () => jack.log({ gender: 'Male', address: 'LA' }))

I made a npm package with some throttling functions:
npm install function-throttler
throttleAndQueue
Returns a version of your function that can be called at most every W milliseconds, where W is wait. Calls to your func that happen more often than W get queued up to be called every W ms
throttledUpdate
Returns a version of your function that can be called at most every W milliseconds, where W is wait. for calls that happen more often than W the last call will be the one called (last takes precedence)
throttle
limits your function to be called at most every W milliseconds, where W is wait. Calls over W get dropped

There is a library suited for this purpose, it's Backburner.js from Ember.
https://github.com/BackburnerJS/
You'd use it so.
var backburner = new Backburner(["task"]); //You need a name for your tasks
function saySomething(words) {
backburner.throttle("task", console.log.bind(console, words)
}, 1000);
}
function mainTask() {
"This will be said with a throttle of 1 second per word!".split(' ').map(saySomething);
}
backburner.run(mainTask)

This throttle function is build on ES6. Callback functions takes arguments (args), and still it works wrapped with throttle function. Be free to customize delay time according to your app needs. 1 time per 100ms is used for development mode, event "oninput" is just an example for frequent case of its use:
const callback = (...args) => {
console.count('callback throttled with arguments:', args);
};
throttle = (callback, limit) => {
let timeoutHandler = 'null'
return (...args) => {
if (timeoutHandler === 'null') {
timeoutHandler = setTimeout(() => {
callback(...args)
timeoutHandler = 'null'
}, limit)
}
}
}
window.addEventListener('oninput', throttle(callback, 100));
P.S. As #Anshul explained: throttling enforces a maximum number of times a function can be called over time. As in "execute this function at most once every 100 milliseconds."

In below example, try clicking the button multiple times, but the myFunc function would be executed only once in 3 sec.
The function throttle is passed with the function to be executed and the delay.It returns a closure, which is stored in obj.throttleFunc.
Now since obj.throttleFunc stores a closure, the value of isRunning is maintained inside it.
function throttle(func, delay) {
let isRunning;
return function(...args) {
let context = this; // store the context of the object that owns this function
if(!isRunning) {
isRunning = true;
func.apply(context,args) // execute the function with the context of the object that owns it
setTimeout(function() {
isRunning = false;
}, delay);
}
}
}
function myFunc(param) {
console.log(`Called ${this.name} at ${param}th second`);
}
let obj = {
name: "THROTTLED FUNCTION ",
throttleFunc: throttle(myFunc, 3000)
}
function handleClick() {
obj.throttleFunc(new Date().getSeconds());
}
button {
width: 100px;
height: 50px;
font-size: 20px;
}
<button onclick="handleClick()">Click me</button>
If we don't want the context or arguments to be passed, then a simpler
version of this would be as following:
function throttle(func, delay) {
let isRunning;
return function() {
if(!isRunning) {
isRunning = true;
func()
setTimeout(function() {
isRunning = false;
}, delay);
}
}
}
function myFunc() {
console.log('Called');
}
let throttleFunc = throttle(myFunc, 3000);
function handleClick() {
throttleFunc();
}
button {
width: 100px;
height: 50px;
font-size: 20px;
}
<button onclick="handleClick()">Click me</button>

I also want to suggest a simple solution for when there is only 1 function you know you will call (for example: Search)
here is what i did in my project
let throttle;
function search() {
if (throttle) {
clearTimeout(throttle);
}
throttle = setTimeout(() => {
sendSearchReq(str)
}, 500);
}
Search is called on input change event

function throttle(targetFunc, delay){
let lastFunc;
let lastTime;
return function(){
const _this = this;
const args = arguments;
if(!lastTime){
targetFunc.apply(_this, args);
lastTime = Date.now();
} else {
clearTimeout(lastFunc);
lastFunc = setTimeout(function(){
targetFunc.apply(_this, args);
lastTime = Date.now();
}, delay - (Date.now() - lastTime));
}
}
}
Try it :
window.addEventListener('resize', throttle(function() {
console.log('resize!!');
}, 200));

CodeSandbox
const { now } = Date;
export default function throttle(func, frameDuration) {
let timeout = null;
let latest;
const epoch = now();
function getDurationToNextFrame() {
const elapsed = now() - epoch;
const durationSinceLastFrame = elapsed % frameDuration;
return frameDuration - durationSinceLastFrame;
}
function throttled(...args) {
latest = () => {
func.apply(this, args);
};
if (!timeout) {
timeout = setTimeout(() => {
latest();
timeout = null;
}, getDurationToNextFrame());
}
}
return throttled;
}

Simple throttle function -
Note- Keep on clicking on the button , You'll see console log at first on click and then only after every 5 seconds until you're keep clicking.
HTML -
<button id='myid'>Click me</button>
Javascript -
const throttle = (fn, delay) => {
let lastTime = 0;
return (...args) => {
const currentTime = new Date().getTime();
if((currentTime - lastTime) < delay) {
return;
};
lastTime = currentTime;
return fn(...args);
}
};
document.getElementById('myid').addEventListener('click', throttle((e) => {
console.log('I am clicked');
}, 5000));

We can also implement using a flag-
var expensive = function(){
console.log("expensive functionnns");
}
window.addEventListener("resize", throttle(expensive, 500))
function throttle(expensiveFun, limit){
let flag = true;
return function(){
let context = this;
let args = arguments;
if(flag){
expensiveFun.apply(context, args);
flag = false;
setTimeout(function(){
flag = true;
}, limit);
}
}
}

Here is a bit modernized and simplified version of #clément-prévost answer
function throttle(func, wait, options = {}) {
let timeout = null;
let previous = 0;
const later = (...args) => {
previous = options.leading === false ? 0 : Date.now();
func(...args);
};
return (...args) => {
const now = Date.now();
if (!previous && options.leading === false) {
previous = now;
}
const remaining = wait - (now - previous);
if (remaining <= 0 || remaining > wait) {
if (timeout) {
clearTimeout(timeout);
timeout = null;
}
previous = now;
func(...args);
} else if (options.trailing !== false) {
clearTimeout(timeout);
timeout = setTimeout(() => later(...args), remaining);
}
};
}
function myFunc(a) {
console.log(`Log: ${a} ${this.val}`);
}
const myFuncThrottled = throttle(myFunc.bind({val: 42}), 1234, {leading: true, trailing: true})
myFuncThrottled(1)
myFuncThrottled(2)
myFuncThrottled(3)

function throttle(CB,ms=300,Id='Identifier for the callback(CB)'){
Id = Id || ""+CB
var N = throttle.N = throttle.N || {}; // Static variable N to store all callbacks ids and their status
if( N[Id] ) return; // already in the queue to run
N[Id] = 1; // add it the queue
setTimeout(()=>{
N[Id] = 0; // remove it from the queue
CB(); // finally call the function
}, ms);
}
for(var i=0;i<100;i++){
throttle(e=>console.log("Hi1"),1e3,'F1');
}
// will only output : Hi1
// this function guarantee the callback to run at least once

Some great solutions here already, but I was looking for a modern version with trailing (and optionally leading) executions, with the last passed arguments provided to each function call:
const throttle = (fn, wait=500, leading=true) => {
let prev, timeout, lastargs;
return (...args) => {
lastargs = args;
if (timeout) return;
timeout = setTimeout(() => {
timeout = null;
prev = Date.now();
// let's do this ... we'll release the stored args as we pass them through
fn.apply(this, lastargs.splice(0, lastargs.length));
// some fancy timing logic to allow leading / sub-offset waiting periods
}, leading ? prev && Math.max(0, wait - Date.now() + prev) || 0 : wait);
};
}
Usage:
x = throttle((...args) => console.log(...args));
let n = 0;
x(++n, 'boom');
x(++n, 'boom');
x(++n, 'boom');

if there will be more than one function defining them one by one would not be maintainable so i would suggest use a helper class to keep values for each
class slowDown {
constructor(cb,timeGap){
this.last = 0
this.run = function(){
let current = Date.now(),
shouldRun = (current - this.last) >= timeGap
if(shouldRun){
cb(current - this.last)
this.last = current
}
}
}
}
// example use
const press = new slowDown(timeElapsed => {
// define function here which you wanted to slow down
console.log("pressed after " + timeElapsed + " ms")
},750)
window.addEventListener("keydown",()=>{
press.run()
})

Below is the simplest throttle I could think of, in 13 LOC. It creates a timeout each time the function is called and cancels the old one. The original function is called with the proper context and arguments, as expected.
function throttle(fn, delay) {
var timeout = null;
return function throttledFn() {
window.clearTimeout(timeout);
var ctx = this;
var args = Array.prototype.slice.call(arguments);
timeout = window.setTimeout(function callThrottledFn() {
fn.apply(ctx, args);
}, delay);
}
}
// try it out!
window.addEventListener('resize', throttle(function() {
console.log('resize!!');
}, 200));

Related

Differences between these two throttle implementations

Approach one:
function throttle(func, timeMs) {
let lastTime = 0;
return function() {
const now = new Date();
if (now - lastTime >= timeMs) {
func();
lastTime = now;
}
};
}
Approach two:
function throttle(func, timeMs) {
let waiting = false;
return function() {
if (!waiting) {
func();
waiting = true;
setTimeout(() => {
waiting = false;
}, timeMs);
}
}
}
One uses the timestamp to keep track of the call while the other uses timeout. Is one better than the other?

How to kill and restart the recursive function in javascript

I am working on knockout js.
In that i have a recursive function which executes a function every minute. for that am using a timer every 60 sec it will execute also same will be reflecting in the UI also.
In my case, if i try to assign or initialize a timer value(observable) which is inside a loop, it doesn't reflecting instead of reflecting it is added to the pipeline and that much time loop is running simultaneously.
In that case i want to kill the loop and again want to restart every time i am changing the timer value.
timerInSec=60;
var loop = function () {
if (this.timer() < 1) {
myFunction()
this.timer(this.timerInSec - 1);
setTimeout(loop, 1000);
} else {
this.timer(this.timer() - 1);
setTimeout(loop, 1000);
}
};
loop();
Here is my solution. Please check.
timerInSec = 60;
const Loop = (function () {
let timer = 0;
let timerId = -1;
const myFunction = function () {
console.log('finished');
}
const fnLog = function (tm) {
console.log('current time = ', tm);
}
const fnProc = function () {
timerId = setTimeout(myFunction, 1000 * timer);
}
return {
start: function (tm = 60) {
this.stop();
timer = tm;
fnProc();
},
stop: function () {
if (timerId !== -1) {
clearTimeout(timerId);
timerId = -1;
}
}
}
})();
Loop.start(timerInSec);
setTimeout(() => {
Loop.start(timerInSec);
}, 500);

Implement your own setInterval and clearInterval methods without using setTimeout/clearTimeout

I was asked in a phone interview to replicate Javascript's setInterval, clearInterval methods by writing my own. I am not allowed to use setTimeout or clearTimeout.
function setInterval(func, wait){
var currentTime = Date.now();
while(currentTime < wait){
currentTime = Date.now() - currentTime;
}
if(currentTime >= wait) {
func();
}
}
function clearInterval(myVar){
myVar = undeclared;
}
function setTimeout(func, ms, ...args){
const start = Date.now();
(function check(){
if(start + ms >= Date.now()){
func(...args);
} else {
requestAnimationFrame(check);
}
})()
}
You might use a pseudorecursive function and requestAnimationFrame to do this. Based on this setTimeout implementation without window.setTimeout you cane easily implement a new setInterval ...
PS: in node you can use process.nectTick instead of requestAnimationFrame
Another possibility is, taking example from here, the usage of Web Workers:
let mySetInterval = function(callback, timeout) {
let blob = new Blob([ "self.addEventListener('message', function(e) { \
let old_date = Date.now(); \
while (Date.now() - old_date <= " + timeout + ") {}; \
self.postMessage(true); \
}, false);" ], { type: "text/javascript" });
let worker = new Worker(window.URL.createObjectURL(blob));
worker.addEventListener("message", function(e) {
if (callback() === false) {
return
}
mySetInterval(callback, timeout)
}, false);
worker.postMessage(true);
};
var a = 0;
mySetInterval(() => { if (a >= 10) { return false } else { console.log(a++) } }, 1000)
console.log(45);
Every 1 second it updates the variable a with +1.
Like you see, in this way it is non-blocking and it will stop when the variable a is 10.
To clear the "interval", in this case simply return false inside the callback. Obviously is not the same as the clearInterval function! Here there is not something like an ID such as for the setInterval function.
Hard Task, you can sort of do what setInterval and setTimeout do with requestAnimationFrame ? Not the same, but could do what you want to do.
var time;
function repeatOften() {
$("<div />").appendTo("body");
time = requestAnimationFrame(repeatOften);
}
$("#start").on("click", function() {
time = requestAnimationFrame(repeatOften);
});
$("#stop").on("click", function() {
cancelAnimationFrame(time);
});

setInterval(function(),time) change time on runtime

I want to change setInterval function time when my code is running.
I try this
<script type="text/javascript">
$(function () {
var timer;
function come() { alert("here"); }
timer = setInterval(come, 0);
clearInterval(timer);
timer = setInterval(come, 10000);
});
</script>
First SetInterval does not work!
You're clearing the interval on the next line, so the first one wont work, as it gets cleared right away :
timer = setInterval(come, 0);
clearInterval(timer);
timer = setInterval(come, 10000);
Also, as gdoron says, setting a interval of nothing isn't really valid, and not a really good idea either, use setTimeout instead, or just run the function outright if no delay is needed.
come();
clearInterval(timer);
timer = setInterval(come, 10000);
You can't. You will need to use setTimeout, and call it repetitively:
var timer; // current timeout id to clear
function come(){ /* do something */};
var time; // dynamic interval
(function repeat() {
come();
timer = setTimeout(repeat, time);
})();
With this you can set a different "interval" to be applied each time the function repeat is executed. Yet, nothing changes if alter time during a timeout, you'd need to stop the timeout for that.
I know this is an old post, but I have implemented a typescript version for changing the interval in run time:
class LoopTimer {
private timer: null | NodeJS.Timer;
private callback: (...args: any[]) => void;
private ms: number;
private started: boolean;
constructor(callback: (...args: any[]) => void, ms: number) {
this.callback = callback;
this.ms = ms;
this.timer = null;
this.started = false;
}
start() {
if (!this.started) {
this.timer = setInterval(this.callback, this.ms);
this.started = true;
}
}
stop() {
if (this.timer) {
clearInterval(this.timer);
this.timer = null;
this.started = false;
}
}
get getStarted(): boolean {
return this.started;
}
setInterval(ms: number) {
this.ms = ms;
if (this.started) {
this.stop();
this.start();
}
}
}
You can use it like this:
The timer will stop and start again when interval is changed.
const myTimer = new LoopTimer(()=>{
console.log("Hello");
}, 100);
myTimer.setInterval(500);
There is no way to directly change the interval at which a function fires. The best you can do is cancel an interval and set a new one with the same function and updated timer. Here's a possible way of doing it:
timer = {
timers:{},
inc:0,
start:function(cb,gap) {
var key = inc;
inc++;
timer.timers[key] = [setInterval(cb,gap),cb];
return key;
},
stop:function(id) {
if( !timer.timers[id]) return;
clearInterval(timer.timers[id][0]);
delete timer.timers[id];
},
change:function(id,newgap) {
if( !timer.timers[id]) return;
clearInterval(timer.timers[id][0]);
setInterval(timer.timers[id][1],newgap);
}
};
Usage:
var myTimer = timer.start(function() {....},1000);
// calls every second
timer.change(myTimer,5000);
// now calls every five seconds
timer = setInterval(come, 0); // zero isn't a valid interval...
You probably wants:
come();
timer = setInterval(come, 10000);
docs on MDN:
delay is the number of milliseconds (thousandths of a second) that the setInterval() function should wait before each call to func. As with setTimeout, there is a minimum delay enforced.
And:
Historically browsers implement setTimeout() "clamping": successive setTimeout() calls with delay smaller than the "minimum delay" limit are forced to the use at least the minimum delay. The minimum delay, DOM_MIN_TIMEOUT_VALUE, is 4 ms (stored in a preference in Firefox: dom.min_timeout_value), with a DOM_CLAMP_TIMEOUT_NESTING_LEVEL of 5ms.
this is mi example, i think is more simple and easy to understand
const timer = {
time: 5, // 5 time in seconds
_time: null,
_timeout: null,
fun: () => {},
start() {
if (this._timeout == null) {
const self = this;
this.fun();
this._timeout = setTimeout(function repeater() {
self.fun();
self._timeout = setTimeout(repeater, 1000 * self.time);
}, 1000 * this.time);
}
},
stop() {
const timeout = this._timeout;
this._timeout = null;
this.set_time(); // set time to default
clearTimeout(timeout);
},
set_time(time) {
if (this._time == null) this._time = this.time;
if (time) {
this.time = time;
} else {
this.time = this._time;
}
},
};
Explication:
time: is the time of interval between every iteration, cycle or next call
_time: this variable save the default value of time, when use stop(), this variable(_time) restore "time"
start: this function start the iteration, if you call again, it will not duplicate.
stop: this function, stop the timeout and set default time and _timeout
set_time: this function set a new value to time, if you not send a parameter, time restore to default value, declare on running in this example is "5"
example:
const timer = {
time: 5, // 5 time in seconds
_time: null,
_timeout: null,
fun: () => {},
start() {
if (this._timeout == null) {
const self = this;
this.fun();
this._timeout = setTimeout(function repeater() {
self.fun();
self._timeout = setTimeout(repeater, 1000 * self.time);
}, 1000 * this.time);
}
},
stop() {
const timeout = this._timeout;
this._timeout = null;
this.set_time(); // set time to default
clearTimeout(timeout);
},
set_time(time) {
if (this._time == null) this._time = this.time;
if (time) {
this.time = time;
} else {
this.time = this._time;
}
},
};
// print time
timer.fun = () =>{
console.log(new Date())
};
timer.set_time(10)
timer.start()
I know this post is old, but i needed something similar, maybe someone needs it.
This is a version without setInterval, based on the code from the other reaction (Niet the Dark Absol).
function timer()
{
var timer = {
running: false,
iv: 5000,
timeout: false,
cb : function(){},
start : function(cb,iv,sd){
var elm = this;
clearInterval(this.timeout);
this.running = true;
if(cb) this.cb = cb;
if(iv) this.iv = iv;
if(sd) elm.execute(elm);
this.timeout = setTimeout(function(){elm.execute(elm)}, this.iv);
},
execute : function(e){
if(!e.running) return false;
e.cb();
e.start();
},
stop : function(){
this.running = false;
},
set_interval : function(iv){
clearInterval(this.timeout);
this.start(false, iv);
}
};
return timer;
}
Usage:
var timer_1 = new timer();
timer_1.start(function(){
//magic here
}, 2000, false);
var timer_2 = new timer();
timer_2.start(function(){
//more magic here
}, 3000, true);
//change the interval
timer_2.set_interval(4000);
//stop the timer
timer_1.stop();
The last parameter of the start function is a boolean if the function needs to be run at 0.
You can also find the script here: https://github.com/Atticweb/smart-interval

javascript: pause setTimeout();

If I have an active timeout running that was set through
var t = setTimeout("dosomething()", 5000)
Is there anyway to pause and resume it?
Is there any way to get the time remaining on the current timeout?
or do I have to in a variable, when the timeout is set, store the current time, then we we pause, get the difference between now and then?
You could wrap window.setTimeout like this, which I think is similar to what you were suggesting in the question:
var Timer = function(callback, delay) {
var timerId, start, remaining = delay;
this.pause = function() {
window.clearTimeout(timerId);
timerId = null;
remaining -= Date.now() - start;
};
this.resume = function() {
if (timerId) {
return;
}
start = Date.now();
timerId = window.setTimeout(callback, remaining);
};
this.resume();
};
var timer = new Timer(function() {
alert("Done!");
}, 1000);
timer.pause();
// Do some stuff...
timer.resume();
Something like this should do the trick.
function Timer(fn, countdown) {
var ident, complete = false;
function _time_diff(date1, date2) {
return date2 ? date2 - date1 : new Date().getTime() - date1;
}
function cancel() {
clearTimeout(ident);
}
function pause() {
clearTimeout(ident);
total_time_run = _time_diff(start_time);
complete = total_time_run >= countdown;
}
function resume() {
ident = complete ? -1 : setTimeout(fn, countdown - total_time_run);
}
var start_time = new Date().getTime();
ident = setTimeout(fn, countdown);
return { cancel: cancel, pause: pause, resume: resume };
}
A slightly modified version of Tim Downs answer. However, since Tim rolled back my edit, I've to answer this myself. My solution makes it possible to use extra arguments as third (3, 4, 5...) parameter and to clear the timer:
function Timer(callback, delay) {
var args = arguments,
self = this,
timer, start;
this.clear = function () {
clearTimeout(timer);
};
this.pause = function () {
this.clear();
delay -= new Date() - start;
};
this.resume = function () {
start = new Date();
timer = setTimeout(function () {
callback.apply(self, Array.prototype.slice.call(args, 2, args.length));
}, delay);
};
this.resume();
}
As Tim mentioned, extra parameters are not available in IE lt 9, however I worked a bit around so that it will work in oldIE's too.
Usage: new Timer(Function, Number, arg1, arg2, arg3...)
function callback(foo, bar) {
console.log(foo); // "foo"
console.log(bar); // "bar"
}
var timer = new Timer(callback, 1000, "foo", "bar");
timer.pause();
document.onclick = timer.resume;
No. You'll need cancel it (clearTimeout), measure the time since you started it and restart it with the new time.
The Timeout was easy enough to find a solution for, but the Interval was a little bit trickier.
I came up with the following two classes to solve this issues:
function PauseableTimeout(func, delay){
this.func = func;
var _now = new Date().getTime();
this.triggerTime = _now + delay;
this.t = window.setTimeout(this.func,delay);
this.paused_timeLeft = 0;
this.getTimeLeft = function(){
var now = new Date();
return this.triggerTime - now;
}
this.pause = function(){
this.paused_timeLeft = this.getTimeLeft();
window.clearTimeout(this.t);
this.t = null;
}
this.resume = function(){
if (this.t == null){
this.t = window.setTimeout(this.func, this.paused_timeLeft);
}
}
this.clearTimeout = function(){ window.clearTimeout(this.t);}
}
function PauseableInterval(func, delay){
this.func = func;
this.delay = delay;
this.triggerSetAt = new Date().getTime();
this.triggerTime = this.triggerSetAt + this.delay;
this.i = window.setInterval(this.func, this.delay);
this.t_restart = null;
this.paused_timeLeft = 0;
this.getTimeLeft = function(){
var now = new Date();
return this.delay - ((now - this.triggerSetAt) % this.delay);
}
this.pause = function(){
this.paused_timeLeft = this.getTimeLeft();
window.clearInterval(this.i);
this.i = null;
}
this.restart = function(sender){
sender.i = window.setInterval(sender.func, sender.delay);
}
this.resume = function(){
if (this.i == null){
this.i = window.setTimeout(this.restart, this.paused_timeLeft, this);
}
}
this.clearInterval = function(){ window.clearInterval(this.i);}
}
These can be implemented as such:
var pt_hey = new PauseableTimeout(function(){
alert("hello");
}, 2000);
window.setTimeout(function(){
pt_hey.pause();
}, 1000);
window.setTimeout("pt_hey.start()", 2000);
This example will set a pauseable Timeout (pt_hey) which is scheduled to alert, "hey" after two seconds. Another Timeout pauses pt_hey after one second. A third Timeout resumes pt_hey after two seconds. pt_hey runs for one second, pauses for one second, then resumes running. pt_hey triggers after three seconds.
Now for the trickier intervals
var pi_hey = new PauseableInterval(function(){
console.log("hello world");
}, 2000);
window.setTimeout("pi_hey.pause()", 5000);
window.setTimeout("pi_hey.resume()", 6000);
This example sets a pauseable Interval (pi_hey) to write "hello world" in the console every two seconds. A timeout pauses pi_hey after five seconds. Another timeout resumes pi_hey after six seconds. So pi_hey will trigger twice, run for one second, pause for one second, run for one second, and then continue triggering every 2 seconds.
OTHER FUNCTIONS
clearTimeout() and clearInterval()
pt_hey.clearTimeout(); and pi_hey.clearInterval(); serve as an easy way to clear the timeouts and intervals.
getTimeLeft()
pt_hey.getTimeLeft(); and pi_hey.getTimeLeft(); will return how many milliseconds till the next trigger is scheduled to occur.
"Pause" and "resume" don't really make much sense in the context of setTimeout, which is a one-off thing. You might want to pause a chained series of setTimeout calls, in which case just don't schedule the next one (perhaps cancel the one that's outstanding via clearTimeout, as below). But setTimeout itself doesn't loop, there's nothing to pause and resume.
If you mean setInterval then no, you can't pause it, you can only cancel it (clearInterval) and then re-schedule it again. Details of all of these in the Timers section of the spec.
// Setting
var t = setInterval(doSomething, 1000);
// Pausing (which is really stopping)
clearInterval(t);
t = 0;
// Resuming (which is really just setting again)
t = setInterval(doSomething, 1000);
/revive
ES6 Version using Class-y syntactic sugar 💋
(slightly-modified: added start())
class Timer {
constructor(callback, delay) {
this.callback = callback
this.remainingTime = delay
this.startTime
this.timerId
}
pause() {
clearTimeout(this.timerId)
this.remainingTime -= new Date() - this.startTime
}
resume() {
this.startTime = new Date()
clearTimeout(this.timerId)
this.timerId = setTimeout(this.callback, this.remainingTime)
}
start() {
this.timerId = setTimeout(this.callback, this.remainingTime)
}
}
// supporting code
const pauseButton = document.getElementById('timer-pause')
const resumeButton = document.getElementById('timer-resume')
const startButton = document.getElementById('timer-start')
const timer = new Timer(() => {
console.log('called');
document.getElementById('change-me').classList.add('wow')
}, 3000)
pauseButton.addEventListener('click', timer.pause.bind(timer))
resumeButton.addEventListener('click', timer.resume.bind(timer))
startButton.addEventListener('click', timer.start.bind(timer))
<!doctype html>
<html>
<head>
<title>Traditional HTML Document. ZZz...</title>
<style type="text/css">
.wow { color: blue; font-family: Tahoma, sans-serif; font-size: 1em; }
</style>
</head>
<body>
<h1>DOM & JavaScript</h1>
<div id="change-me">I'm going to repaint my life, wait and see.</div>
<button id="timer-start">Start!</button>
<button id="timer-pause">Pause!</button>
<button id="timer-resume">Resume!</button>
</body>
</html>
I needed to calculate the elapsed and remaining time to show a progress-bar. It was not easy using the accepted answer. 'setInterval' is better than 'setTimeout' for this task. So, I created this Timer class that you can use in any project.
https://jsfiddle.net/ashraffayad/t0mmv853/
'use strict';
//Constructor
var Timer = function(cb, delay) {
this.cb = cb;
this.delay = delay;
this.elapsed = 0;
this.remaining = this.delay - self.elapsed;
};
console.log(Timer);
Timer.prototype = function() {
var _start = function(x, y) {
var self = this;
if (self.elapsed < self.delay) {
clearInterval(self.interval);
self.interval = setInterval(function() {
self.elapsed += 50;
self.remaining = self.delay - self.elapsed;
console.log('elapsed: ' + self.elapsed,
'remaining: ' + self.remaining,
'delay: ' + self.delay);
if (self.elapsed >= self.delay) {
clearInterval(self.interval);
self.cb();
}
}, 50);
}
},
_pause = function() {
var self = this;
clearInterval(self.interval);
},
_restart = function() {
var self = this;
self.elapsed = 0;
console.log(self);
clearInterval(self.interval);
self.start();
};
//public member definitions
return {
start: _start,
pause: _pause,
restart: _restart
};
}();
// - - - - - - - - how to use this class
var restartBtn = document.getElementById('restart');
var pauseBtn = document.getElementById('pause');
var startBtn = document.getElementById('start');
var timer = new Timer(function() {
console.log('Done!');
}, 2000);
restartBtn.addEventListener('click', function(e) {
timer.restart();
});
pauseBtn.addEventListener('click', function(e) {
timer.pause();
});
startBtn.addEventListener('click', function(e) {
timer.start();
});
Typescript implementation based on top rated answer
/** Represents the `setTimeout` with an ability to perform pause/resume actions */
export class Timer {
private _start: Date;
private _remaining: number;
private _durationTimeoutId?: NodeJS.Timeout;
private _callback: (...args: any[]) => void;
private _done = false;
get done () {
return this._done;
}
constructor(callback: (...args: any[]) => void, ms = 0) {
this._callback = () => {
callback();
this._done = true;
};
this._remaining = ms;
this.resume();
}
/** pauses the timer */
pause(): Timer {
if (this._durationTimeoutId && !this._done) {
this._clearTimeoutRef();
this._remaining -= new Date().getTime() - this._start.getTime();
}
return this;
}
/** resumes the timer */
resume(): Timer {
if (!this._durationTimeoutId && !this._done) {
this._start = new Date;
this._durationTimeoutId = setTimeout(this._callback, this._remaining);
}
return this;
}
/**
* clears the timeout and marks it as done.
*
* After called, the timeout will not resume
*/
clearTimeout() {
this._clearTimeoutRef();
this._done = true;
}
private _clearTimeoutRef() {
if (this._durationTimeoutId) {
clearTimeout(this._durationTimeoutId);
this._durationTimeoutId = undefined;
}
}
}
You could look into clearTimeout()
or pause depending on a global variable that is set when a certain condition is hit. Like a button is pressed.
<button onclick="myBool = true" > pauseTimeout </button>
<script>
var myBool = false;
var t = setTimeout(function() {if (!mybool) {dosomething()}}, 5000);
</script>
You could also implement it with events.
Instead of calculating the time difference, you start and stop listening to a 'tick' event which keeps running in the background:
var Slideshow = {
_create: function(){
this.timer = window.setInterval(function(){
$(window).trigger('timer:tick'); }, 8000);
},
play: function(){
$(window).bind('timer:tick', function(){
// stuff
});
},
pause: function(){
$(window).unbind('timer:tick');
}
};
If you're using jquery anyhow, check out the $.doTimeout plugin. This thing is a huge improvement over setTimeout, including letting you keep track of your time-outs with a single string id that you specify and that doesn't change every time you set it, and implement easy canceling, polling loops & debouncing, and more. One of my most-used jquery plugins.
Unfortunately, it doesn't support pause/resume out of the box. For this, you would need to wrap or extend $.doTimeout, presumably similarly to the accepted answer.
I needed to be able to pause setTimeout() for slideshow-like feature.
Here is my own implementation of a pausable timer. It integrates comments seen on Tim Down's answer, such as better pause (kernel's comment) and a form of prototyping (Umur Gedik's comment.)
function Timer( callback, delay ) {
/** Get access to this object by value **/
var self = this;
/********************* PROPERTIES *********************/
this.delay = delay;
this.callback = callback;
this.starttime;// = ;
this.timerID = null;
/********************* METHODS *********************/
/**
* Pause
*/
this.pause = function() {
/** If the timer has already been paused, return **/
if ( self.timerID == null ) {
console.log( 'Timer has been paused already.' );
return;
}
/** Pause the timer **/
window.clearTimeout( self.timerID );
self.timerID = null; // this is how we keep track of the timer having beem cleared
/** Calculate the new delay for when we'll resume **/
self.delay = self.starttime + self.delay - new Date().getTime();
console.log( 'Paused the timer. Time left:', self.delay );
}
/**
* Resume
*/
this.resume = function() {
self.starttime = new Date().getTime();
self.timerID = window.setTimeout( self.callback, self.delay );
console.log( 'Resuming the timer. Time left:', self.delay );
}
/********************* CONSTRUCTOR METHOD *********************/
/**
* Private constructor
* Not a language construct.
* Mind var to keep the function private and () to execute it right away.
*/
var __construct = function() {
self.starttime = new Date().getTime();
self.timerID = window.setTimeout( self.callback, self.delay )
}(); /* END __construct */
} /* END Timer */
Example:
var timer = new Timer( function(){ console.log( 'hey! this is a timer!' ); }, 10000 );
timer.pause();
To test the code out, use timer.resume() and timer.pause() a few times and check how much time is left. (Make sure your console is open.)
Using this object in place of setTimeout() is as easy as replacing timerID = setTimeout( mycallback, 1000) with timer = new Timer( mycallback, 1000 ). Then timer.pause() and timer.resume() are available to you.
function delay (ms) { return new Promise(resolve => setTimeout(resolve, s)); }
"async" working demo at:
site zarsoft.info
You can do like below to make setTimeout pausable on server side (Node.js)
const PauseableTimeout = function(callback, delay) {
var timerId, start, remaining = delay;
this.pause = function() {
global.clearTimeout(timerId);
remaining -= Date.now() - start;
};
this.resume = function() {
start = Date.now();
global.clearTimeout(timerId);
timerId = global.setTimeout(callback, remaining);
};
this.resume();
};
and you can check it as below
var timer = new PauseableTimeout(function() {
console.log("Done!");
}, 3000);
setTimeout(()=>{
timer.pause();
console.log("setTimeout paused");
},1000);
setTimeout(()=>{
console.log("setTimeout time complete");
},3000)
setTimeout(()=>{
timer.resume();
console.log("setTimeout resume again");
},5000)
class pausable_timeout {
constructor(func, milliseconds) {
this.func = func;
this.date_ms = new Date().valueOf();
this.timeout = setTimeout(func, milliseconds);
this.time_left = milliseconds;
};
pause() {
clearTimeout(this.timeout);
const elapsed_time = new Date().valueOf() - this.date_ms;
this.time_left -= elapsed_time;
};
unpause() {
this.timeout = setTimeout(this.func, this.time_left);
this.date_ms = new Date().valueOf();
};
};
const timer = new pausable_timeout(() => /* your code */, 3000 /* your timeout in milliseconds */);
timer.pause();
timer.unpause();
The programme is rather simple. We will create a class containing two functions, the pause function and the unpause function.
The pause function will clear the setTimeout and store the time that has elapsed between the start and now in the time_left variable. The unpause function will recreate a setTimeout by putting the time_left time as an argument.
If anyone wants the TypeScript version shared by the Honorable #SeanVieira here, you can use this:
public timer(fn: (...args: any[]) => void, countdown: number): { onCancel: () => void, onPause: () => void, onResume: () => void } {
let ident: NodeJS.Timeout | number;
let complete = false;
let totalTimeRun: number;
const onTimeDiff = (date1: number, date2: number) => {
return date2 ? date2 - date1 : new Date().getTime() - date1;
};
const handlers = {
onCancel: () => {
clearTimeout(ident as NodeJS.Timeout);
},
onPause: () => {
clearTimeout(ident as NodeJS.Timeout);
totalTimeRun = onTimeDiff(startTime, null);
complete = totalTimeRun >= countdown;
},
onResume: () => {
ident = complete ? -1 : setTimeout(fn, countdown - totalTimeRun);
}
};
const startTime = new Date().getTime();
ident = setTimeout(fn, countdown);
return handlers;
}
I created this code in TypeScript for slider feature:
class TimeoutSlider {
private callback: () => void;
private duration: number;
private timeReaming: number;
private startTime: number | null = null;
private timerId: NodeJS.Timeout | null = null;
constructor(callback: () => void, duration: number) {
this.callback = callback;
this.duration = duration;
this.timeReaming = duration;
}
public start() {
this.clear();
this.startTime = new Date().getTime();
this.timerId = setTimeout(this.callback, this.duration);
}
public pause() {
if (!this.startTime) {
throw new Error("Cannot pause a timer that has not been started");
}
this.clear();
this.timeReaming = this.duration - (new Date().getTime() - this.startTime);
}
public resume() {
this.clear();
this.startTime = new Date().getTime();
this.timerId = setTimeout(this.callback, this.timeReaming);
}
private clear() {
if (this.timerId) {
clearTimeout(this.timerId);
this.timerId = null;
}
}
}
I don't think you'll find anything better than clearTimeout. Anyway, you can always schedule another timeout later, instead 'resuming' it.
If you have several divs to hide, you could use an setInterval and a number of cycles to do like in:
<div id="div1">1</div><div id="div2">2</div>
<div id="div3">3</div><div id="div4">4</div>
<script>
function hideDiv(elm){
var interval,
unit = 1000,
cycle = 5,
hide = function(){
interval = setInterval(function(){
if(--cycle === 0){
elm.style.display = 'none';
clearInterval(interval);
}
elm.setAttribute('data-cycle', cycle);
elm.innerHTML += '*';
}, unit);
};
elm.onmouseover = function(){
clearInterval(interval);
};
elm.onmouseout = function(){
hide();
};
hide();
}
function hideDivs(ids){
var id;
while(id = ids.pop()){
hideDiv(document.getElementById(id));
}
}
hideDivs(['div1','div2','div3','div4']);
</script>

Categories