How do I pass a function as a parameter without the function executing in the "parent" function or using eval()? (Since I've read that it's insecure.)
I have this:
addContact(entityId, refreshContactList());
It works, but the problem is that refreshContactList fires when the function is called, rather than when it's used in the function.
I could get around it using eval(), but it's not the best practice, according to what I've read. How can I pass a function as a parameter in JavaScript?
You just need to remove the parenthesis:
addContact(entityId, refreshContactList);
This then passes the function without executing it first.
Here is an example:
function addContact(id, refreshCallback) {
refreshCallback();
// You can also pass arguments if you need to
// refreshCallback(id);
}
function refreshContactList() {
alert('Hello World');
}
addContact(1, refreshContactList);
If you want to pass a function, just reference it by name without the parentheses:
function foo(x) {
alert(x);
}
function bar(func) {
func("Hello World!");
}
//alerts "Hello World!"
bar(foo);
But sometimes you might want to pass a function with arguments included, but not have it called until the callback is invoked. To do this, when calling it, just wrap it in an anonymous function, like this:
function foo(x) {
alert(x);
}
function bar(func) {
func();
}
//alerts "Hello World!" (from within bar AFTER being passed)
bar(function(){ foo("Hello World!") });
If you prefer, you could also use the apply function and have a third parameter that is an array of the arguments, like such:
function eat(food1, food2) {
alert("I like to eat " + food1 + " and " + food2 );
}
function myFunc(callback, args) {
//do stuff
//...
//execute callback when finished
callback.apply(this, args);
}
//alerts "I like to eat pickles and peanut butter"
myFunc(eat, ["pickles", "peanut butter"]);
Example 1:
funct("z", function (x) { return x; });
function funct(a, foo){
foo(a) // this will return a
}
Example 2:
function foodemo(value){
return 'hello '+value;
}
function funct(a, foo){
alert(foo(a));
}
//call funct
funct('world!',foodemo); //=> 'hello world!'
look at this
To pass the function as parameter, simply remove the brackets!
function ToBeCalled(){
alert("I was called");
}
function iNeedParameter( paramFunc) {
//it is a good idea to check if the parameter is actually not null
//and that it is a function
if (paramFunc && (typeof paramFunc == "function")) {
paramFunc();
}
}
//this calls iNeedParameter and sends the other function to it
iNeedParameter(ToBeCalled);
The idea behind this is that a function is quite similar to a variable. Instead of writing
function ToBeCalled() { /* something */ }
you might as well write
var ToBeCalledVariable = function () { /* something */ }
There are minor differences between the two, but anyway - both of them are valid ways to define a function.
Now, if you define a function and explicitly assign it to a variable, it seems quite logical, that you can pass it as parameter to another function, and you don't need brackets:
anotherFunction(ToBeCalledVariable);
There is a phrase amongst JavaScript programmers: "Eval is Evil" so try to avoid it at all costs!
In addition to Steve Fenton's answer, you can also pass functions directly.
function addContact(entity, refreshFn) {
refreshFn();
}
function callAddContact() {
addContact("entity", function() { DoThis(); });
}
I chopped all my hair off with that issue. I couldn't make the examples above working, so I ended like :
function foo(blabla){
var func = new Function(blabla);
func();
}
// to call it, I just pass the js function I wanted as a string in the new one...
foo("alert('test')");
And that's working like a charm ... for what I needed at least. Hope it might help some.
I suggest to put the parameters in an array, and then split them up using the .apply() function. So now we can easily pass a function with lots of parameters and execute it in a simple way.
function addContact(parameters, refreshCallback) {
refreshCallback.apply(this, parameters);
}
function refreshContactList(int, int, string) {
alert(int + int);
console.log(string);
}
addContact([1,2,"str"], refreshContactList); //parameters should be putted in an array
You can also use eval() to do the same thing.
//A function to call
function needToBeCalled(p1, p2)
{
alert(p1+"="+p2);
}
//A function where needToBeCalled passed as an argument with necessary params
//Here params is comma separated string
function callAnotherFunction(aFunction, params)
{
eval(aFunction + "("+params+")");
}
//A function Call
callAnotherFunction("needToBeCalled", "10,20");
That's it. I was also looking for this solution and tried solutions provided in other answers but finally got it work from above example.
Here it's another approach :
function a(first,second)
{
return (second)(first);
}
a('Hello',function(e){alert(e+ ' world!');}); //=> Hello world
In fact, seems like a bit complicated, is not.
get method as a parameter:
function JS_method(_callBack) {
_callBack("called");
}
You can give as a parameter method:
JS_method(function (d) {
//Finally this will work.
alert(d)
});
The other answers do an excellent job describing what's going on, but one important "gotcha" is to make sure that whatever you pass through is indeed a reference to a function.
For instance, if you pass through a string instead of a function you'll get an error:
function function1(my_function_parameter){
my_function_parameter();
}
function function2(){
alert('Hello world');
}
function1(function2); //This will work
function1("function2"); //This breaks!
See JsFiddle
Some time when you need to deal with event handler so need to pass event too as an argument , most of the modern library like react, angular might need this.
I need to override OnSubmit function(function from third party library) with some custom validation on reactjs and I passed the function and event both like below
ORIGINALLY
<button className="img-submit" type="button" onClick=
{onSubmit}>Upload Image</button>
MADE A NEW FUNCTION upload and called passed onSubmit and event as arguments
<button className="img-submit" type="button" onClick={this.upload.bind(this,event,onSubmit)}>Upload Image</button>
upload(event,fn){
//custom codes are done here
fn(event);
}
By using ES6:
const invoke = (callback) => {
callback()
}
invoke(()=>{
console.log("Hello World");
})
If you can pass your whole function as string, this code may help you.
convertToFunc( "runThis('Micheal')" )
function convertToFunc( str) {
new Function( str )()
}
function runThis( name ){
console.log("Hello", name) // prints Hello Micheal
}
You can use a JSON as well to store and send JS functions.
Check the following:
var myJSON =
{
"myFunc1" : function (){
alert("a");
},
"myFunc2" : function (functionParameter){
functionParameter();
}
}
function main(){
myJSON.myFunc2(myJSON.myFunc1);
}
This will print 'a'.
The following has the same effect with the above:
var myFunc1 = function (){
alert('a');
}
var myFunc2 = function (functionParameter){
functionParameter();
}
function main(){
myFunc2(myFunc1);
}
Which is also has the same effect with the following:
function myFunc1(){
alert('a');
}
function myFunc2 (functionParameter){
functionParameter();
}
function main(){
myFunc2(myFunc1);
}
And a object paradigm using Class as object prototype:
function Class(){
this.myFunc1 = function(msg){
alert(msg);
}
this.myFunc2 = function(callBackParameter){
callBackParameter('message');
}
}
function main(){
var myClass = new Class();
myClass.myFunc2(myClass.myFunc1);
}
I have some ambiguity in Javascript callback functions.
The first code is structured as follows:
function firstFunction()
{
var message = "something";
secondFunction(message);
}
function secondFunction(message)
{
var myButton = document.getElementById("my-button");
myButton.addEventListener('click',thirdFunction(message));
}
function thirdFunction(message)
{
console.log("the messages is: "+message);
}
When I run the script above, the thirdFunction gets executed without clicking the button.
After some research, I read about the closure in Javascript. Then I changed the code to the following structure:
function firstFunction()
{
var message = "something";
secondFunction(message);
}
function secondFunction(message)
{
var myButton = document.getElementById("my-button");
myButton.addEventListener('click',thirdFunction);
}
function thirdFunction(message)
{
return function(){
console.log("the messages is: "+message);
}
}
I got the expected result. The thirdFunction is executed only when the button is clicked.
I am not sure if I my second code structure is correct? I am not sure if I'm getting the closure concept correctly as I never returned a function in conventional programming before. This is a new concept to me. Please, correct me if I'm wrong.
EDIT:
Some of the solutions suggest writing it like this:
myButton.addEventListener('click', function() { thirdFunction(message) });
For code readability, I am trying to avoid this. I prefer to place the code for the thirdFunction outside the secondFunction.
Use an anonymous function to make the closure in the correct environment:
function secondFunction(message)
{
var myButton = document.getElementById("my-button");
myButton.addEventListener('click', function() {
thirdFunction(message)
});
}
If I want to break a callback implementation out of a method's parameter footprint for cleaner code I can do (for example)
foo.bar(a, callback(), b);
function callback() {
stuff;
}
instead of
foo.bar(a, function() {
stuff;
}, b);
But what do I do if the method passes something into the callback like three.js's loader functions? (http://threejs.org/docs/#Reference/Loaders/OBJMTLLoader)
foo.bar(callback(object));
function callback(object) {
object.stuff();
}
doesn't seem to work.
Got it. The format should be:
foo.bar(callback);
function callback(object) {
object.stuff();
}
The 2 snippets you've posted are actually different - when you pass an anonymous function as an argument it isn't run immediately, but in the "foo.bar" function itself. However, when you pass it as "callback()", that function runs immediately (it is useful in some cases, for example: if the function returns another function). So, pass it without "()".
For example:
function a(){
alert(1);
}
function b(callback){
callback(); //run the callback function.
}
b(a);
And, if you want to see an example of the second option:
function a(){
return function() {
alert(1);
};
}
a()(); //Alert
b(a()); //Alert
b(a) //nothing happens
Preety straight forward question, though I can't find the answer anywhere
I tried these two ways:
setInterval(function(){object/*or this*/.method()},500)
and
setInterval('object/*or this*/.method()',500)
setInterval in fact expects a method as the first argument, though there is an alternative syntax where the first argument can be a string of code (not recommended by most)
If you're having issues with that code, it may have to do with the scope of 'this'
setInterval(function(){this.method()},500)
In the above code, 'this' will refer to the closure itself, and wouldn't be the same as 'this.method' occurring outside of that closure. For example, the following would work:
function MyClass() {
this.thingy = 'yep this is a thingy'
}
var myClass = new MyClass()
// Will log 'MyClass yep this is a thingy'
setInterval(function() { console.log('MyClass', myClass.thingy) }, 1000)
Whereas the following will not work (presuming instantiating the object and calling foo()):
function MyOtherClass() {
this.thingy = 'also a thingy'
}
// Will log 'MyOtherClass undefined'
MyOtherClass.prototype.foo = function() {
setInterval(function() { console.log('MyOtherClass', this.thingy) }, 1000)
}
The second example will work if we get around using 'this' within the closure (presuming instantiating the object and calling bar()):
MyOtherClass.prototype.bar = function() {
var that = this
setInterval(function() { console.log('MyOtherClass', that.thingy) }, 1000)
}
Also be sure that setInterval is being passed the name of a function:
setInterval(someFunction, 500)
rather than executing a function as an argument
setInterval(someFunction(), 500)
This last line of code is usually a mistake, unless someFunction() returns a function itself ;)
The difference between your 2 ways for passing a function to setInterval is whether you want to pass your function as refrence of just copy of it. Allow me to explain it by example:
-1 Referring(demo):
var obj = {
testMethod: function () {
console.log('function (testMethod): intial output');
}
}
setInterval(function () {
obj.testMethod()
}, 1000);
obj.testMethod = function () {
console.log('function (testMethod): changed output');
}
when you run this code, the result 'll be execution of the modified version of testMethod. Because here you dont copy the function! Instead, you refer to it. So whenever function implementation is changed, the last modified version is executed.
-2 Copying(demo):
var obj = {
testMethod: function () {
console.log('function (testMethod): intial output');
}
}
setInterval(obj.testMethod, 1000);
obj.testMethod = function () {
console.log('function (testMethod): changed output');
}
Here all you do is you are passing a copy of the last defined version of the function testMethod to setInterval. So whatever changes you do to testMethod, the result of setInterval will not be changed.
Just wondering if there is anyway to fire some code when a function is called, without adding the code to the function, for example:
function doSomething(){
//Do something
}
//Code to call when doSomething is called
You can wrap the function :
(function(){
var oldFunction = doSomething;
doSomething = function(){
// do something else
oldFunction.apply(this, arguments);
}
})();
I use an IIFE here just to avoid polluting the global namespace, it's accessory.
Well, yes, it's not actually hard to do. The crucial thing is that a function's name is just an identifier like any other. You can redefine it if you want to.
var oldFn = doSomething;
doSomething = function() {
// code to run before the old function
return oldFn.apply(this, arguments);
// code to run after the old function
};
NB that it's better to do oldFn.apply(this, arguments) rather than just oldFn. In many cases it won't matter, but it's possible that the context (i.e. the value of this inside the function) and the arguments are important. Using apply means they are passed on as if oldFn had been called directly.
What about something like:
function doSomething(){
doSomething.called = true;
}
//call?
doSomething();
if(doSomething.called) {
//Code to call when doSomething is called
}
I know you said you don't want to modify the original function, but consider adding a callback. Then you can execute code based on different results in your function (such as onSucess and onError):
function doSomething(onSuccess, onError){
try {
throw "this is an error";
if(onSuccess) {
onSuccess();
}
} catch(err) {
if(onError) {
onError(err);
}
}
}
Then, when you call doSomething, you can specify what you want done with inline functions:
doSomething(function() {
console.log("doSomething() success");
}, function(err) {
console.log("doSomething() error: " + err);
});