Javascript closures -- what is the difference between these - javascript

EDIT
With the number of responses saying "you can make private things!" below, I'm going to add this to the top as well:
I know you can emulate private variables within a closure. That is not what I'm asking. I'm asking, given the two examples below where I'm "exporting" EVERYTHING from the closure, what is the fundamental difference between these two examples.
Given these two methods of creating objects/methods:
var test = {}
test = (function(){
var a_method = function(print_me){
return "hello "+print_me;
}
return {print_me: a_method};
})();
test.print_me2 = function(print_me2){
return "hello "+print_me2;
}
test.print_me('world');
>>> returns "hello world"
test.print_me2('world');
>>> returns "hello world"
I understand that the first method allows for private variables (which as a python developer at heart i don't really care to use), but both seem rather equivilent to me, only the first one looks "cooler" (as in all the big javascript people seem to be doing it that way) and the second way looks very passé.
So, like, what is the difference?
I've looked through the closure questions here -- most of them center around what or why do you use them; I understand their utility, I just want to know why you'd do the first over the second and what benefits it has.
I'd prefer hard evidence over conjecture -- not looking for a "this is how the cool kids are doing it" or "i heard that mozilla does better memory usage when you use a closure", but rather qualitative evidence towards one being 'better' than the other.

The difference between the methods is the anonymous function wrapper that creates a closure, but also that the first method replaces the entire object while the second method just sets a property in the existing method.
There are different ways of adding methods to an object. You can put them there when you create the object:
var test = {
print_me: function(text) { return "hello " + text; }
};
You can add a method as a property to an existing object:
var test = {};
test.print_me = function(text) { return "hello " + text; };
You can make a constructor function, and add methods to its prototype:
function Test() {}
Test.prototype.print_me = function(text) { return "hello " + text; };
var test = new Test();
As Javascript has a prototype based object model, the last one is how methods were intended to be created. The other ways are just possible because an object property can have a function as value.
You can use a function wrapper around any code where you want local variables, so you can do that with the second way of setting the property:
test.print_me2 = (function(){
var method = function(print_me2) {
return "hello "+print_me2;
}
return method;
})();

In the first example, the closure is used to keep local variables out of the global scope.
var thisIsGlobal = 2;
var test = (function () {
var thisIsLocal = 3;
return function () {
return thisIsLocal;
};
}());
What you get is a function test which when invoked returns the value of the local variable thisIsLocal. There is no way to change the value of the variable. You can be look at it as to a private variable.

The first method you present (an anonymous function which is executed immediately) keeps your program out of the global scope. Within that private scope you can decide what properties you want to expose as your API:
(function(win){
var a, b, c = 1;
function init(){
return aa(b);
}
function exit(){
if(cc()){
return a;
}
}
function aa(){
}
function bb(){
}
function cc(){
}
win.myProgram = {init: init,
exit: exit };
})(window);
// We now have access to your API from within the global namespace:
myProgram.init();
//etc

Related

Why Self-Executing Anonymous Functions

Today I came a cross the self executing functions, than somehow I ended up knowing about
Self-Executing Anonymous Functions, then I've read this article: http://briancrescimanno.com/how-self-executing-anonymous-functions-work/
The thing is that I don't know WHY to use Self-Executing Anonymous Functions because if I need to do something like:
var test = "a";
(function(foo) {
alert(foo);
})(test);
I could just make something like:
var test = "a";
alert(foo);
Or did I miss anything?
also this can be done to any code inside the function, but I used alert() to make simple
Update:
Even thought I've already accepted and answer I would like to share something I've found, if anyone came across this question later :)
Using this notation we can also make an endless loop like following:
(function loop(){
// do something here
loop();
)();
There are a couple of reasons why one would use an IIFE:
1) No littering
var a = 'foo';
alert(a);
vs
(function() {
var a = 'foo';
alert(a);
}())
Both examples do the same thing, but in the second example there is no a variable inside the outer scope.
2) State capturing
var a = 'foo';
window.setTimeout(function() { alert(a); }, 1);
a = 'bar';
vs
var a = 'foo';
window.setTimeout( (function(a_copy) {
return function() { alert(a_copy); }
}(a)), 1);
a = 'bar';
The first example alerts bar, while the second alerts foo. You will find this technique used especially with loops.
Your initial example isn't worth to be executed in an anonymous function, so its a bad example to understand WHY to use this technique. Here is a good example to explore state capturing:
var list = [{id: 1, data: null}, ...];
for (var i = 0; i < list.length; i++) {
(function(item) {
// start async request, for each item in the list, in the order of the list
asyncAjax("get-data-from-somewhere.json?what=" + list[i].id, function (response) {
// thats the on success callback, which gets executed when the server responded
// each item will have different response times, therefore the order of response is various
// if we would use simply list[i].data, it wouldn't work correctly, because "i" would has the value of list.length, because the iteration has been finished yet.
item.data = response;
});
})(list[i]); // the function will preserve the reference to the list item inside, until the function has been fully executed
}
When writing sync. code, you can always fallback to classic object oriented style of structering your code. So you can avoid closures / instant-anonymous function calls. But as soon as you use async. mechanics, closures get very handy and make your code looking more clean, off course only if you can read and understand closures :)
By the way, you could also simply write:
function(private) {}(outer)
is the same as
(function(private) {})(outer)
but the second is better, because its simply more obvious for the reader.
The syntax you describe is commonly referred to as an "immediately invoked function expression", or IIFE.
One of the common use cases is to emulate private variables:
var ns = (function () {
var x = 1; // "private"
return {
getX: function () {
return x;
}
}
}());
ns.getX(); // 1
ns.x; // undefined because x is "private"
In that example the x variable is local to the IIFE. It's not directly accessible outside of it. However, since it is referred to by the getX method, which is accessible outside of the IIFE (because it's part of the returned object) a reference to x is kept alive. This is what is usually meant by the term "closure".
Self executing functions are not really useful if you just do an alert inside.
Consider something like this:
(function(foo) {
var a = ..
// do something with a and foo
})(test);
The advantage here is that a is "private" inside the method and cannot be used outside the method. So a doesn't end up as a global variable and can't be overwritten by some other piece of javascript which uses a variable of the same name.

Putting arrays into custom objects

I am attempting to build my first custom object and it looks something like this:
function URLObject()
{
this.syllables = new Array();
etc...
this.AddtoSyllables = AddtoSyllables;
function AddtoSyllables(AWord)
{
var SylCount = this.syllables.length;
alert("This is SylCount: " + SylCount);
}
}
var myobj = new URLObject();
myobj.AdtoSyllables("text");
The execution of the above code results in the JS engine printing out the following:
This is Sylcount: NAN
-or-
This is SylCount: undefined.
I have looked at information in Head First Javascript, in the Javascript bible, and on various JS websites. All of them cover in exhaustive detail the use of arrays of objects, but none of them discuss arrays within objects.
And yet I am doing something wrong here and I do not know what. Can anyone help?
Here you go:
function URLObject()
{
this.syllables = [];
etc...
}
URLObject.prototype.addToSyllables = function(aWord) {
var SylCount = this.syllables.length;
alert("This is SylCount: " + SylCount);
}
var myobj = new URLObject();
myobj.adtoSyllables("text");
.prototype adds the function declared after it to every object constructed by the constructor function. (in your case every object that was instantiated by new URLObject())
Firstly, the code as posted actually works for me on Chrome and Firefox; so this must depend on the JavaScript engine, or else there's something funky going on.
Update: I suspect what may be confusing you is some separate call to AddtoSyllables (in code you haven't shown us) where suddenly this.syllables is no longer defined. This is where the behavior of this can get confusing. I've created a jsFiddle to hopefully explain how it works a bit better for you.
http://jsfiddle.net/J3tUb/
That said, it is often very possible to write code like this without having to use this (or the prototype) at all. For instance:
function createURLObject() {
// Use closed-over locals instead of attaching properties.
var syllables = new Array();
function AddToSyllables(AWord) {
// Since syllables is closed over, it is accessible here
// (but WON'T be accessible outside this scope).
syllables.push(AWord);
return syllables.length;
}
// Expose whatever functionality you want to be "public"
// in the returned object.
return {
AddToSyllables: AddToSyllables
};
}
var myObj = createURLObject();
myObj.AddToSyllables("text");
It is, of course, valuable to understand JavaScript's quirky (and surprising, to most developers coming from other languages) behavior with respect to this. That said, once you do understand it, I suspect you will find that it can often be avoided altogether.
you need to do this :
function URLObject()
{
var that = this;
that.AddtoSyllables = AddtoSyllables;
function AddtoSyllables(AWord)
etc...
Like this you can add method and attributes to one object.
The issue you are having is that the function AddtoSyllables is not a member function or method of the URLObject. It is just a nested function with no object attachments, so all usages of this will result in returning the dom window object. The correct way of declaring the AddtoSyllables function is this:
function URLObject()
{
//...
}
URLObject.prototype.AddtoSyllables = function (AWord)
{
var SylCount = this.syllables.length;
alert("This is SylCount: " + SylCount);
}
To explain the reasons of the behavior in the question, I'd like to clarify that objects in javascript are treated like a map, dictionary or a key-value pair (use the term what suits you best). Using the syntax x.y = value; is equivalent putting the value value into the map x with key y. Having the code:
this.AddtoSyllables = AddtoSyllables;
function AddtoSyllables(AWord)
{
var SylCount = this.syllables.length;
alert("This is SylCount: " + SylCount);
}
adds the AddtoSyllables function as an entry to the object this points to.
The code
myobj.AdtoSyllables(...)
is equivalent to
myobj["AdtoSyllables"](...) // now a retreiaval operation
or even
var fn = myobj["AdtoSyllables"];
fn (...);
Inside the AdtoSyllables function, this is used. Against common expectations, it is not a pointer to the myobj.
The cause of this is that AddtoSyllables is treated as a static method of the URLObject class (as OOP guys would understand it), or even a loose static function (like in C). To make JS treat it like a member of the URLObject object (an instance method to OOP guys), JS must be told to do so. This is achieved through the URLObject.prototype.AddtoSyllables = .... which equivalents to declaration of an instance method.
From an alternative point of view:
function foo() { /* some code using `this` */ }
var bar = {};
var baz = {};
bar.foo = foo; // same as bar["foo"] = foo;
baz.foo = foo; // same az baz["foo"] = foo;
In the above code, this usages inside foo will neither point to bar, nor baz. At the same time bar.foo will point to the very same instance as baz.foo, for foo is also an object.

JavaScript library pattern

I'm trying to figure out the basic pattern for creating a JavaScript library (class). I want to do it in such a way that it doesn't pollute the global namespace with a bunch of junk, but allowing for the class to have instance variables and public methods that modify those instance variables.
Consider the following toy example. I want to make a class Foo. It should contain an instance member, bar, which is a number. There should be a constructor for Foo that takes a number and initializes its instance bar with that number. There should be an instance method that I can call on a Foo object to modify bar. Maybe the code that uses the library looks like this:
var foo1 = new Foo(1);
var foo2 = new Foo(2);
console.log(foo1.bar); // should print "1"
console.log(foo2.bar); // should print "2"
foo2.changeBar(42);
console.log(foo1.bar); // should print "1"
console.log(foo2.bar); // should print "42"
The resultant foo.js would be used by a Web app and therefore included via a script tag in the HTML.
I've done a bit of searching with Google but I have yet to find a single, concise, generic outline of how to design a JavaScript class (used as a library).
(function () {
Foo = function (num) {
this.changeBar(num);
};
var privateMethod = function (x) {
if (this.bar === 999) {
this.bar = x;
}
};
Foo.prototype.changeBar = function (num) {
this.bar = num;
privateMethod.call(this, 1);
};
}());
That is the simplest way of doing it. You don't need to include the definition in a closure, more of a style thing.
Building up on Evan's answer, to showcase some more possibilites. Most normal cases only use some of these though.
(function() {
//When we create variables inside a function they can only be seen by other
// inner functions. Wraping all our code here makes sure noone gets to see
// the private stuff.
//The first opening parenthesis is required for Javascript to parse it
//correctly though
//this is the constructor function
//Note how we set a global variable (Foo) to make it available outside.
Foo = function(num, another_num) {
this.changeBar(num);
//sometimes you will want to make a method that modifies a variable
//that can't be acessed via this.xxx. You can use a closure here for that
//(there are some performance considerations though)
this.baz = function(){
console.log(another_num);
}
//Setting methods "by hand" like this is also useful if you want to
//do multiple inheritance, since you can just set all methods of the
//second class by hand here if you want.
}
//Things on Foo.prototype will be available for all Foo objects,
// via prototypal inheritance magic.
Foo.prototype.changeBar = function(num) {
this.bar = num;
}
var a_secret_variable = 42;
function my_private_function(){
console.log(a_secret_variable);
}
//All private variables can be normaly used (by functions that can see them).
Foo.prototype.use_magic = function(){
my_private_function();
}
}());
//The "fake" function is imediatelly called,
//so in the end all it does is create a inner scope.
The Module Pattern is probably the most popular pattern used today.
var Foo = (function() {
var _thisIsAPrivateVar;
function thisIsAPrivateMethod() {
}
return {
thisIsAPublicMethod : function() {
// can still access the "private" variable and method
}
};
})();

Accessing "pseudo-globals" by their name as a string

I am now in the process of removing most globals from my code by enclosing everything in a function, turning the globals into "pseudo globals," that are all accessible from anywhere inside that function block.
(function(){
var g = 1;
var func f1 = function () { alert (g); }
var func f2= function () { f1(); }
})();
(technically this is only for my "release version", where I append all my files together into a single file and surround them with the above....my dev version still has typically one global per js file)
This all works great except for one thing...there is one important place where I need to access some of these "globals" by string name. Previously, I could have done this:
var name = "g";
alert (window[name]);
and it did the same as
alert(g);
Now -- from inside the block -- I would like to do the same, on my pseudo-globals. But I can't, since they are no longer members of any parent object ("window"), even though are in scope.
Any way to access them by string?
Thanks...
Basically no, as answered indirectly by this question: Javascript equivalent of Python's locals()?
Your only real option would be to use eval, which is usually not a good or even safe idea, as described in this question: Why is using the JavaScript eval function a bad idea?
If the string name of those variables really and truly is defined in a safe way (e.g. not through user-input or anything), then I would recommend just using eval. Just be sure to think really long and hard about this and whether there is not perhaps a better way to do this.
You can name the function you are using to wrap the entire code.
Then set the "global" variable as a member of that function (remember functions are objects in JavaScript).
Then, you can access the variable exactly as you did before....just use the name of the function instead of "window".
It would look something like this:
var myApp = new (function myApp(){
this.g = "world";
//in the same scope
alert ( "Hello " + this["g"]);
})();
//outside
alert ( "Hello " + myApp["g"]);
if you want to access something in a global scope, you have to put something out there. in your case it's probably an object which references your closed off function.
var obj1 = new (function(){
var g = 1;
var func f1 = function () { alert (g); }
var func f2= function () { f1(); }
})();
you can add a method or property as a getter for g. if the value of g isn't constant you might do like
this.getG = function() { return g; };
you can work from there to access items by name, like
alert( obj1["getG"]() );
alert( window["obj1"]["getG"]() );

How to get class object's name as a string in Javascript?

Let's say I instantiate an object in Javascript like this:
var myObj = new someObject();
Now, is it possible to obtain the var object's name as string 'myObj' from within one of the class methods?
Additional details (edited):
The reason why I would like to get the name of the variable holding reference to the object is that my new myObj would create a new clickable DIV on the page that would need to call a function myObj.someFunction(). As I insert the new DIV I need to know the name of the variable holding reference to the object. Is there maybe a better way of doing this?
You are right, sorry for the mixup in terminology.
The reason why I would like to get the name of the variable holding reference to the object is that my new myObj would create a new clickable DIV on the page that would need to call a function myObj.someFunction(). As I insert the new DIV I need to know the name of the variable holding reference to the object. Is there maybe a better way of doing this?
Shog9 is right that this doesn't make all that much sense to ask, since an object could be referred to by multiple variables. If you don't really care about that, and all you want is to find the name of one of the global variables that refers to that object, you could do the following hack:
function myClass() {
this.myName = function () {
// search through the global object for a name that resolves to this object
for (var name in this.global)
if (this.global[name] == this)
return name
}
}
// store the global object, which can be referred to as this at the top level, in a
// property on our prototype, so we can refer to it in our object's methods
myClass.prototype.global = this
// create a global variable referring to an object
var myVar = new myClass()
myVar.myName() // returns "myVar"
Note that this is an ugly hack, and should not be used in production code. If there is more than one variable referring to an object, you can't tell which one you'll get. It will only search the global variables, so it won't work if a variable is local to a function. In general, if you need to name something, you should pass the name in to the constructor when you create it.
edit: To respond to your clarification, if you need to be able to refer to something from an event handler, you shouldn't be referring to it by name, but instead add a function that refers to the object directly. Here's a quick example that I whipped up that shows something similar, I think, to what you're trying to do:
function myConstructor () {
this.count = 0
this.clickme = function () {
this.count += 1
alert(this.count)
}
var newDiv = document.createElement("div")
var contents = document.createTextNode("Click me!")
// This is the crucial part. We don't construct an onclick handler by creating a
// string, but instead we pass in a function that does what we want. In order to
// refer to the object, we can't use this directly (since that will refer to the
// div when running event handler), but we create an anonymous function with an
// argument and pass this in as that argument.
newDiv.onclick = (function (obj) {
return function () {
obj.clickme()
}
})(this)
newDiv.appendChild(contents)
document.getElementById("frobnozzle").appendChild(newDiv)
}
window.onload = function () {
var myVar = new myConstructor()
}
Short answer: No. myObj isn't the name of the object, it's the name of a variable holding a reference to the object - you could have any number of other variables holding a reference to the same object.
Now, if it's your program, then you make the rules: if you want to say that any given object will only be referenced by one variable, ever, and diligently enforce that in your code, then just set a property on the object with the name of the variable.
That said, i doubt what you're asking for is actually what you really want. Maybe describe your problem in a bit more detail...?
Pedantry: JavaScript doesn't have classes. someObject is a constructor function. Given a reference to an object, you can obtain a reference to the function that created it using the constructor property.
In response to the additional details you've provided:
The answer you're looking for can be found here: JavaScript Callback Scope (and in response to numerous other questions on SO - it's a common point of confusion for those new to JS). You just need to wrap the call to the object member in a closure that preserves access to the context object.
You can do it converting by the constructor to a string using .toString() :
function getObjectClass(obj){
if (typeof obj != "object" || obj === null) return false;
else return /(\w+)\(/.exec(obj.constructor.toString())[1];}
You might be able to achieve your goal by using it in a function, and then examining the function's source with toString():
var whatsMyName;
// Just do something with the whatsMyName variable, no matter what
function func() {var v = whatsMyName;}
// Now that we're using whatsMyName in a function, we could get the source code of the function as a string:
var source = func.toString();
// Then extract the variable name from the function source:
var result = /var v = (.[^;]*)/.exec(source);
alert(result[1]); // Should alert 'whatsMyName';
If you don't want to use a function constructor like in Brian's answer you can use Object.create() instead:-
var myVar = {
count: 0
}
myVar.init = function(n) {
this.count = n
this.newDiv()
}
myVar.newDiv = function() {
var newDiv = document.createElement("div")
var contents = document.createTextNode("Click me!")
var func = myVar.func(this)
newDiv.addEventListener ?
newDiv.addEventListener('click', func, false) :
newDiv.attachEvent('onclick', func)
newDiv.appendChild(contents)
document.getElementsByTagName("body")[0].appendChild(newDiv)
}
myVar.func = function (thys) {
return function() {
thys.clickme()
}
}
myVar.clickme = function () {
this.count += 1
alert(this.count)
}
myVar.init(2)
var myVar1 = Object.create(myVar)
myVar1.init(55)
var myVar2 = Object.create(myVar)
myVar2.init(150)
// etc
Strangely, I couldn't get the above to work using newDiv.onClick, but it works with newDiv.addEventListener / newDiv.attachEvent.
Since Object.create is newish, include the following code from Douglas Crockford for older browsers, including IE8.
if (typeof Object.create !== 'function') {
Object.create = function (o) {
function F() {}
F.prototype = o
return new F()
}
}
As a more elementary situation it would be nice IF this had a property that could reference it's referring variable (heads or tails) but unfortunately it only references the instantiation of the new coinSide object.
javascript: /* it would be nice but ... a solution NOT! */
function coinSide(){this.ref=this};
/* can .ref be set so as to identify it's referring variable? (heads or tails) */
heads = new coinSide();
tails = new coinSide();
toss = Math.random()<0.5 ? heads : tails;
alert(toss.ref);
alert(["FF's Gecko engine shows:\n\ntoss.toSource() is ", toss.toSource()])
which always displays
[object Object]
and Firefox's Gecko engine shows:
toss.toSource() is ,#1={ref:#1#}
Of course, in this example, to resolve #1, and hence toss, it's simple enough to test toss==heads and toss==tails. This question, which is really asking if javascript has a call-by-name mechanism, motivates consideration of the counterpart, is there a call-by-value mechanism to determine the ACTUAL value of a variable? The example demonstrates that the "values" of both heads and tails are identical, yet alert(heads==tails) is false.
The self-reference can be coerced as follows:
(avoiding the object space hunt and possible ambiguities as noted in the How to get class object's name as a string in Javascript? solution)
javascript:
function assign(n,v){ eval( n +"="+ v ); eval( n +".ref='"+ n +"'" ) }
function coinSide(){};
assign("heads", "new coinSide()");
assign("tails", "new coinSide()");
toss = Math.random()<0.5 ? heads : tails;
alert(toss.ref);
to display heads or tails.
It is perhaps an anathema to the essence of Javascript's language design, as an interpreted prototyping functional language, to have such capabilities as primitives.
A final consideration:
javascript:
item=new Object(); refName="item"; deferAgain="refName";
alert([deferAgain,eval(deferAgain),eval(eval(deferAgain))].join('\n'));
so, as stipulated ...
javascript:
function bindDIV(objName){
return eval( objName +'=new someObject("'+objName+'")' )
};
function someObject(objName){
this.div="\n<DIV onclick='window.opener."+ /* window.opener - hiccup!! */
objName+
".someFunction()'>clickable DIV</DIV>\n";
this.someFunction=function(){alert(['my variable object name is ',objName])}
};
with(window.open('','test').document){ /* see above hiccup */
write('<html>'+
bindDIV('DIVobj1').div+
bindDIV('DIV2').div+
(alias=bindDIV('multiply')).div+
'an aliased DIV clone'+multiply.div+
'</html>');
close();
};
void (0);
Is there a better way ... ?
"better" as in easier? Easier to program? Easier to understand? Easier as in faster execution? Or is it as in "... and now for something completely different"?
Immediately after the object is instantiatd, you can attach a property, say name, to the object and assign the string value you expect to it:
var myObj = new someClass();
myObj.name="myObj";
document.write(myObj.name);
Alternatively, the assignment can be made inside the codes of the class, i.e.
var someClass = function(P)
{ this.name=P;
// rest of the class definition...
};
var myObj = new someClass("myObj");
document.write(myObj.name);
Some time ago, I used this.
Perhaps you could try:
+function(){
var my_var = function get_this_name(){
alert("I " + this.init());
};
my_var.prototype.init = function(){
return my_var.name;
}
new my_var();
}();
Pop an Alert: "I get_this_name".
This is pretty old, but I ran across this question via Google, so perhaps this solution might be useful to others.
function GetObjectName(myObject){
var objectName=JSON.stringify(myObject).match(/"(.*?)"/)[1];
return objectName;
}
It just uses the browser's JSON parser and regex without cluttering up the DOM or your object too much.

Categories

Resources