In the module pattern in JavaScript "Immediately-Invoked Function Expressions" (also known as self-executing anonymous functions) are used as self executing functions that return an object.
How can a self-executing function hide private variables and only expose the returned object. Why does this not happen with a normal JavaScript function?
So in the following mini module, why could we not achieve the same concept of encapsulation without the enclosing ()()?
var Module = (function () {
var privateVariable = "foo",
privateMethod = function () {
alert('private method');
};
return {
PublicMethod: function () {
alert(privateVariable);
privateMethod();
}
};
})();
How can a self-executing function hide private variables and only expose the returned object. Why does this not happen with a normal JavaScript function?
It does happen with normal JavaScript functions.
function MakeModule() {
var privateVariable = "foo",
privateMethod = function () {
alert('private method');
};
return {
PublicMethod: function () {
alert(privateVariable);
privateMethod();
}
};
}
var Module = MakeModule();
would work just fine.
The only difference is that the anonymous function introduces one less global variable and allows for itself to be garbage collected while MakeModule can't be collected unless explicitly deleted by the author.
The privateness is because of closures. The "var privateVariable" is closed over by "PublicMethod", so only that function can access the variable because only it has it in its closure. It cannot be referenced by anything else and is "private"
This happens not only in "Immediately-Invoked Function Expressions" but also in normal function calls. It is just a way to immediately create the closure when defining the module instead of doing it later when you call the outer function.
Also see this post from Douglas Crockford himself: http://javascript.crockford.com/private.html
You can define a anonymous function via named function.
Example:
//factorial
(function(n){
var self = function(n){
//call self
return n > 0 ? (self(n-1) * n) : 1;
}
return self;
})()
Related
I have this code and wondering how can we access properties and method of IIFE
var app = (function () {
function app() {
this.name = 'fasf';
}
app.prototype.greeting = function () {
this.a = 'hello world';
console.log(name);
window.alert('hello world');
};
app.prototype.sayhello = function () {
var j = 't';
console.log(this.a);
};
return app;
}());
also about this piece of code
var t = (function greet(name){
name = name || 'enter your name'
console.log('Hello ' + name);
})();
With the code you have, you can just do this:
// create an instance of the app object
var a = new app();
// call methods on it
a.greeting();
a.sayhello();
Your IIFE returns the internal app constructor function and then assigns that to a variable named app. So you can then do new app() to create an instance of that object. Once you create an instance of that object, you can then call any of the methods on the prototype.
In your second IIFE, nothing is returned from the IIFE so t is undefined.
The IIFE itself does not have properties and methods. That's just an immediately invoked function expression. In your first one, it returns a constructor function which is then assigned to a variable so that variable can be used to create objects with that constructor function. There are no properties or methods of the IIFE itself.
The way the IIFE pattern works is that you use local variables for the stuff you want to keep private. Local variables are are only visible inside the function they were defined in they cannot be read or written from anywhere else in the program.
If you want to expose things inside the IIFE outside the IIFE one way to do it is via the return value of the immediately-invoked function:
var stuff = (function(){
// These variables and functions are not
// visible from the outside:
var myvar1 = "Hello";
var myvar2 = "World";
var func1(){
return myvar1 + " " + myvar2;
}
// But we can return a record that contains a
// reference to our private functions, making them
// visible outside this block.
return {
thefunc: func1
}
}());
//Now we can call the "thefunc" function we exported
console.log( stuff.thefunc() );
I made sure to not repeat any variable names to avoid confusion.
In your first example, there isn't much of a point to using the IIFE pattern since there are no local variables that you are making private. You could have just defined the app function outside the IIFE and it would have worked the same.
The second case could have been written more clearly by putting the function in a separate line:
function greet(name){
name = name || 'enter your name'
console.log('Hello ' + name);
}
var t = greet();
Not only is it a bit weird to immediately invoke a named function but in your particular example, the "default argument value" pattern suggests that this function was designed to be called multiple times, sometimes passing the "name" argument and sometimes not passing it (in which case it defaults to the "enter your name" string.
When using IIFE pattern (closure) you usually keep local variables and function private as properly mentions by hugomg's answer.
In case you need to selectively make public some specific property you could consider using "The Revealing Module Pattern" instead, more info here.
Below a simple example showing this pattern:
var app = (function () {
var privateVar = 'hello world';
function sayhelloPrivate() {
console.log(privateVar);
}
function sayhelloPublic() {
console.log(privateVar);
}
// reveal public pointers to
// private functions and properties
return {
sayhello : sayhelloPublic
};
})();
app.sayhello();
//app.sayhelloPrivate(); this will not work as function is private and not visible outside module
I've the following javascript code
var globalConfiguration = null;
function loadFile(filePath) {
}
function onLoadPage() {
}
function getConfiguration() {
}
function process() {
}
I want to use IIFE to encolose all my functions in a closure to avoid cluttering the global name space,so I wrote :
(function(window){
var globalConfiguration = null;
function loadFile(filePath) {
}
function onLoadPage() {
}
function getConfiguration() {
}
function process() {
}
});
However, I do not understand this notion very well, now in my HTML page how would I call my function onLoadPage?
You can't without putting it in the global namespace somehow.
My recommendation to structure code like this:
function ExampleHelper() {
(function(scope) {
scope.globalConfiguration = null;
scope.loadFile = function(filePath) {
};
scope.onLoadPage = function() {
};
scope.getConfiguration = function() {
};
scope.process = function() {
};
})(this);
}
var foo = new ExampleHelper(); // "foo" now contains all of your functions
foo.onLoadPage();
Now that you have enclosed the module, you need to decide what you want to expose to the outside world. Anything you want to expose, you can export. Also, you need to decide what context (in this case, window) that you want to attach to. Then pass the context in right away (thus completing the IIFE).
For example:
(function(window){
var exports = {};
var globalConfiguration = null;
function loadFile(filePath) {
}
function onLoadPage() {
}
function getConfiguration() {
}
function process() {
}
exports.getConfiguration = getConfiguration;
window.myModule = exports;
})(window);
Attaching to the passed in window object is one way to export things out in a controlled fashion. So, you do need to pass the context (in this case window) to the IIFE. Perhaps, window will not always be the context for the call.
After running this code, myModule will be available on the global scope.
You can set your function to window.onload callback.
(function(window) {
var globalConfiguration = null;
window.onload = onLoadPage;
function loadFile(filePath) {}
function onLoadPage() {
alert('hello world');
}
function getConfiguration() {}
function process() {}
}(window));
This is called chaining of functions/methods and is usually done for better readability of the code and to avoid the usage of temporary variables to hold the return value of each function.
Check this post on chaining methods which helped me to understand the chaining better.
I see you wanted to use closures to avoid cluttering the global object. However, do note that we write functions for reusability. Though you create a closure, ensure that your methods inside the outer function are abstracted such that they can be used by other parts of the code.
For ex: from your code, you have a loadFile method which could be reused.
Now to see how you can use the methods you described in a chain.
Assumptions: (since i don't know why you wrote those methods, i am making some assumptions).
loadFile(filepath) - returns a file object
onPageLoad() - once the page is loaded, it returns the object or id of the input file
getConfiguration() - gets the file path
process() - process the file
onPageLoad().loadFile(getConfiguration()).process();
The important part is that the scope of the object is set correctly in the chaining. In order to do this, each method must return the reference to appropriate object.
Hope this helps.
Is there a way to print the value of temp (a closure variable) from a function defined outside the closure but referenced within the closure without passing temp as a variable to funcA?
var funcA, funcB;
funcA = function () {
console.log(temp);
}
funcB = function () {var temp, funcC;
temp = 1;
funcC = funcA;
funcC();
}
funcB(); // temp is undefined.
This works, but only because funcA is defined within funcB:
funcB = function () {var temp, funcA, funcC;
temp = 1;
funcA = function () {
console.log(temp);
}
funcC = funcA;
funcC();
}
funcB(); // 1
I'm trying to find a way to pull some function definitions out of outer functions to streamline code that's getting a little complex. Can I define funcA outside of funcB but still reference the temp variable without having to pass parameters?
I read that javascript does not have dynamic run time scoping, that it's only lexical scoping, but with referencing a function (funcA via funcC) within funcB, is there a way to meet the lexical scope requirement and provide access to the scoped variables for funcB?
Using Akinkunle Allen's comment, I came up with this which seems to solve my problem.
function funcB () {
var funcA, funcB, temp;
funcA = function () {
console.log(temp);
}
funcB = function () {var funcC;
temp = 1;
funcC = funcA;
funcC();
}
return funcB();
}
funcB(); // 1
Yes, and No.
The keyword var for declaring variables behaves different depending upon what the current scope is. When var is executed on the global scope it is optional. The variable becomes a property of the global object. When executing in the browser this global object is window.
So the following in global space has the same result.
var temp = 1;
window.temp = 1;
this.temp = 1;
All the above is just window.temp because of the global context.
When you use var inside a function then the variable is attached to the function. All functions in Javascript are objects so local var variables will live as long as the parent function is still being used somewhere.
Javascript will walk the hierarchy of executing scopes to find a variable identifier. So any inner functions can access their outer function variables (as in your example).
What you can do is play around with the this reference in functions.
The identifier this in Javascript is dynamic (meaning that you can change it). I can pass a variable to an unknown function that was declared outside the calling function. Since the this.temp is used to reference the variable the function funcA is able to display the value.
funcB = function(otherFunc)
{
this.temp = 1;
otherFunc();
}
funcA = function()
{
alert(this.temp);
}
funcB(funcA);
http://jsfiddle.net/thinkingmedia/dfLvj/
See the above jsfiddle example.
What you can do with this is change the meaning of this on the fly. Here is a better example.
funcB = function(otherFunc)
{
var temp = {
message: "Hello World!"
};
var foo = otherFunc.bind(temp);
foo();
}
funcA = function()
{
alert(this.message);
}
funcB(funcA);
http://jsfiddle.net/thinkingmedia/K5Pw6/
Dynamically changing this can have a lot of benefits by allowing a function to accept closure references that will be executed with a custom this reference.
An example might be a click event handler where this is the DOM element that triggered the event.
En example can be found in Twitter'a typeahead.js here:
function () {
// ...
return this.each(initialize);
function initialize() {
// ...
}
}
Questions:
What are the scopes and what function sees what?
What is the reason for using such a construct (usage scenarios and advantages)?
Javascript has function based scope, which means that every thing defined inside a function is available from the first line, since the definition is "hoisted" by the complier.
That goes for both variable and function definitions - variable values however, are not available until after assignment.
You can read all about javascript scoping and hoisting here
This means that the function initialize is available from the first line of the wrapping anonymous function.
There is no real reason, and no advantages, for doing it that way, unless you count the code structure as an advantage.
Personally I don't see any reason to do this. For me even it looks a little bit weird. Martin is right. You should be careful, because the defined variables are not accessible like functions. For example this doesn't work:
var getValue = function(func) {
return func();
}
var f = function() {
return getValue(now);
var now = function() {
return 10;
}
}
alert(f());
However, this works:
var getValue = function(func) {
return func();
}
var f = function() {
return getValue(now);
function now() {
return 10;
}
}
alert(f());
Is there a way to create private global variables in JavaScript? I have tried looking around, and I keep bumping into talk of constructors - which don't seem too global.
Thanks
Not sure what your use case is. I'll assume you have a js script file containing some functions and variables and you want to expose some of those globally, but keep the rest private to your script file. You can achieve this with a closure. Basically you create a function that you execute immediately. Inside the function you place your original code. You then export the functions you need into the global scope.
// Define a function, evaluate it inside of parenthesis
// and execute immediately.
(function(export) {
var myPrivateVariable = 10;
function myPrivateFunction(param) {
return param + myPrivateVariable;
}
export.myGlobalFunction = function(someNumber) {
return myPrivateFunction(someNumber);
};
})(this); // The *this* keyword points to *window* which
// is *the* global scope (global object) in a web browser
// Here it is a parameter - the *export* variable inside the function.
// This is executed in the global scope
myGlobalFunction(2); // yields 12 (i.e. 2 + 10)
myPrivateVariable; // Error, doesn't exist in the global scope
myPrivateFunction(2) // Error, doesn't exist in the global scope
To answer your question, no, that is not possible as there are no access modifiers in javascript. A variable declared in global scope is accessible to any function.
As pointed out in the comments to this answer, you can create objects which have private members. Crockford has a page on private members in Javascript. He uses the following code to illustrate his point:
function Container(param) {
// private method
function dec() {
if (secret > 0) {
secret -= 1;
return true;
} else {
return false;
}
}
this.member = param;
var secret = 3;
var that = this;
// privileged method
this.service = function () {
return dec() ? that.member : null;
};
}
In the above example, param, secret, and that are all private in that they are not accessible from the outside. To be more clear, these variables can only be accessed by privileged or private methods, the difference being that privileged methods can be called from any instance of the object. As is suggested in the comments, this is achievable by using closures.
Quoting from Crockford for a quick explanation on closures, but you can find plenty of related questions.
What this means is that an inner function always has access to the
vars and parameters of its outer function, even after the outer
function has returned.
in order to have private members. you need to use closures.
following code help you understanding the concept.
function CustomArray () {
this.array = [];
var privateData = 'default data';
this.getPrivateData = function () {
return privateData;
};
this.setPrivateData = function (data) {
privateData = data;
};
};
CustomArray.prototype.push = function (data) {
this.array.push(data);
};
CustomArray.prototype.unshift = function (data) {
this.array.unshift(data);
};
CustomArray.prototype.pop = function () {
this.array.pop();
};
CustomArray.prototype.shift = function () {
this.array.shift();
};
CustomArray.prototype.print = function () {
console.log(this.array.join(','));
};
var array = new CustomArray();
array.push(10);
array.push(20);
array.push(30);
array.push(5);
array.unshift(3);
array.unshift(2);
array.unshift(1);
array.unshift(0);
array.pop();
array.shift();
array.print();
console.log(array.getPrivateData());// default data
array.setPrivateData('am new private data');
console.log(array.getPrivateData());//am new private data