There is a (fundamental, in my opinion) difference between object literals and functions, the "private" variables. Since an object can't be instantiated(because it is already an instance of Object
) it has no possibility to have its own (new) scope. It is a base concept of advanced JS programming. Having a new scope allows you to do almost everything(you can declare your own window
, document
or whatever you want except the JS keywords inside your own scope). Now, some simple examples:
Let's assume you want to create a large number of instances of the same object(using as few lines as possible):
function MyObj(i) {
var privateCounter = "I am the instantiated object " + i + " .";
this.counter = function() {
return privateCounter;
};
}
var MyObjList = [],
ObjLitList = [];
for (var i = 0; i < 100; i++) {
MyObjList.push(new MyObj(i));
ObjLitList.push({counter: "I am the literal object number " + i + "."});
}
Now you have 200 objects that are almost, but not precisely, the same thing. You can extend them as you prefer, because functions are objects, but in the case of the function you cannot access the private
variable directly. Let's see which are the advantages of a function:
- It is treated like an
Object
- It has its own
Prototype
- It has private variables
And the Object
s?
- It is an
Object
- It doesn't have its own
Prototype
, but you can declare the functions and extend the object itself
- It doesn't have private variables
Apart the private vars, they are not much different from each other.
Let's see what a function's prototype can do:
MyObj.prototype.setX = function(x) {
this.x = x;
}
Using the prototype allows you to create an only instance of an anonymous function(which can be named too and then assigned) which will be shared across instances. How can you do the same thing with object literals?
function setX(x) {
this.x = x;
}
var obj = {
setX: setX
};
As you can see you have to create the object defining everytime a property which is setX
. Otherwise, you can extend Object.prototype
itself(but there is a long debate about extending the prototype of native JS objects).
So which is the best way? There is no one, it depends on what you have to do, what you need from your script, which of the two you feel more comfortable with.
I prefer writing my own functions and treat them like classes, because they are more readable and I am able to use "private" variables. I don't know anyone using literals instead of functions though.
As for the questions:
Which is the best preferred way of programming(object literals vs constructors vs prototype)
Answered.
can a code with constructor and protoype be written using just object literals without using constructor and protoype.
Yes, you can if you don't need private variables(and if the script isn't too big. Imagine jQuery written as an Object literal :D).
what is the signifiance of anonymous function.
Oh well, I can answer with an example:
//code
myNamedFunction();
//code
function myNamedFunction() {
alert("I'm defined everywhere! :)");
}
This works and won't generate a TypeError
.
myAnonymousFunction();
var myAnonymousFunction = function() {
alert("I'm defined after this declaration :(");
}
myAnonymousFunction(); // works!
This will cause a Uncaught TypeError: undefined is not a function
, because myAnonymousFunction
is only a reference to the effective function(which is unnamed, so it is not callable from the script).
There are a lot of things to say about this argument, and a good point to start advanced programming is Javascript Garden. Other good readings are Basics of OOP in JS - NetTutsPlus, Working with Objects - MDN and OOP in JS - Phrogz
Hope this helps!
Sidenote: functions also have a good advantage since they can change their context(this
) just with a function(call
for example), while objects can't.