In JavaScript, why is [ ] preferred over new Array();?

Question

I remember reading somewhere (I think it was in one of Crockford's papers) that using an array literal [] is better than using the new Array(); notation.

But I can't really remember any advantages of one over the other.

Can anyone please explain to me on why the former is preferred over the latter?

Here is one reason I can think of on why [] is better than new Array();:

var Array = function () { };

Overriding the Array object will break code...!

Any more reasons?

Answer 1

Brevity

It has less bytes to transfer over the wire, less bytes to interpret, less mental resources to parse it.

Less is more.

Consistency

What is the difference between these two lines of code?

var arr = [5];
var arr = new Array(5);

According to here new Array(5); will not return an array with a 5 in it, instead it will return a 5 element array, with all the elements being undefined. Whereas these two lines return identical arrays.

var arr = [5,6];
var arr = new Array(5,6);

Answer 2

One good reason is because the Array constructor has completely non-intuitive behavior. For example:

var a = new Array(5);
console.log(a.length); //prints "5"
console.log(a[0]); //prints "undefined"

var b = new Array(1,2);
console.log(b.length); //prints "2"
console.log(b[0]); //prints "1"

In this case, a ends up being an array of size 5 with all elements undefined, and b ends up being an array of size 2, with the values [1,2].

var c = new Array("5");
console.log(c.length); //prints "1"
console.log(c[0]); //prints "5"

And here, you end up with a single-element array, containing "5"

In general, you should probably never use the constructors on built-in types in Javascript. They all have weird edge cases like this. For example:

var s = new String("Hello");
var l = "Hello";
console.log(typeof(s)); // prints "object"
console.log(typeof(l)); // prints "string"