Array of Objects

class Dims {
public static void main(String[] args) {
int[][] a = {{1,2,}, {3,4}};

Object[][] o1 = a;//why this illegal?

but

Object[]o1 = a;//legal??
}
}

Remember in Java you don't have multidimesional arrays, you have arrays of arrays.

int[][] a = {{1,2,}, {3,4}}; // This is just an array, and every of it's elements is an array. But is is just and array.

Object[][] o1 = a; // You are trying to assing an array of arrays to an array. Different Types!!!

Object[]o1 = a; // You are trying to assing an array to an array. That's OK.

Can some one say how to distinguish between arrays of arrays and just arrays in the above case

Bye & Regards,
Praveen

Originally posted by saied ims:
class Dims {
public static void main(String[] args) {
int[][] a = {{1,2,}, {3,4}};

Object[][] o1 = a;//why this illegal?

but

Object[]o1 = a;//legal??
}
}

An int is not an Object, so you can't assign an array of array of ints to an array of array of Objects.

An array is an Object (Array extends Object), so you can assign an array of Arrays to an array of Objects.

Hope this is clear (too many arrays...)

Cheers !
--JA

Originally posted by Javier Sanchez Cerrillo:
Remember in Java you don't have multidimesional arrays, you have arrays of arrays.

int[][] a = {{1,2,}, {3,4}}; // This is just an array, and every of it's elements is an array. But is is just and array.

Object[][] o1 = a; // You are trying to assing an array of arrays to an array. Different Types!!!

Object[]o1 = a; // You are trying to assing an array to an array. That's OK.

a IS an array of array...

Cheers!
--JA

Hi,

I don't fully understand this.

This is a 2 dimensional array of ints

This code won't compile because you are trying to assign a 2-D array of ints to a 2-D array of Objects.

This is the part I don't understand. Why is it legal to assign a 2-D int array to a 1-D Object array?

Thanks

me too why??

This is the part I don't understand. Why is it legal to assign a 2-D int array to a 1-D Object array?

If p is a primitive type, then: Object > p[]. It means p[] array can be assigned to Object.

What if, p is 2-D primitive array and Object is an 1-D array. (Your case)

In that case we can say that Object[] > p[][]. So here again 2-D p array can be assigned to 1-D Object array.

Similary this also holds. Object [][] > p[][][]

Naseem

Thanks Naseem,

I still don't understand this.

Every array, regarless of dimensions is essentially just an object so I can understand why it would be legal to assign an array of any dimension to an Object reference variable.

However, why would it make sense to assign a 2-D int array to a 1-D Object array? What would we be trying to achieve when doing this? Does the 1-D Object array become transformed into a 2-D int array?
How could we access the 2-D int array from the 1-D Object array again afterwards.

Sorry, lots of questions but I am really confused!!

Thanks

Originally posted by d jones:

How could we access the 2-D int array from the 1-D Object array again afterwards.

In this case, the elements of the object array will be "int array references".

Check the Below Code

Originally posted by d jones:
However, why would it make sense to assign a 2-D int array to a 1-D Object array? What would we be trying to achieve when doing this? Does the 1-D Object array become transformed into a 2-D int array?

JLS says if S and T are both reference types, then S[] > T[] iff S > T. [Source: JLS 4.10.3 Subtyping among Array Types]

So you can assign T in S. This is legal S = T;

e.g.,

Now fit all these cases on S[] and T[].

1. Can you assign int[] to Object[]? Certainly No. as both are not reference types. One is array of primitive and other is of object references.

2. What about this? int[][] to Object[]. 1-D int array is an object. So
int[][] is nothing but T[] where T is reference of int[]. Now can you assign T[] to Object[]. Yes of course.

How could we access the 2-D int array from the 1-D Object array again afterwards.

Just go through this code.

Naseem

1-D int array is an object. So
int[][] is nothing but T[] where T is reference of int[]. Now can you assign T[] to Object[].

What do you mean by this?

Could you recommend any extra reading about what is going on here. I really don't undersand this.

Is this relevant for SCJP 1.4?

Many Thanks

Hi,

Can anyone help with this?

Many Thanks

From JLS 4.10.3, if S and T are both reference types, then S[] > T[] iff S > T. it means T can be assigned to S.

In the same way a 2-D primitive array (int[][]) can be assigned to 1-D object array as int[][] is nothing but a 1-D array of object references.

Naseem

class casting
{
public static void main(String args[])
{
int[][] a=new int[3][2];
int[] b=new int[2];
Object o;
b=a[2];
o=a[2]; // How this line works
o=a; // Even this line works
a=(int[][])o;
}
}
The above code works fine.
How can

Hi Vinod,

Every array is an object.

So
o=a[2];
a[2] is a 1-D array which holds an array. Every array extends from Object so we can assign any array to an object reference.

a is a 2-D array, which is just an array of arrays. Again this extens from Object so we can assign the 2-D array to the object reference variable.

Just to clarify this ....

Can somebody please correct me if I am wrong:

int[][] intArray={{2,3,4}, {10,11,12}};

Object[] objArray=intArray;

Here will the objArray have 2 elements, each holding a 1-D array

And in the case:
int[][][] intArray={{{1,2},{1,2},{1,2}}, {{1,2},{1,2},{1,2}}};
Object[] objArray=intArray;

Here will the objArray have 2 elements, each holding a 2-D array

Thanks