Implementing my own Integer.toBinaryString(int n) method

Our senior developer gave us (trainees/jr. developers) a small exercise and that is to make our own Integer.toBinaryString(int n) implementation. This is the answer I came up with. I would just like to hear comments/suggestions/opinions on this. Especially, if there is a way to optimize my answer.

public static String toBinaryString(int n){

 String binary = "";

 if(n == 0) return "0";

 // I know, I'm desperate :(
 if(n == Integer.MIN_VALUE) return "10000000000000000000000000000000";

 if(n < 0){

     // Get the positive equivalent
     int val = n * -1;

     // Convert into binary
     String initial = toBinaryString(val);
     String inverted = "";

     // Get 1's complement
     for(char chars : initial.toCharArray()){
        inverted += String.valueOf(((chars == '1') ? '0' : '1'));
     }

     int carry = 0;

     /*Check least significant bit.
     If 0, simply change it to 1.
     If 1, perform addition of 0b1*/
     if(inverted.charAt(inverted.length()-1) == '1'){

         boolean carriedOver = false;

         for(char chars : new StringBuilder(inverted).reverse().toString().toCharArray()){

             if(carriedOver){
                 binary = chars + binary;
                 continue;
             }

             if(carry > 0){
                 if(chars == '1'){
                     binary = "0" + binary;
                     continue;
                 }else{
                    binary = "1" + binary;
                    carriedOver = true;
                    continue;
                 }
             }

             binary = "0" + binary;
             carry += 1;
         }
     }else{
          StringBuilder sb = new StringBuilder(inverted);
          sb.setCharAt(inverted.length()-1, '1');

          binary = sb.toString();
     }

     return String.format("%32s", binary).replace(" ", "1");
 }

 // Convert to binary
 while(n > 0){
     binary = (n & 1) + binary;
     n /= 2;
 }

 return binary;
 }

If you were our senior developer, would you accept this as a valid answer? Why? Why not?

The exercise calls for bit shifting. Only bit shifting, nothing else, really. Your main tools are:

• checking if the last bit is 0 or 1 with: num & 1
• then shift by one bit to the right: num >> 1

A naive implementation could go like this:

    String result = "";
    while (num > 0) {
        result = (num & 1) + result;
        num >>= 1;
    }
    return result;

But that won’t work for negative numbers. A simple tweak can fix that:

    String result = "";
    while (num != 0) {
        result = (num & 1) + result;
        num >>>= 1;
    }
    return result;

Instead of the signed bit shift operator >>, we need to use the unsigned bit shift operator >>>, to shift the negative bit just like all the others. And we changed the condition to != 0 instead of > 0.

But this won’t work for 0. But only for 0. So you can add a simple condition to handle that.

Lastly, string concatenation is inefficient. We can do better using a StringBuilder.
But a StringBuilder only has an append method, doesn’t have prepend. It has an insert method, but that won’t be efficient.
A simple solution is to append the bits and reverse at the end.

String toBinaryString(int num) {
    if (num == 0) {
        return "0";
    }

    StringBuilder builder = new StringBuilder(32);
    while (num != 0) {
        builder.append(num & 1);
        num >>>= 1;
    }
    return builder.reverse().toString();
}

In any case, the StringBuilder is not a critical piece here.
You could use a char[] with 32 elements to store the digits,
and transform that to a string to return.