Bit level operations, get a bit from a short value

Question

I'm trying to get the following result 00000000000000101 from this input 5.

This my code and it doesn't make sense:

public static int get16Bits(int x)
    {
        System.out.println((x & 0xffff) - ((x & 0x8000) << 1));
        return (x & 0xffff) - ((x & 0x8000) << 1);

    }

How i can get the 16bits of integer value?

Which *16bits* do you mean? Your question doesn't seem to make much sense. `00000000000000101` is just the binary representation of `5` with some leading zeros. So for this example input your method should just return the input as far as I guess. — SpaceTrucker, Sep 22 '17 at 12:49
Try this: `String.format("%016d", Integer.parseInt(Integer.toBinaryString(x)))` — Procrastinator, Sep 22 '17 at 12:50
@procrastinator This is what he wants? Nice, if you guessed correctly you should post an answer, I will upvote. — Oleg, Sep 22 '17 at 12:51
@Oleg that's what I understood from the problem. Ain't sure whether they meant the same or different. — Procrastinator, Sep 22 '17 at 12:53

Procrastinator · Answer 1 · 2017-09-25T07:23:38.300

8

Try changing your return statement a bit:

    public static String get16Bits(int x)
    {
        return (String.format("%016d", Integer.parseInt(Integer.toBinaryString(x))));

    }

This should help you.

edited Sep 25 '17 at 07:23

answered Sep 22 '17 at 12:52

Procrastinator

2,526
30
27
36

yup it's working. but not perfect... what about negative numbers ? also i need to get result without methods. – Sep 22 '17 at 12:53
1

change your method signature or your return type, because it does not compile. – Schidu Luca Sep 22 '17 at 12:54
@android without using any of the methods, it'd be difficult. – Procrastinator Sep 22 '17 at 12:58
@procrastinator it would be mode easier and much funny and more useful. – Sep 22 '17 at 12:58

Sergey Kalinichenko · Answer 2 · 2017-09-22T13:04:11.860

6

The first part is self-explanatory: the upper 16 bits are removed with & 0xffff masking.

Subtracting (x & 0x8000) << 1 is a trick that keeps the proper sign of the 16-bit number in the 32-bit result: for example, 0xffff, a -1 in 16-bit representation, gets converted to 0xffffffff, which is also -1 in 32-bit representation. This is better than using a conditional, because it is branch-free.

Another trick that you could use is

(((x & 0xFFFF) << 16) >> 16)

This lets the sign bit of 16-bit number "touch" the sign bit of 32-bit number, letting the right shift to sign-extend the result.

Note: If you are looking for a binary String representation of the number, only x & 0xffff masking is necessary, because the upper 16 bits are dropped from the string result anyway. This Q&A explains how to obtain a binary representation of an integer with the appropriate number of leading zeros.

edited Sep 22 '17 at 13:04

answered Sep 22 '17 at 12:54

Sergey Kalinichenko

714,442
84
1,110
1,523

Well, thanks for help, but it's still not perfect solution...! negative numbers didn't converted. – Sep 22 '17 at 13:07
@android Could you provide an example of a negative number that failed to convert? – Sergey Kalinichenko Sep 22 '17 at 13:08
-5 would be 1111111111111011 – Sep 22 '17 at 13:08
2

@android That's the proper two's complement representation of negative five, though: you invert `0000000000000101` which becomes `1111111111111010`, and add `1` to the result, producing `1111111111111011` – Sergey Kalinichenko Sep 22 '17 at 13:11

score 1 · Accepted Answer · answered Sep 22 '17 at 15:11

We can try this code below, we will use for loop :

public void getBits(int number){

      String bits = "";
      for (int i = 0; i < 16; i++) {
          bits = (number & 1) + bits;
          number >>= 1;
      }
      System.out.println("The bits are " + bits);
  }

score 0 · Answer 4 · answered Sep 22 '17 at 12:54

0

Try below :

String.format("%16s", Integer.toBinaryString(5)).replace(' ', '0')

answered Sep 22 '17 at 12:54

Raju Sharma

2,496
3
23
41

Bit level operations, get a bit from a short value

4 Answers4