Win a copy of Functional Reactive Programming this week in the Other Languages forum!
  • Post Reply
  • Bookmark Topic Watch Topic
  • New Topic

~ operator

 
Kedar Dravid
Ranch Hand
Posts: 333
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Consider:

char v1 = '(';
int i = ~v1;
System.out.println(i);

Why does the above print a negative value (-41) when the answer we get is of the form 0xffffffd6 (the actual characters may be different), after changing the 1s to 0s and vice-versa.
 
Joe Borderi
Ranch Hand
Posts: 151
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
What does operator~ take as its operand?
 
Surasak Leenapongpanit
Ranch Hand
Posts: 341
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
0xffd6 = 1111 1111 1101 0110
-----------------------------
The most left bit is 1, means that this is negative integer.

0000 0000 0010 1001 = - (32 + 8 + 1) = -41
 
Sumithab Baskaran
Ranch Hand
Posts: 52
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Can someone please explain the basic steps involved in the ~ operation.
For example,
int x = 9;
x = ~ x;

Step 1: Convert 9 to binary: 1001
Step 2: Since the signed bit is one, the answer should be a negative integer
Step 3: Now convert 0's to 1 and one's to 0. That is, 0110
Step 4: The answer: -6.
This is wrong. The answer is -10.
What am I doing wrong here? Please explain.
Thanks
 
Barry Gaunt
Ranch Hand
Posts: 7729
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator


Now we have a bit pattern that we must convert to decimal. The high bit is 1 so it's negative. So let's take the two's complement and then afterwards call the number negative again.


So our answer is the negative of this: -10.
 
Jay Pawar
Ranch Hand
Posts: 411
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Sumitabh,
Follow the steps below


Step 1> Represent 9 in 32 bit integer form.
0000 0000 0000 0000 0000 0000 0000 1001 = 9

Step 2> One's complement of 9
1111 1111 1111 1111 1111 1111 1111 0110

Step 3> To get the final answer do 2's complement on the result obtained in step 2 and add the negative sign.


Hope this helps you.

Always represent number in full 32 bit integer form This is very important when you are trying to play around with bit operators.
 
Barry Gaunt
Ranch Hand
Posts: 7729
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
In fact a bit of algebra "proves" it: Start with a positive integer x.

-x == ~x + 1 // the negative is found by flipping the bits (~x) and adding 1

Subtracting 1 from both sides:

-x - 1 == ~x

That is:
~x == -(x + 1)

So for example: ~9 == -(9 + 1) == -10.

Exercise: does it work if x < 0? for Integer.MAX_VALUE? for Integer.MIN_VALUE?
[ December 06, 2004: Message edited by: Barry Gaunt ]
 
Sumithab Baskaran
Ranch Hand
Posts: 52
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Thanks Barry for the short cut method. ~x = -(x+1).
It works for x <0, x= Integer.MAX_VALUE, Integer.MIN_VALUE.

So, this is all I need to know for ~x. Or do I definitely need to do it the long way - converting to one's compliment and then adding 1 to the result? Let me know so that I can move on.
Thanks once again.
 
Jay Pawar
Ranch Hand
Posts: 411
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Sumitabh,
For Integer.MIN_VALUE , the formula doesn't hold good. However, for ~Integer.MAX_VALUE will give you Integer.MIN_VALUE as per the formula.



In short,

~Integer.MIN_VALUE = Integer.MAX_VALUE
~Integer.MAX_VALUE = Integer.MIN_VALUE
 
Barry Gaunt
Ranch Hand
Posts: 7729
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Originally posted by Jay Pawar:
Sumitabh,
For Integer.MIN_VALUE , the formula doesn't hold good. However, for ~Integer.MAX_VALUE will give you Integer.MIN_VALUE as per the formula.



In short,

~Integer.MIN_VALUE = Integer.MAX_VALUE
~Integer.MAX_VALUE = Integer.MIN_VALUE



But interestingly:


prints: true true

You have to think in a 32bit universe you see. And -2147483649 is not in that universe.
 
  • Post Reply
  • Bookmark Topic Watch Topic
  • New Topic