Conversation started Apr 13, 2020 at 11:08.
John Rennie
John Rennie
Apr 13, 2020 11:08
Do you know what hexadecimal notation is?
Aladdin
Aladdin
just that address are represented by them
i don't know in detail
John Rennie
John Rennie
A byte has eight bits, and this neatly splits into two 4 bit chunks (these are often called nibbles)
4 bits can store a number from 0 to 15.
In hexadecimal we write the value of the nibble as 0-9 then A-F.
Aladdin
Aladdin
ok
John Rennie
John Rennie
i.e. if the value in the four bits is in the range 0 to 9 we just write it as 0 to 9. If the value is 10 we write it as A, 11 we write as B, and so on up to 15 we write as F.
Aladdin
Aladdin
ah okay
John Rennie
John Rennie
Apr 13, 2020 11:13
So any byte can be written in hex as two characters 0-9, A-F
Aladdin
Aladdin
ah now it makes sense
John Rennie
John Rennie
The reason we do this is that it makes it easy to see what the byte is because hex is easily converted to binary.
Aladdin
Aladdin
ok
John Rennie
John Rennie
So for example a space character has the ascii value 32, and in binary that is 00100000
If we split this into two we get: 
0010 0000
  2    0
Aladdin
Aladdin
ok
John Rennie
John Rennie
Apr 13, 2020 11:16
Where I've written the corresponding hex digit below each nibble.
That's why 32 decimal is 0x20 hex.
(The 0x prefix just means hex format - it isn't part of the number)
Aladdin
Aladdin
ah okay makes sense now
John Rennie
John Rennie
So anyway, if the four bytes start out as 0x00000000 and we set c[0] = ' ' (32) then the four bytes become 0x20000000.
And if we set c[1] to c[3] to space as well then the four bytes become 0x20202020
OK so far?
Aladdin
Aladdin
yes
John Rennie
John Rennie
If we now use those four bytes as an int then we get the int with the value 0x20202020 and that is the number 538976288 that your program printed.
Aladdin
Aladdin
ok
John Rennie
John Rennie
Apr 13, 2020 11:21
You're probably wondering why on earth anyone would do this, and that's a good question.
Aladdin
Aladdin
yeah.This seems kinda complicated enough
John Rennie
John Rennie
I think there are very few good uses for unions.
Possibly when programming microcontrollers, but very rarely for programs on PCs.
But I guess it's something you need to know to pass the C exams.
Aladdin
Aladdin 
// Unions
#include <stdio.h>
union xyz
{
  int x;
  char y;
  struct x1{
	  int x2;
	  float y2;
  }xx1;
};
int main()
{
 union xyz xy;
  xy.x=321;
  printf("x = %d\t y= %d",xy.x,xy.y);
  xy.y='Z';
  printf("x = %d\t y= %d",xy.x,xy.y);
}
can u explain this if u are free
it seems slightly different from the previous programme
John Rennie
John Rennie
Let me change the program slightly: 
#include <stdio.h>

union xyz
{
  int x;
  char y;
  struct x1{
	  int x2;
	  float y2;
  }xx1;
};

int main()
{
 union xyz xy;
  xy.x=321;
  printf("x = %X\t y= %X\n",xy.x,xy.y);
  xy.y='Z';
  printf("x = %X\t y= %X\n",xy.x,xy.y);

  return 0;
}
I get: 
D:\rhs\c>rat
x = 141  y= 41
x = 15A  y= 5A
Aladdin
Aladdin
yeah even i get the same
John Rennie
John Rennie
Apr 13, 2020 11:31
xy.x is an int, so when we set xy.x = 321 we set the first four bytes to 321, which is 0x141 hex. So the four bytes are now: 
0x00000141
OK so far?
Aladdin
Aladdin
ok
John Rennie
John Rennie
And xy.y is a char, so when you look at the value of xy.y you get just the first byte, which is 0x41. That's why when you print xy.y it prints 41.
Ah, there's something I need to make clear.
Aladdin
Aladdin
ok
John Rennie
John Rennie
In C numbers are stored with the bytes in reverse order. So the number 312 is 0x141 in hex, but it's actually stored as the four bytes:
41 01 00 00
i.e. the bytes are stored in reverse order.
Aladdin
Aladdin
ok
John Rennie
John Rennie
Apr 13, 2020 11:37
That's why when I say xy.y is the first byte it's the byte with the value 0x41
Aladdin
Aladdin
ah let me think over this
John Rennie
John Rennie
I'm not sure I'd worry too much about unions. You rarely use them in real life.
Aladdin
Aladdin
@JohnRennie Will u be back later
John Rennie
John Rennie
Yes, I'll be around later today and tomorrow as usual.
Aladdin
Aladdin
ok
xy.x is an int, so when we set xy.x = 321 we set the first four bytes to 321, which is 0x141 hex. So the four bytes are now:
can u tell how u got 141
John Rennie
John Rennie
Apr 13, 2020 11:55
0x141 is the same number as 312 decimal.
Aladdin
Aladdin
binary of 321 is:
101000001
if we take 8 bits from this
we get:
John Rennie
John Rennie
@Aladdin Split that into nibbles and you get: 
0001 0100 0001
Aladdin
Aladdin
0100,0001
why extra 000
?
John Rennie
John Rennie
Because a nibble is always four bits.
I'm showing all four bits even though some are zero
Aladdin
Aladdin
won't this exceed 4 bytes
John Rennie
John Rennie
Apr 13, 2020 12:00
0001  0100  0001
  1     4     1
@Aladdin I'm not sure why you say that?
Aladdin
Aladdin
0x00000141
so it is:
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0001 0100 0001
John Rennie
John Rennie
No, it's just: 
0000 0000 0000 0000 0000 0001 0100 0001
Eight nibbles = four bytes
Aladdin
Aladdin
oh i get it
In C numbers are stored with the bytes in reverse order. So the number 312 is 0x141 in hex, but it's actually stored as the four bytes:
41 01 00 00
reverse of 0000 0000 0000 0000 0000 0001 0100 0001
John Rennie
John Rennie
Yes, just to be awkward most processors reverse the order when they store ints.
I need to go now. 'll be around later.
 
Conversation ended Apr 13, 2020 at 12:07.