«

Hello everybody, I have question here about bit'pos. I saw this
function from a place.

function bt_to_natural(bv: in bit_vector) return natural is
variable result : natural:=0;
begin
for index in bv'range loop
result :=result*2+bit'pos(bv(index));
end loop;
return result;
end bv_to_natural;

I am searching the bit'pos on the internet. And I found one
explaination is :
signal 'pos is return the position (number) of the input.
bit' pos('0') would return 0; bit'pos('1') would return 1;
I am a little bit confused about it. I think 'pos should be the
position of the bit. But I take my definition into above function. I
cannot figure it out.

Is there anybody to help me figure it out? Could you give a real
example for me? For example how to convert 10101010 to the integer.
Many thanks.

Zhi

schrieb:

> Hello everybody, I have question here about bit'pos. I saw this
> function from a place.

Consult the VHDL LRM (Language Reference Manual)


http://www.microlab.ch/courses/vlsi/...76_14.HTM#14.1

bit is an enumaration type:

type BIT is ('0', '1'); -- package standard

So bit'pos is a way to convert the type bit to integer:

bit'pos('0')=0
bit'pos('1')=1

Attrributs pos and val can be used to convert enumaration types:
variable c: character;#
variable b: bit;
variable i: integer;

c:=character'val(bit'pos(b)+character'pos('0')); --
character'pos('0')=48 in ASCII character set
b:=bit'val(character'pos(c)-character'pos('0'));

('val is the opposite of 'pos)

Hubble.

Thank you, Hubble. I understand your explanation..
But I am still confused it a little bit. If I take an example of
bit_vector ("10101010"). I want to convert it to integer. According to
the function:
1. result:=0+0;
2 result:=2*0+1;
3 result:=2*1+0;
.........................Does it like that? I think my way is wrong.
Because It definitedly cannot get the expected result.
Could you give me a clear example like that ? Since the function can
convert bitvector to natural, how does it work? Thanks again.

Hubble wrote:
> schrieb:
>
> > Hello everybody, I have question here about bit'pos. I saw this
> > function from a place.
>
> Consult the VHDL LRM (Language Reference Manual)
>
>
> http://www.microlab.ch/courses/vlsi/...76_14.HTM#14.1
>
> bit is an enumaration type:
>
> type BIT is ('0', '1'); -- package standard
>
> So bit'pos is a way to convert the type bit to integer:
>
> bit'pos('0')=0
> bit'pos('1')=1
>
> Attrributs pos and val can be used to convert enumaration types:
> variable c: character;#
> variable b: bit;
> variable i: integer;
>
> c:=character'val(bit'pos(b)+character'pos('0')); --
> character'pos('0')=48 in ASCII character set
> b:=bit'val(character'pos(c)-character'pos('0'));
>
> ('val is the opposite of 'pos)
>
> Hubble.

wrote:
> Thank you, Hubble. I understand your explanation..
> But I am still confused it a little bit. If I take an example of
> bit_vector ("10101010"). I want to convert it to integer. According to
> the function:
> 1. result:=0+0;
> 2 result:=2*0+1;
> 3 result:=2*1+0;

For it to function properly, bv'range must be max downto 0. It must be
sequencing from MSB to LSB.

with an input bit vector of "1100"

1. index = 0, result = 0*2 + 1 = 1
2. index = 1, result = 1*2 + 1 = 3
3. index = 2, result = 3*2 + 0 = 6
4. index = 3, result = 6*2 + 0 = 12

I got it. Thank you all very much ^_^

D Stanford wrote:
> wrote:
> > Thank you, Hubble. I understand your explanation..
> > But I am still confused it a little bit. If I take an example of
> > bit_vector ("10101010"). I want to convert it to integer. According to
> > the function:
> > 1. result:=0+0;
> > 2 result:=2*0+1;
> > 3 result:=2*1+0;
>
> For it to function properly, bv'range must be max downto 0. It must be
> sequencing from MSB to LSB.
>
> with an input bit vector of "1100"
>
> 1. index = 0, result = 0*2 + 1 = 1
> 2. index = 1, result = 1*2 + 1 = 3
> 3. index = 2, result = 3*2 + 0 = 6
> 4. index = 3, result = 6*2 + 0 = 12

D Stanford a écrit :

> wrote:
> > Thank you, Hubble. I understand your explanation..
> > But I am still confused it a little bit. If I take an example of
> > bit_vector ("10101010"). I want to convert it to integer. According to
> > the function:
> > 1. result:=0+0;
> > 2 result:=2*0+1;
> > 3 result:=2*1+0;
>
> For it to function properly, bv'range must be max downto 0. It must be
> sequencing from MSB to LSB.
It works with any range, doesn't it ?

JD.

> > For it to function properly, bv'range must be max downto 0. It must be
> > sequencing from MSB to LSB.
> It works with any range, doesn't it ?
>
> JD.

A good point. It would have been better stated max downto min instead
of 0.