need the unsigned value from dl.call()

[eliss]

I'm using dl.call() to call a C function in an external library. It's
working great so far except for one function, which returns an
unsigned int in the C version. However, in python it returns a signed
value to me. How can I get the unsigned value from this? I haven't
brushed up on my two's complement in a while, so I was hoping someone
could give me a hand.

Thanks

eliss

[Larry Bates]

eliss wrote:

I'm using dl.call() to call a C function in an external library. It's
working great so far except for one function, which returns an
unsigned int in the C version. However, in python it returns a signed
value to me. How can I get the unsigned value from this? I haven't
brushed up on my two's complement in a while, so I was hoping someone
could give me a hand.

Thanks

eliss

It is returning 32 bits. If the sign bit (bit 32) is on it appears as a
negative number. Test for negative and multiply the absolute value * 2.
That should get you the unsigned value you want in a long.

-Larry

[Diez B. Roggisch]

Larry Bates wrote:

eliss wrote:

>I'm using dl.call() to call a C function in an external library. It's
working great so far except for one function, which returns an
unsigned int in the C version. However, in python it returns a signed
value to me. How can I get the unsigned value from this? I haven't
brushed up on my two's complement in a while, so I was hoping someone
could give me a hand.

Thanks

eliss

It is returning 32 bits. If the sign bit (bit 32) is on it appears as a
negative number. Test for negative and multiply the absolute value * 2.
That should get you the unsigned value you want in a long.

Erm... Nope.

All bits set is -1 - so according to your recipe, that would be abs(-1) * 2
= 2

I'd suggest this formula:

if value < 0:
value = 2^32 + value + 1

Diez

[Larry Bates]

Diez B. Roggisch wrote:

Larry Bates wrote:

>eliss wrote:

>>I'm using dl.call() to call a C function in an external library. It's
working great so far except for one function, which returns an
unsigned int in the C version. However, in python it returns a signed
value to me. How can I get the unsigned value from this? I haven't
brushed up on my two's complement in a while, so I was hoping someone
could give me a hand.

Thanks

eliss

It is returning 32 bits. If the sign bit (bit 32) is on it appears as a
negative number. Test for negative and multiply the absolute value * 2.
That should get you the unsigned value you want in a long.

Erm... Nope.

All bits set is -1 - so according to your recipe, that would be abs(-1) * 2
= 2

I'd suggest this formula:

if value < 0:
value = 2^32 + value + 1

Diez

Thanks for the correction. You are of course correct.

-Larry

[eliss]

On Dec 11, 12:38 pm, Larry Bates <[e-mail]:

Diez B. Roggisch wrote:

Larry Bates wrote:

eliss wrote:
I'm using dl.call() to call a C function in an external library. It's
working great so far except for one function, which returns an
unsigned int in the C version. However, in python it returns a signed
value to me. How can I get the unsigned value from this? I haven't
brushed up on my two's complement in a while, so I was hoping someone
could give me a hand.

>Thanks

>eliss
It is returning 32 bits. If the sign bit (bit 32) is on it appears as a
negative number. Test for negative and multiply the absolute value * 2.
That should get you the unsigned value you want in a long.

Erm... Nope.

All bits set is -1 - so according to your recipe, that would be abs(-1) * 2
= 2

I'd suggest this formula:

if value < 0:
value = 2^32 + value + 1

Diez

Thanks for the correction. You are of course correct.

-Larry

Hi thanks for the responses but I'm afraid I don't see how either
formula works.

Lets say I get the return value of -5, which is 1011b when it should
be 11. Then according to the formula it would be 2^4-5+1=12

But it should be 11...

[eliss]

On Dec 11, 2:28 pm, eliss <[e-mail]:

On Dec 11, 12:38 pm, Larry Bates <[e-mail]:

Diez B. Roggisch wrote:

Larry Bates wrote:

>eliss wrote:
>>I'm using dl.call() to call a C function in an external library. It's
>>working great so far except for one function, which returns an
>>unsigned int in the C version. However, in python it returns a signed
>>value to me. How can I get the unsigned value from this? I haven't
>>brushed up on my two's complement in a while, so I was hoping someone
>>could give me a hand.

>>Thanks

>>eliss
>It is returning 32 bits. If the sign bit (bit 32) is on it appears as a
>negative number. Test for negative and multiply the absolute value * 2.
>That should get you the unsigned value you want in a long.

Erm... Nope.

All bits set is -1 - so according to your recipe, that would be abs(-1) * 2
= 2

I'd suggest this formula:

if value < 0:
value = 2^32 + value + 1

Diez

Thanks for the correction. You are of course correct.

-Larry

Hi thanks for the responses but I'm afraid I don't see how either
formula works.

Lets say I get the return value of -5, which is 1011b when it should
be 11. Then according to the formula it would be 2^4-5+1=12

But it should be 11...

Seems like the simple formula of:

if value < 0:
value = 2^32 + value

might just work. Thanks :)

[Gabriel Genellina]

En Tue, 11 Dec 2007 19:28:34 -0300, eliss <el***********@gmail.com>
escribi�:

On Dec 11, 12:38 pm, Larry Bates <[e-mail]:

>Diez B. Roggisch wrote:

Larry Bates wrote:
eliss wrote:
working great so far except for one function, which returns an
unsigned int in the C version. However, in python it returns a

signed

>>value to me. How can I get the unsigned value from this? I haven't

I'd suggest this formula:

if value < 0:
value = 2^32 + value + 1

Hi thanks for the responses but I'm afraid I don't see how either
formula works.

Lets say I get the return value of -5, which is 1011b when it should
be 11. Then according to the formula it would be 2^4-5+1=12

But it should be 11...

Yes, both formulae were wrong, omit the "+1" in the last one.
Another way is to use a bitwise and (&) with a number whose bits are all
1's.
For 4 bits (your example), you need 1111b = 0xF

py-5 & 0xF
11

For a 32bit number, you have to use x & 0xFFFFFFFF (or 0xFFFFFFFFL on
older Python versions)

>>-5 & 0xffffffff

4294967291L

--
Gabriel Genellina

[MRAB]

On Dec 11, 10:51 pm, eliss <[e-mail]:

On Dec 11, 2:28 pm, eliss <[e-mail]:

On Dec 11, 12:38 pm, Larry Bates <[e-mail]:

Diez B. Roggisch wrote:
Larry Bates wrote:

eliss wrote:
>I'm using dl.call() to call a C function in an external library. It's
>working great so far except for one function, which returns an
>unsigned int in the C version. However, in python it returns a signed
>value to me. How can I get the unsigned value from this? I haven't
>brushed up on my two's complement in a while, so I was hoping someone
>could give me a hand.

>Thanks

>eliss
It is returning 32 bits. If the sign bit (bit 32) is on it appears as a
negative number. Test for negative and multiply the absolute value * 2.
That should get you the unsigned value you want in a long.

Erm... Nope.

All bits set is -1 - so according to your recipe, that would be abs(-1) * 2
= 2

I'd suggest this formula:

if value < 0:
value = 2^32 + value + 1

Diez

Thanks for the correction. You are of course correct.

-Larry

Hi thanks for the responses but I'm afraid I don't see how either
formula works.

Lets say I get the return value of -5, which is 1011b when it should
be 11. Then according to the formula it would be 2^4-5+1=12

But it should be 11...

Seems like the simple formula of:

if value < 0:
value = 2^32 + value

might just work. Thanks :)

You're working in Python so that should be:

if value < 0:
value = 2 ** 32 + value

because ^ is exclusive-or. :-)