[squeak-dev] 32 bits single precision float number to a byte array

Hi all,

i'm looking for a way to convert a float (seen as a 32 bits single
precision float number) as a byte array of length 4 ?

Thank you,
--
Serge Stinckwich
UMI UMMISCO 209 (IRD/UPMC), Hanoi, Vietnam
Smalltalkers do: [:it | All with: Class, (And love: it)]
http://blog.doesnotunderstand.org/

Serge Stinckwich wrote:
> i'm looking for a way to convert a float (seen as a 32 bits single
> precision float number) as a byte array of length 4 ?

Given that there is no builtin 32bit single precision float type in
Squeak, how is this float stored? If it's in a float array you can just
get the bits out of it using #basicAt: and store that. E.g.,

   fltArray := FloatArray with: 42.0.
   word := fltArray basicAt: 1.
   bytes := (ByteArray new: 4) unsignedLongAt: 1 put: word; yourself.

Cheers,
   - Andreas

Andreas Raab a écrit :
There is a primitive failed in ByteArray>>integerAt:put:size:signed with
  your solution ...

Seems to work if i add bigEndian:true :

fltArray := FloatArray with: 42.0.
word := fltArray basicAt: 1.
bytes := (ByteArray new: 4) unsignedLongAt: 1 put: word bigEndian:true;
yourself.

Cheers,
--
Serge Stinckwich
UMI UMMISCO 209 (IRD/UPMC), Hanoi, Vietnam
Smalltalkers do: [:it | All with: Class, (And love: it)]
http://doesnotunderstand.org/

Serge Stinckwich wrote:
> There is a primitive failed in ByteArray>>integerAt:put:size:signed with
>  your solution ...

Ah, yes. This is the platform specific access for FFI and friends.

> Seems to work if i add bigEndian:true :

You're right. That is the correct solution.

Cheers,
   - Andreas

> fltArray := FloatArray with: 42.0.
> word := fltArray basicAt: 1.
> bytes := (ByteArray new: 4) unsignedLongAt: 1 put: word bigEndian:true;
> yourself.
>
> Cheers,

On Thu, Feb 26, 2009 at 01:15:36AM -0800, Andreas Raab wrote:
Note also that this does not work on a 64 bit host (e.g. 64 bit Linux).

Dave