How do you create a single or double float from a Lisp array of bytes? Thanks, Joel -------------------------------------------------------------------------- - for hire: mac osx device driver ninja, kernel extensions and usb drivers ---------------------+------------+--------------------------------------- http://wagerlabs.com | @wagerlabs | http://www.linkedin.com/in/joelreymont ---------------------+------------+---------------------------------------
Joel Reymont wrote:
> How do you create a single or double float from a Lisp array of bytes?
>
There may be an easier method (i.e., using undocumented functions), but
this seems to work for me assuming IEEE 754 on a little-endian arch:
(defun unsigned-bytes-to-float (bytes)
#-ieee-floating-point (error "fixme")
(let ((ft (ecase (length bytes)
(4 :lisp-single-float)
(8 :lisp-double-float))))
(fli:with-dynamic-foreign-objects
((x (:unsigned :byte)
:initial-contents (coerce bytes 'list)
:nelems (length bytes)))
(fli:with-coerced-pointer
(y :type ft) x
(fli:dereference y)))))
CL-USER 361 > (= pi (unsigned-bytes-to-float (reverse #(#x40 #x09 #x21
#xfb #x54 #x44 #x2d #x18))))
T
Mike
Hi Joel, On 26 sept. 2011, at 21:32, Joel Reymont wrote: > How do you create a single or double float from a Lisp array of bytes? You could try the ieee-floats library: http://common-lisp.net/project/ieee-floats/ Works well for me. Best Regards, Cam