On 03/23/2014 09:22 AM, Ralf Hemmecke wrote:
> Thanks for your answer, but writing just "z:=" or "z:% :=" *makes* a
> difference on my computer. Without the ":%" the build will hang during
> doc build. So I guess the compiler generates different code.
And indeed, it does:
with "z := ..."
(SDEFUN |INT;exquo;2$U;50| ((|x| $) (|y| $) ($ |Union| $ "failed"))
(SPROG ((|z| NIL))
(SEQ
(COND ((ZEROP |y|) (CONS 1 "failed"))
(#1='T
(SEQ (LETT |z| (INTEXQUO |x| |y|) |INT;exquo;2$U;50|)
(EXIT
(COND ((NULL |z|) (CONS 1 "failed"))
(#1# |z|)))))))))
with "z: % := ..."
(SDEFUN |INT;exquo;2$U;50| ((|x| $) (|y| $) ($ |Union| $ "failed"))
(SPROG ((|z| ($)))
(SEQ
(COND ((ZEROP |y|) (CONS 1 "failed"))
(#1='T
(SEQ (LETT |z| (INTEXQUO |x| |y|) |INT;exquo;2$U;50|)
(EXIT
(COND ((NULL |z|) (CONS 1 "failed"))
(#1# (CONS 0 |z|))))))))))
Main difference is in the last line.
Looks to me as if the compiler does not automatically add the missing
type information ":%" to z although the "pretend %" on the right hand
side makes it clear what type the left hand side is supposed to have.
It wouldn't be a problem for me to *require* explicit type declaration
for the first type a variable appears, but then the compiler should
shout when a type declaration is missing and not silently produce bad code.
Ralf
--
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra system" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/fricas-devel.
For more options, visit https://groups.google.com/d/optout.
