Re: [Csnd-dev] new parser

it seems that the new parser is clever enough to look at the output 
type.  of course it would be welcome if someone with insights could confirm.

i compared this simple instrument:

instr 1
   aSig = oscili(0.2,400)
   kSig = oscili(10,0.5)
endin

when running with -v i get this in Csound 6 (showing that both lines use 
the a-rate version of oscili):
instr 1 allocated at 0x559ea703b550
	lclbas 0x559ea703b6e8, opds 0x559ea703b748
op (oscili.a) allocated at 0x559ea703b748
opadr = 0x7f4e813166c7
op (oscili.a) allocated at 0x559ea703fa90
opadr = 0x7f4e813166c7
op (=.down) allocated at 0x559ea7043dd8
opadr = 0x7f4e8132af96
insert(): tp->act_instance = 0x559ea703b550
psave beg at 0x559ea703b6c0
    ending at 0x559ea703b6c0
init oscili.a:
init oscili.a:
init =.down:

when running the same with csound 7 the parser distinguishes between .a 
and .k for both cases:
instr 1 allocated at 0x565430a18020
         lclbas 0x565430a181d8, opds 0x565430a18240
op (oscili.a) allocated at 0x565430a18240
oscili.a init = 0x7f72367788ba
oscili.a perf = 0x7f7236779985
op (oscili.kk) allocated at 0x565430a1c5b8
oscili.kk init = 0x7f72367788ba
oscili.kk perf = 0x7f723677981c
insert(): tp->act_instance = 0x565430a18020
psave beg at 0x565430a181b0
    ending at 0x565430a181b0
init oscili.a:
init oscili.kk:




On 17/12/2024 20:51, joachim heintz wrote:
> for functional style of code writing, we need the type specifier to 
> choose the appropriate version of an opcode, like
> aSig = oscili:a(.2,200)
> versus
> kSig = oscili:k(10,1)
> 
> is this still necessary in CS7?  or can the new parser "look" at the 
> output and see that here is a-rate necessary, but there k-rate?
> 
> cheers -
>      joachim