Re: playing with IFS

On 2026-04-12 16:26, dana wrote:
> On Sun 12 Apr 2026, at 18:10, Ray Andrews wrote:
>> I hear you.  But as often as not I've made some simple syntax error
>> that's reparable without having to dig into the agenda
> you often rewrite or cut parts out of what you're actually doing so even
> finding syntax errors requires digging.
Believe it or not, it seems polite to cut the problem down to the 
minimal specimen.  But of course that can be taken too far and I end up 
showing something that misses the crux of the problem cuz I 
misunderstand the problem.  In this case I had both the illegal space -- 
sometimes they're optional, sometimes they're mandatory and sometimes 
they are verboten -- and a stupid error:

BTW, here's the actual thing:

[ "$nospaces" ] && _execute_output=( ${{f}"$( sed -r 
's/(\x1b\[K[:-]\x1b\[m\x1b\[K)[[:space:]]*/\1 /g' <<< 
${(F)_execute_output} )"} )

Yup, it throws an error alright.  This is rather better:

[ "$nospaces" ] && _execute_output=( ${(f)"$( sed -r 
's/(\x1b\[K[:-]\x1b\[m\x1b\[K)[[:space:]]*/\1 /g' <<< 
${(F)_execute_output} )"} )

BTW Mark, I trust that sed can't be done native.  Fantastic if it could 
be tho.

Oh ... as to speed, 20% improvement, but in the real world microscopic.  
But on principle I'd avoid the pipe if possible.

Thank you gentlemen