On Thu, 10 Jul 2025 at 01:36, Chet Ramey <chet.ra...@case.edu> wrote:
> If the expression hadn't contained $x, nothing would have been expanded,
> because expanding a word without any expansion characters is a waste of
> time. This arguably causes a problem in this case; it would have been
> more consistent to run the whole string through word expansion and quote
> the subscript in all cases.
>
Or alternatively, always defer expansions, so that such generated quoting
is unnecessary.
It's all very well to say "leave out the '$'", but that can't work when the
key for a hash itself has to be computed.
Compare:
$ declare -A hash=() foo=( [bar]=4 ) ; declare -a keys=( [0]=humbug
[4]=hokeypokey )
$ i= ; (( i=foo[bar], hash[${keys[i]}]++ ))
$ declare -p hash
declare -A hash=([*humbug*]="1" )
$ declare -A hash=() foo=( [bar]=4 ) ; declare -a keys=( [0]=humbug
[4]=hokeypokey )
$ i= ; (( hash[${keys[i=foo[bar]]}]++ ))
$ declare -p hash
declare -A hash=([*hokeypokey*]="1" )
Advanced users hopefully understand that there are two numeric contexts
that are evaluated at different times, despite being apparently nested one
within the other, but it's hardly obvious to a new user, even after they've
read the manual.
It would be more useful - and less confusing - if both produced the second
answer.
Sure, this wouldn't be exactly compatible with previous behaviour, but then
neither is the current behaviour.
-Martin
