local/non-local and sub shell exit codes

local/non-local and sub shell exit codes

[René Neumann]

Hi all,

I just encountered this strange behavior under zsh-5.5.1:

% foo=$(false); echo $?
1
% local foo=$(false); echo $?
0

Is this intended? As I'm pretty sure that this behavior is new...

Best regards,
René

Re: local/non-local and sub shell exit codes

[René Neumann]

Ok, I have to take it back. Found the following in the manpage (after
discovering, that 'local' is an alias for 'typeset'):

---8<---
Unlike parameter assignment statements, typeset's exit status on an
assignment that involves a command substitution does not reflect the
exit status of the command substitution. Therefore, to test for an error
in a command substitution, separate the declaration of the parameter
from its initialization:
  # WRONG
  typeset var1=$(exit 1) || echo "Trouble with var1"

  # RIGHT
  typeset var1 && var1=$(exit 1) || echo "Trouble with var1"
--->8---

- René

Am 10.06.2018 um 15:18 schrieb René Neumann:
> Hi all,
> 
> I just encountered this strange behavior under zsh-5.5.1:
> 
> % foo=$(false); echo $?
> 1
> % local foo=$(false); echo $?
> 0
> 
> Is this intended? As I'm pretty sure that this behavior is new...
> 
> Best regards,
> René
> 

Re: local/non-local and sub shell exit codes

[dana]

On 10 Jun 2018, at 08:34, René Neumann <lists@necoro.eu> wrote:
>Ok, I have to take it back. Found the following in the manpage (after
>discovering, that 'local' is an alias for 'typeset'):

FYI, although that note was only added in 2016*, the behaviour is definitely
older than that. That's how it works in bash as well, and shellcheck warns you
about it by default.

* A result of this discussion: http://www.zsh.org/mla/workers/2016/msg00140.html

dana