Hacker News Reader

Comment by js2 | Hacker News Reader

The biggest issue with "set -e" is that it's inconsistent. It doesn't always cause the shell to exit where you'd expect it to, and in some cases, it causes the shell to exit where you would NOT expect it to. This is the most surprising:

    set -e
    i=0
    (( i++ ))  # shell will exit here

Still, I tend to use it because even though it doesn't handle every case, I'd still usually rather my scripts exit when a command fails.

View on HN

Another variant which I overlooked to often is if you have a construct which returns an error but does not exit the script later move it in a function and now it does exit the script.

  set -e
  [[ 0 == 1 ]] && true
  echo Status: $?

will print "Status: 1" but

  set -e
  func() {
    [[ 0 == 1 ]] && true
  }
  func
  echo Status: $?

will print nothing but exit after func returns.

There's quite a few situations where it does NOT exit when you'd expect it to. This is one of those. What you're seeing there is a variant of the "Short-Circuiting Considerations" discussed here:

http://redsymbol.net/articles/unofficial-bash-strict-mode/

Basically, when you short circuit, it will not cause the script to exit. However, the short circuit still affects the return value of the function, or the script itself if it's the last line of the script.

So your short circuit is causing the function to return non-zero, which then causes the script to exit.

This doesn't cause an exit for me on GNU bash (3.2.57)

The change was made with bash-4.1:

  This is a terse description of the new features added
  to bash-4.1 since the release of bash-4.0. [...]

  j. The [[ and (( commands are now subject to the setting
     of `set -e' and the ERR trap."

https://git.savannah.gnu.org/cgit/bash.git/tree/NEWS#n1078

Test program:

  $ cat foo.sh
  #!/bin/bash
  set -eux
  
  echo "$BASH_VERSION"
  
  i=0
  
  (( i++ ))
  echo "?=$? i=$i"
  
  (( i++ ))
  echo "?=$? i=$i"

Running with 3.2.57:

  $ /bin/bash foo.sh
  + echo '3.2.57(1)-release'
  3.2.57(1)-release
  + i=0
  + ((  i++  ))
  + echo '?=1 i=1'
  ?=1 i=1
  + ((  i++  ))
  + echo '?=0 i=2'
  ?=0 i=2

Running with 5.1.16:

  $ /opt/homebrew/bin/bash foo.sh
  + echo '5.1.16(1)-release'
  5.1.16(1)-release
  + i=0
  + ((  i++  ))

So the arithmetic operation returns non-zero in both cases where it evaluates to 0, but only under bash >= 4.1 does it cause bash to terminate.

The moral of the story is to be careful with arithmetic operations that might evaluate to zero, or use this:

(( operation )) || true

I copy/pasted the example from the Google bash scripting guide because I was on mobile earlier, but I just confirmed that it exits immediately after the arithmetic operation with bash 5.1.12.

Reply on news.ycombinator.com