I am using a preflight bash script in packagemaker :
run cp -pf "/folder/to/my/db" "/Library/Application Support/app/db
The run function (that I found on StackOverflow by the way) :
run() { $*; code=$?; [ $code -ne 0 ] && echo "command [$*] failed with error code $code\nERROR: $@\n"; }
The command cp returns a 64 code. What is this 64 status please?
How can I resolve that?
The problem is that you don't have a folder Support/app/db for the command to copy files /folder/to/my/db and /Library/Application to.
Replace the misguided (almost always wrong) $* with the correct "$@":
run()
{
"$@"
code=$?
[ $code -ne 0 ] && echo "command [$*] failed with error code $code\nERROR: $@\n"
}
Plain $* breaks words at spaces; "$@" preserves spaces in arguments. Most often, $* is not the right notation (though it would be fine in the echo where you used $@). It isn't clear to me why the command's arguments are being listed twice in the error message.
The error reporting would be improved by adding >&2 to the end to redirect the output to standard error, which is where error messages belong. (I'd remove the repetition while I'm at it.) Note that using $* inside the argument to echo is entirely appropriate.
[ $code -ne 0 ] && echo "command [$*] failed with error code $code" >&2
In fact, the run() function can be simplified still more; the variable code really isn't needed:
run()
{
"$@" || echo "command [$*] failed with error code $?" >&2
}
If you want the script to exit too, then you can use:
run()
{
"$@" || { echo "command [$*] failed with error code $?" >&2; exit 1; }
}
The { ...; } notation treats the commands within as a unit for I/O redirection without starting a sub-shell.
See also How to iterate over arguments in a Bash script.
