I am using set -e
to stop execution of a script on first error.
The problem is that this does not tell me what went wrong.
How can update a bash scr
Instead of set -e
, use an ERR trap; you can pass $BASH_LINENO
in to get the specific line number on which the error occurred. I provide a script taking advantage of this in my answer at https://stackoverflow.com/a/185900/14122
To summarize:
error() {
local sourcefile=$1
local lineno=$2
# ...logic for reporting an error at line $lineno
# of file $sourcefile goes here...
}
trap 'error "${BASH_SOURCE}" "${LINENO}"' ERR
make err.sh
set -e
trap 'echo "ERROR: $BASH_SOURCE:$LINENO $BASH_COMMAND" >&2' ERR
include it (. err.sh
) in all your scripts.
replace any
... | while read X ; do ... ; done
with
while read X ; do ... ; done < <( ... )
in your scripts for the trap to give the correct line number/command in the error message
You can't use set -e
by itself because processing will immediately stop after any error. Take a look at the Set Builtin section of the Bash Reference Manual for more information about the -x and -v options, which you can use for debugging.
Something like:
set -e
set -v
will exit on any error, while showing you each input line as it is read. It will not, however, show you just the line with the error. For that, you will need to do your own explicit error checking.
For example:
set +e
if false; then
real_exit_status=$?
echo 'Some useful error message.' >&2
exit $real_exit_status
fi
set -ex
will show (all) lines as they are executed and stop at the first command returning nonzero (not as part of if/while/until
constructs).
Have you tried with --verbose
?
bash --verbose script.sh