NBardelot
commented
7 years ago Note:
For the moment in order to achieve the same goal I have to use the following wrapper script:
#!/bin/bash
clean_exit() {
cat "$1"
/bin/rm -f "$ERR_FILE" "$OUT_FILE"
exit "$2"
}
OUT_FILE=$(mktemp)
ERR_FILE=$(mktemp)
bats $@ 1>"$OUT_FILE" 2>"$ERR_FILE"
RETVAL=$?
if [ $RETVAL -ne 0 ]
then
if [ $(wc -l "$ERR_FILE" | awk '{ print $1 }') -gt 0 ]
then
clean_exit "$ERR_FILE" 1 # Abnormal error (probably a syntax error)!
else
clean_exit "$OUT_FILE" 2 # One or more tests fail
fi
fi
clean_exit "$OUT_FILE" 0
Potherca
When one or more tests fail at the moment the bats command exits with an error code of 1. If there is a syntax error in the bats file, the behaviour is the same.
The feature would consist of a flag (for example "-i|--ignore-failures"), so that the exit code is 0 even if some tests fail, allowing the caller to differenciate between abnormal behaviour (exit code > 0) and tests failures (exit code = 0, but number of failures > 0 as printed in stdout).