This way tests continue to run after one test script does not report success. ---
I failed to notice this when MANUALLY testing the test change in question. This is quick fix and makes notmuch-test & test-lib.sh have interdependency how to modify the test script name for output file. I'll think whether there is something to be made about this, meanwhile I'm open to suggestions to that. too. test/notmuch-test | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/test/notmuch-test b/test/notmuch-test index b843712..ecf81a9 100755 --- a/test/notmuch-test +++ b/test/notmuch-test @@ -32,12 +32,14 @@ fi trap 'e=$?; kill $!; exit $e' HUP INT TERM # Run the tests for test in $TESTS; do + this_test=${test%.sh} + this_test=${this_test#T[0-9][0-9][0-9]-} $TEST_TIMEOUT_CMD ./$test "$@" & wait $! # If the test failed without producing results, then it aborted, # so we should abort, too. RES=$? - if [[ $RES != 0 && ! -e "test-results/${test%.sh}" ]]; then + if [[ $RES != 0 && ! -e "test-results/$this_test" ]]; then exit $RES fi done -- 1.8.0