hi,

that would be a nice feature. My testsuite basically depends on wether if the first test finishes succesfully. If this is not the case then other tests may fail too. Of course one could add the first test in the first bats file and call the other bats files only if the first test works out.. but well.

One way to workaround that missing option would be to kill $PPID (bats) if one test condition
fails. Then the entire run exits after the failed function. Its not very clean but seems to work,
and killing bats does not tend to leave any temporary files around too.
Example:

checkstate() {
    if [ $1 == 0 ]; then
        return 0
    else
        kill $PPID
        return $status
    fi
}
@test "foo_fails" {
    run cat /etc/nonexistant
    checkstate $status
}
@test "bar" {
    run cat /etc/fstab
    checkstate $status
}
@test "baz" {
    run cat /etc/fstab
}

1..3
not ok 1 foo_fails
# (from function `checkstate' in file test.batss, line 6,
#  in test file test.batss, line 11)
Terminated
#   `checkstate $status' failed
# bats warning: Executed 1 instead of expected 3 tests

Maybe someone has a better solution, but this one would allow to have certain tests fail the complete suite and others not, depending on wether the test is using the checkstate function to check return of the run command. In case multiple bats files are used one can have a common special function included via source, that terminates PPID

Add an option to fail the entire run as soon as first test fails #209

Description

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions