Denton Liu <liu.denton@xxxxxxxxx> writes: > test_expect_success 'open-quoted pathname' ' > echo "\"a test=a" >.gitattributes && > - test_must_fail attr_check a a > + attr_check a unspecified > ' I offhand do not think of a reason why "git check-attr" must be given freedom to choose from more than one answers (which is what the original test that said "as long as it does not report 'a', we are happy" allowed the command to), so it probably is OK to make the tests more strict like this patch does. Thanks.
