Hello Rikki, Eric, * Eric Blake wrote on Tue, Feb 10, 2009 at 01:58:43PM CET: > > AT_XFAIL_IF([AT_EXPR_RESPONSE_TEST_P([test.1], ['"i,j&i<j"'], [test.1c])]) > > According to Rikki McQueary on 2/9/2009 3:20 PM: > > > > My question is: Is this correct? And also in the output instead of the > > test passing with an ok is there a way to signify the test failed but > > was expected to fail? > > Yes, other than the quoting, that is the correct way to mark a test as > conditionally failing.
One should maybe add that the argument to AT_XFAIL_IF is evaluated before the test group is started. This is arguably a limitation or a bug in Autotest, but anyway needs to be taken into account. Cheers, Ralf -- I've lost an unknown number of emails due to some server trouble. If you've written to me during the last week, consider resending. _______________________________________________ Autoconf mailing list Autoconf@gnu.org http://lists.gnu.org/mailman/listinfo/autoconf
