+ + if len(assoc_info) < 1: + ret_val, linux_cs = check_sblim(server, virt) + if ret_val != PASS: + status = FAIL + raise Exception(" '%s' returned (%d) '%s' objects" % \ + (len(assoc_info), reg_classname)) + else: + status = XFAIL_RC(bug_sblim) + raise Exception("Known failure") + + for inst in assoc_info: + status, verify_ectp_list = verify_fields(inst, verify_ectp_list) + if status != PASS: + raise Exception("Failed to verify instance") + + if status == PASS: + for k, l in verify_ectp_list.iteritems(): + if len(l) != 0: + status = FAIL + raise Exception("%s items weren't returned: %s" % (k, l)) + + except Exception, detail: + logger.error("Exception: %s" % detail) + if status == PASS: + status = FAIL
No need to check status == PASS here, can directly assign status = FAIL. otherwise looks good , +1 for me.
I did this because if an XFAIL is thrown, we overwrite the failure type as FAIL. But I agree - it's an ugly way of preserving the XFAIL status. I'll resend with a fix for this. -- Kaitlin Rupert IBM Linux Technology Center kaitlin@linux.vnet.ibm.com