(re-sending the message for the second time without a list of rules,
which caused posting to bounce at mx1.eu.apache.org: due to hitting
[can't name the hitting rules to avoid hitting them again])
I'm going through tests which report "skipped: (no reason given)",
checking to see what it takes to enable them.
Just got across t/rule_names.t, which looks abandoned and
suffered from several problems. I think I fixed them all now:
$ svn ci -m 't/rule_names.t : missing boilerplate "use SATest",
allow_user_rules=1, too early call to Test::plan;
fix tainted PATH in SATest::probably_unused_spamd_port()'
Sending t/SATest.pm
Sending t/rule_names.t
Transmitting file data ..
Committed revision 1428892.
The test is off by default, needs to be explicitly enabled in t/config:
# rule_names.t fails continually due to changes in the rulesrc repository;
# TODO: we need a better way to test this.
run_rule_name_tests=y
It reports 18 warnings:
Found anti-pattern: P_116 =
[...]
# Failed test 3140 in SATest.pm at line 762 fail #18
Output can be examined in: log/d.rule_names/1
Note: rule_name failures may be only cosmetic
but must be fixed before release
Failed 18/7834 subtests
Test Summary Report
-------------------
t/rule_names.t (Wstat: 0 Tests: 7834 Failed: 18)
Failed tests: 116, 200, 290, 301-302, 602, 604, 794, 2514
2531, 2633, 2750, 2796, 2818, 2913, 2956
3101, 3140
Files=1, Tests=7834, 56 wallclock secs ( 0.97 usr 0.09 sys + 54.93 cusr 0.37
csys = 56.36 CPU)
Result: FAIL
It's probably no big deal, but I wonder what is
the purpose of this test???
It is looking for spammy-looking names of the rules
or something else???
[P.S. now after trying to send this posting for the second
time, I'm beginning to understand a possible rationale
for this test]
Mark
