On Fri, Jan 04, 2019 at 12:13:20AM +0200, Adrian Bunk wrote:
> The failures all make sense if the result is
> $(currentyear)-11-03 09:23:38.000
> and this is compared with
> 2018-11-03 09:23:38.000
>
> The upstream fix might not work for the complete lifetime of buster...
> https://github.com/tstack/lnav/commit/7dfa0e9dd4dfe2bf97c413c6c5dc03d42aeebc96#diff-e12cb0b1f4ccdad5b40998150689c681
Good catch. So it would be like a time-bomb.
I would just disable the test and forget about it until upstream
has a (really) long term solution.
(Trivial but untested) patch follows:
--- a/test/test_sql.sh
+++ b/test/test_sql.sh
@@ -786,17 +786,6 @@ EOF
cat ${test_dir}/logfile_syslog.0 | run_test ${lnav_test} -n \
- -c ";select log_time from syslog_log where log_procname = 'automount'"
-
-check_output "querying against stdin is not working?" <<EOF
- log_time
-2018-11-03 09:23:38.000
-2018-11-03 09:23:38.000
-2018-11-03 09:23:38.000
-EOF
-
-
-cat ${test_dir}/logfile_syslog.0 | run_test ${lnav_test} -n \
-c ";select log_time from syslog_log where log_procname = 'sudo'"
check_output "single result is not working?" <<EOF
