On Thu, Oct 27, 2016 at 10:02 AM, Peter Eisentraut
<peter.eisentraut@2ndquadrant.com> wrote:
> Here is a series of patches to add NLS to the remaining bin programs,
> which were moved from contrib a while ago. (If you're missing pgbench,
> I'm skipping that for now because it's more complicated.) I'll add this
> to the commit fest.
I have been looking at this patch set, and that's definitely a good
idea to do this change.
1) 0001 for pg_archivecleanup is missing nothing.
2) For 0002 and pg_test_fsync, I am seeing a missing entry: printf(NA_FORMAT, "n/a*\n");
3) For pg_test_timing, the doc changes could be into a separate
change, but that's fine to group them as well. I am seeing no missing
strings for translations.
4) 0004 and pg_upgrade... In check.c, three places like that: if (!db_used) {
fprintf(script,"Database: %s\n", active_db->db_name); db_used = true; }
In exec.c:
#endif fprintf(log, "command: %s\n", cmd);
#ifdef WIN32
+GETTEXT_FLAGS = \
+ pg_fatal:1:c-format \
+ pg_log:2:c-format \
+ prep_status:1:c-format \
+ report_stats:2:c-forma
s/report_stats/report_status/
In info.c, missing some entries in report_unmatched_relation() when
reporting unmatching relations?
In parseCommandLine() of option.c, the "pg_upgrade run on" string
needs to be handled.
In util.c, doesn't pg_log_v() need to handle strings used in fprintf?
In version.c, this one: if (!db_used) { fprintf(script, "Database: %s\n",
active_db->db_name); db_used = true; }
5) 0005 and pg_xlogdump, I am not seeing a missing entry.
--
Michael
