Fix 32-bit warnings in timelib

This has already been fixed upstream in
https://github.com/derickr/timelib/commit/4df262e7585b2ad2c6b01a23190974f9aea298ed
so I'm backporting this change.

diff --git a/ext/date/lib/parse_tz.c b/ext/date/lib/parse_tz.c
index c8e501151f43b58aee4811b92ed203ec0bebb7ae..65a4d0692fd7aece0be03809f02eb22ef90a95d0 100644 (file)
--- a/ext/date/lib/parse_tz.c
+++ b/ext/date/lib/parse_tz.c
@@ -393,7 +393,7 @@ void timelib_dump_tzinfo(timelib_tzinfo *tz)
                tz->type[0].isgmtcnt
                );
        for (i = 0; i < tz->bit64.timecnt; i++) {
-               printf ("%016lX (%20ld) = %3d [%5ld %1d %3d '%s' (%d,%d)]\n",
+               printf ("%016" PRIX64 " (%20" PRId64 ") = %3d [%5ld %1d %3d '%s' (%d,%d)]\n",
                        tz->trans[i], tz->trans[i], tz->trans_idx[i],
                        (long int) tz->type[tz->trans_idx[i]].offset,
                        tz->type[tz->trans_idx[i]].isdst,
@@ -404,7 +404,7 @@ void timelib_dump_tzinfo(timelib_tzinfo *tz)
                        );
        }
        for (i = 0; i < tz->bit64.leapcnt; i++) {
-               printf ("%016lX (%20ld) = %d\n",
+               printf ("%016" PRIX64 " (%20ld) = %d\n",
                        tz->leap_times[i].trans,
                        (long) tz->leap_times[i].trans,
                        tz->leap_times[i].offset);