msgtime: avoid obviously out-of-range dates (for now)

Wacky dates show up in lore for valid messages.  Lets ignore
them and let future generations deal with Y10K and time-travel
problems.

1 files changed, 5 insertions, 1 deletions

diff --git a/lib/PublicInbox/MsgTime.pm b/lib/PublicInbox/MsgTime.pm
index 479aaa4e..9f432644 100644
--- a/lib/PublicInbox/MsgTime.pm
+++ b/lib/PublicInbox/MsgTime.pm
@@ -38,7 +38,7 @@ sub str2date_zone ($) {
         if ($date =~ /(?:[A-Za-z]+,?\s+)? # day-of-week
                         ([0-9]+),?\s+  # dd
                         ([A-Za-z]+)\s+ # mon
-                        ([0-9]{2,})\s+ # YYYY or YY (or YYY :P)
+                        ([0-9]{2,4})\s+ # YYYY or YY (or YYY :P)
                         ([0-9]+)[:\.] # HH:
                                 ((?:[0-9]{2})|(?:\s?[0-9])) # MM
                                 (?:[:\.]((?:[0-9]{2})|(?:\s?[0-9])))? # :SS
@@ -67,6 +67,10 @@ sub str2date_zone ($) {
 
                 $ts = timegm($ss // 0, $mm, $hh, $dd, $mon, $yyyy);
 
+                # 4-digit dates in non-spam from 1900s and 1910s exist in
+                # lore archives
+                return if $ts < 0;
+
                 # Compute the time offset from [+-]HHMM
                 $tz //= 0;
                 my ($tz_hh, $tz_mm);