Gilles Detillieux (firstname.lastname@example.org)
Mon, 8 Feb 1999 13:53:12 -0600 (CST)
According to Geoff Hutchison:
> 2) I can't reproduce the bus error in Deserialize on Solaris. I tried on
> two boxes and it compiled cleanly (after fixing a typo in Soundex.cc) and
> ran without problems. I also didn't experience the segfault (!) the same
> user saw in the call to strftime.
That problem really sounds like a corrupt database to me. We really ought
to fix Deserialize to check that it doesn't go past the end of its input
string. But, what do we do if it does? If we silently ignore the error,
then the user won't find out right away that the database is corrupt.
There's already a test & error message for bad tags, so I'd say do the
same for a bad length or missing data. I may take a stab at it after my
meeting this afternoon, but I can't promise.
> That said, I'll roll a snapshot as another pre-release. Let's beat on it
> over the weekend and if there's nothing obviously wrong, it's going out on
> Monday evening as 3.1.0. If other people could try running on Solaris or
> Digital Unix, I'd appreciate it.
I didn't see this message until today, but I was running the 020399
snapshot over the weekend without any trouble. The latest snapshot
compiles fine, but I have yet to beat on it.
If it's not too late, I'd like to slip in this little tweak to the new
parse_date function in htnotify. It will recognize an ISO 8601 date
(yyyy-mm-dd) even if iso_8601 isn't set, as long as a four-digit year
is given. Without this fix, it would treat ambiguous cases as
yyyy-dd-mm instead. I hope to get a documentation patch to you before
the end of the day as well, explaining date formats.
--- htdig-3.1.0-020799/htnotify/htnotify.cc.gilles Fri Feb 5 19:17:47 1999
+++ htdig-3.1.0-020799/htnotify/htnotify.cc Mon Feb 8 13:09:57 1999
@@ -319,6 +319,11 @@
// Default to American standard when not specified in conf,
// so expect mm dd [yy]yy.
sscanf(scandate.get(), "%d%d%d", &mm, &dd, &yy);
+ if (mm > 31 && dd <= 12 && yy <= 31)
+ // probably got yyyy-mm-dd instead of mm/dd/yy
+ t = mm; mm = dd; dd = yy; yy = t;
// OK, we took our best guess at the order the y, m & d should be.
-- Gilles R. Detillieux E-mail: <email@example.com> Spinal Cord Research Centre WWW: http://www.scrc.umanitoba.ca/~grdetil Dept. Physiology, U. of Manitoba Phone: (204)789-3766 Winnipeg, MB R3E 3J7 (Canada) Fax: (204)789-3930 ------------------------------------ To unsubscribe from the htdig3-dev mailing list, send a message to firstname.lastname@example.org containing the single word "unsubscribe" in the SUBJECT of the message.
This archive was generated by hypermail 2.0b3 on Mon Feb 08 1999 - 12:15:47 PST