Control: forwarded 503658 http://savannah.gnu.org/bugs/index.php?29391
Control: tags 503658 = l10n
Control: found 503658 2.12-2
Control: retitle 503658 Pathetically slow egrep with --ignore-case and UTF8

I'm pretty sure this is actually another bug, since the original forwarded has been long fixed in grep upstream and I can still see that issue in Wheezy. This is most likely http://savannah.gnu.org/bugs/index.php?29391 (-i and utf8 slowness, speedup idea) which apparently does not yet have a committed fix.

postmaster@lxmhs62:~$ time grep -i stoer...@domain.de /var/log/mail.log

real    0m22.063s
user    0m21.941s
sys     0m0.068s
postmaster@lxmhs62:~$ time LC_ALL=C grep -i stoer...@domain.de /var/log/mail.log

real    0m1.635s
user    0m1.580s
sys     0m0.056s
postmaster@lxmhs62:~$ export | grep LC
declare -x LC_COLLATE="de_DE.UTF-8"
declare -x LC_CTYPE="de_DE.UTF-8"
declare -x LC_MEASUREMENT="de_DE.utf8"
declare -x LC_MESSAGES="de_DE.UTF-8"
declare -x LC_MONETARY="de_DE.utf8"
declare -x LC_NUMERIC="de_DE.utf8"
declare -x LC_TIME="de_DE.utf8"

This only happens with "-i", as described in the Bugreport at gnu.org

postmaster@lxmhs62:~$ time grep stoer...@domain.de /var/log/mail.log

real    0m0.171s
user    0m0.128s
sys     0m0.044s
postmaster@lxmhs62:~$ time LC_ALL=C grep stoer...@domain.de /var/log/mail.log

real    0m0.169s
user    0m0.108s
sys     0m0.060s


--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Reply via email to