Thank you for keeping me updated of the progress (or lack thereof?).
If I can do anything, like testing the patch in Japanese locale, please
let me know.
Sincerely,
Chiaki Ishikawa
Aníbal Monsalve Salazar wrote:
forwarded 503658 http://savannah.gnu.org/bugs/?14472
thanks
On Tue, Oct 28, 2008 at 02:57:05AM +0900, Chiaki wrote:
Ubuntu had a similar bug report with the title "huge performance hit
for -i with UTF-8 locales"
https://bugs.launchpad.net/ubuntu/+source/grep/+bug/75695
Caution: In it, it was suggested that open() calls to load proper
locale-specific runtime libraries may also add to the overhead.
It is true, but the large amount of search time as observed can't be
explained by these open() calls alone. The search time is roughly a *
s + b (s being the input size, a and b are constants. 'b' here denotes
the overhead of initial open() calls for dynamic loading. Here the
problem is the VERY LARGE 'a' in the case of ja_JP.ujis and other
locales, it seems and 'b' is not much of a problem.)
The ubuntu report states the bug is known upstream and the patch has
been proposed. (and the comment was dated 2008 summer).
http://savannah.gnu.org/bugs/?14472
I hope the fix will be made available to users soon.
--
int main(void){int j=2008;/*(c)2008 cishikawa. */
char t[] ="<CI> @abcdefghijklmnopqrstuvwxyz.,\n\"";
char *i ="b>qtCIuqivb,gcwe...@.ietciuqi\"tqkvv is>dnamz";
while(*i)((j+=(int)strchr(t,*i++)-(int)t),(j%=sizeof t-1),
(putchar(t[j])));return 0;}/* under GPL */
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
