This patch works significantly faster than vanilla 0.22.5
and no error or problem found since it tooks about 15 days
I've applied this patch on Windows XP and MacOS X ( It was announced for Japanese, sourceforge.jp users. )
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I've found a bottleneck in the 'add_line' function of Classifier/MailParse.pm.
It is the routine which is handling the 'spaced_out' trick.
I've posted the patch in the patch section:
http://sourceforge.net/tracker/index.php?func=detail&aid=1767506&group_id=63137&atid=502958
After applying this patch, the e-mail classification speeds up about 1.5
times.
While I can't understand what '\38' means, I replaced it with '&'.
If there's no problems, I want to commit the patch.
Naoki
This patch works significantly faster than vanilla 0.22.5
and no error or problem found since it tooks about 15 days
I've applied this patch on Windows XP and MacOS X ( It was announced for Japanese, sourceforge.jp users. )
Hi all,
I've merged the patch to the b0_22_2 branch and the MAIN branch.
Naoki