You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@spamassassin.apache.org by si...@apache.org on 2004/10/13 10:36:08 UTC

svn commit: rev 54736 - spamassassin/trunk/lib/Mail/SpamAssassin

Author: sidney
Date: Wed Oct 13 01:36:07 2004
New Revision: 54736

Modified:
   spamassassin/trunk/lib/Mail/SpamAssassin/TextCat.pm
Log:
bug 3776: limit TextCat input to 10000 bytes to prevent excess memory and cpu usage

Modified: spamassassin/trunk/lib/Mail/SpamAssassin/TextCat.pm
==============================================================================
--- spamassassin/trunk/lib/Mail/SpamAssassin/TextCat.pm	(original)
+++ spamassassin/trunk/lib/Mail/SpamAssassin/TextCat.pm	Wed Oct 13 01:36:07 2004
@@ -54,7 +54,8 @@
   my $maxp = $opt_t;
 
   # create ngrams for input
-  my @unknown = create_lm($input);
+  # limit to 10000 characters, enough for accuracy and still fast enough
+  my @unknown = create_lm(substr($input, 0, 10000));
 
   # load language models once
   if (! @nm) {