svn commit: rev 9346 - incubator/spamassassin/trunk/lib/Mail/SpamAssassin

[fe...@apache.org]

Author: felicity
Date: Wed Mar 10 19:28:15 2004
New Revision: 9346

Modified:
   incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Bayes.pm
   incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStore.pm
   incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStoreDBM.pm
   incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStoreSQL.pm
Log:
tokenize_line was making lots of subroutine calls to get_magic_re in its tight loop of token checking.  let's just make the call once and cache it.

Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Bayes.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Bayes.pm	(original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/Bayes.pm	Wed Mar 10 19:28:15 2004
@@ -363,11 +363,19 @@
     }
   }
 
+  # cache the magic token regexp...
+  my $magic_re = $self->{store}->get_magic_re();
+
   foreach my $token (split) {
     $token =~ s/^[-'"\.,]+//;        # trim non-alphanum chars at start or end
     $token =~ s/[-'"\.,]+$//;        # so we don't get loads of '"foo' tokens
 
-    next if ( $self->{store}->is_magic_token($token) ); # skip false magic tokens
+    # Skip false magic tokens
+    # TVD: we need to do a defined() check since SQL doesn't have magic
+    # tokens, so the SQL BayesStore returns undef.  I really want a way
+    # of optimizing that out, but I haven't come up with anything yet.
+    #
+    next if ( defined $magic_re && /$magic_re/ );
 
     # *do* keep 3-byte tokens; there's some solid signs in there
     my $len = length($token);

Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStore.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStore.pm	(original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStore.pm	Wed Mar 10 19:28:15 2004
@@ -685,18 +685,17 @@
   die "touches_cleanup: not implemented\n";
 }
 
-=head2 is_magic_token
+=head2 get_magic_re
 
-public instance (Boolean) is_magic_token (string $token)
+public instance get_magic_re (String)
 
 Description:
-This method determines if a given token is "magic" or special to the
-implementation.
+This method returns a regexp which indicates a magic token.
 
 =cut
 
-sub is_magic_token {
-  my ($self, $token) = @_;
+sub get_magic_re {
+  my ($self) = @_;
   die "is_magic_token: not implemented\n";
 }
 

Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStoreDBM.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStoreDBM.pm	(original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStoreDBM.pm	Wed Mar 10 19:28:15 2004
@@ -919,14 +919,6 @@
   return qr/^\*\*[A-Z]+$/;
 }
 
-sub is_magic_token {
-  my ($self, $token) = @_;
-
-  my $magic_re = $self->get_magic_re($self->{db_version});
-
-  return ($token =~ /$magic_re/);
-}
-
 # provide a more generalized public insterface into the journal sync
 
 sub sync {

Modified: incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStoreSQL.pm
==============================================================================
--- incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStoreSQL.pm	(original)
+++ incubator/spamassassin/trunk/lib/Mail/SpamAssassin/BayesStoreSQL.pm	Wed Mar 10 19:28:15 2004
@@ -892,20 +892,13 @@
   return 1;
 }
 
-=head2 is_magic_token
-
-public instance (Boolean) is_magic_token (string $token)
-
-Description:
-This method determines if a given token is "magic" or special to the
-implementation.
+=head2 get_magic_re
 
 =cut
 
-sub is_magic_token {
-  my ($self, $token) = @_;
-
-  return 0; # nothing is magic
+sub get_magic_re {
+  my ($self) = @_;
+  undef;
 }
 
 =head2 sync