popfile-commit Mailing List for POPFile - Automatic Email Classification (Page 154)

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

Update of /cvsroot/popfile/engine/Classifier
In directory sc8-pr-cvs1:/tmp/cvs-serv24532/Classifier

Modified Files:
	Bayes.pm 
Log Message:
Walk through and make sure that everytime we use a variable inside a regular expression we surround it with \Q and \E to prevent special characters inside the variable

Index: Bayes.pm
===================================================================
RCS file: /cvsroot/popfile/engine/Classifier/Bayes.pm,v
retrieving revision 1.103
retrieving revision 1.104
diff -C2 -d -r1.103 -r1.104
*** Bayes.pm	24 Feb 2003 23:40:19 -0000	1.103
--- Bayes.pm	27 Feb 2003 00:40:12 -0000	1.104
***************
*** 290,299 ****

      if ( $word ne '' ) {
! 		$word =~ /^(.)/;
! 		my $i = ord($1);

! 		$self->{matrix}{$bucket}[$i] = '' if ( !defined($self->{matrix}{$bucket}[$i]) );
! 		$self->{matrix}{$bucket}[$i] .= "|$word $value|" if ( ( $self->{matrix}{$bucket}[$i] =~ s/\|\Q$word\E (L?[\-\.\d]+)\|/\|$word $value\|/ ) == 0 );
! 	}
  }

--- 290,299 ----

      if ( $word ne '' ) {
!         $word =~ /^(.)/;
!         my $i = ord($1);

!         $self->{matrix}{$bucket}[$i] = '' if ( !defined($self->{matrix}{$bucket}[$i]) );
!         $self->{matrix}{$bucket}[$i] .= "|$word $value|" if ( ( $self->{matrix}{$bucket}[$i] =~ s/\|\Q$word\E (L?[\-\.\d]+)\|/\|$word $value\|/ ) == 0 );
!     }
  }

***************
*** 373,381 ****
      update_constants($self);

! 	# unclassified will always have the color black, note that unclassified is not
! 	# actually a bucket 
! 	
! 	$self->{colors}{unclassified} = 'black';
! 	
      print "Corpus loaded with $self->{full_total} entries\n" if $self->{debug};
  }
--- 373,381 ----
      update_constants($self);

!     # unclassified will always have the color black, note that unclassified is not
!     # actually a bucket 
!     
!     $self->{colors}{unclassified} = 'black';
!     
      print "Corpus loaded with $self->{full_total} entries\n" if $self->{debug};
  }
***************
*** 429,448 ****

              if ( /^([^ ]+) (.+)$/ )  {
!             	my $type  = $1;
!             	my $value = $2;

!             	$value =~ s/\\(\?|\*|\||\(|\)|\[|\]|\{|\}|\^|\$|\.)/$1/g;
                  $self->{magnets}{$bucket}{$type}{$value} = 1;
              } else {

!             	# This branch is used to catch the original magnets in an
!             	# old version of POPFile that were just there for from
!             	# addresses only

!             	if ( /^(.+)$/ ) {
! 					my $value = $1;
! 					$value =~ s/\\(\?|\*|\||\(|\)|\[|\]|\{|\}|\^|\$|\.)/$1/g;
! 	                $self->{magnets}{$bucket}{from}{$1} = 1;
! 	            }
              }
          }
--- 429,448 ----

              if ( /^([^ ]+) (.+)$/ )  {
!                 my $type  = $1;
!                 my $value = $2;

!                 $value =~ s/\\(\?|\*|\||\(|\)|\[|\]|\{|\}|\^|\$|\.)/$1/g;
                  $self->{magnets}{$bucket}{$type}{$value} = 1;
              } else {

!                 # This branch is used to catch the original magnets in an
!                 # old version of POPFile that were just there for from
!                 # addresses only

!                 if ( /^(.+)$/ ) {
!                     my $value = $1;
!                     $value =~ s/\\(\?|\*|\||\(|\)|\[|\]|\{|\}|\^|\$|\.)/$1/g;
!                     $self->{magnets}{$bucket}{from}{$1} = 1;
!                 }
              }
          }
***************
*** 538,544 ****
          for my $type (sort keys %{$self->{magnets}{$bucket}}) {

! 	    # You cannot use @ or $ inside a \Q\E regular expression and hence
! 	    # we have to change the $magnet and the text we are comparing against
! 	    # by changing the $ and @ signs to .

              my $noattype;
--- 538,544 ----
          for my $type (sort keys %{$self->{magnets}{$bucket}}) {

!         # You cannot use @ or $ inside a \Q\E regular expression and hence
!         # we have to change the $magnet and the text we are comparing against
!         # by changing the $ and @ signs to .

              my $noattype;
***************
*** 547,551 ****
              $noattype =~ s/[@\$]/\./g;

! 	    for my $magnet (sort keys %{$self->{magnets}{$bucket}{$type}}) {
                  my $regex;

--- 547,551 ----
              $noattype =~ s/[@\$]/\./g;

!         for my $magnet (sort keys %{$self->{magnets}{$bucket}{$type}}) {
                  my $regex;

***************
*** 699,703 ****

      if ( ( $total != 0 ) && ( $score{$ranking[0]} > log($self->{unclassified} * $total) ) ) {
! 		$class = $ranking[0];
      }

--- 699,703 ----

      if ( ( $total != 0 ) && ( $score{$ranking[0]} > log($self->{unclassified} * $total) ) ) {
!         $class = $ranking[0];
      }

***************
*** 761,775 ****
      while ( <$mail> ) {
          my $line;
! 		my $fileline;

          $line = $_;

! 		# This is done so that we remove the network style end of line CR LF 
! 		# and allow Perl to decide on the local system EOL which it will expand
! 		# out of \n when this gets written to the temp file
! 		
! 		$fileline = $line;
! 		$fileline =~ s/[\r\n]//g;
! 		$fileline .= "\n";

          # Check for an abort
--- 761,775 ----
      while ( <$mail> ) {
          my $line;
!         my $fileline;

          $line = $_;

!         # This is done so that we remove the network style end of line CR LF 
!         # and allow Perl to decide on the local system EOL which it will expand
!         # out of \n when this gets written to the temp file
!         
!         $fileline = $line;
!         $fileline =~ s/[\r\n]//g;
!         $fileline .= "\n";

          # Check for an abort
***************
*** 835,839 ****
          if ( $self->{configuration}->{configuration}{subject} ) {
              # Don't add the classification unless it is not present
!             if ( !( $msg_subject =~ /\[$classification\]/ ) && 
                   ( $self->{parameters}{$classification}{subject} == 1 ) &&
                   ( $self->{parameters}{$classification}{quarantine} == 0 ) )  {
--- 835,839 ----
          if ( $self->{configuration}->{configuration}{subject} ) {
              # Don't add the classification unless it is not present
!             if ( !( $msg_subject =~ /\[\Q$classification\E\]/ ) && 
                   ( $self->{parameters}{$classification}{subject} == 1 ) &&
                   ( $self->{parameters}{$classification}{quarantine} == 0 ) )  {
***************
*** 852,856 ****
      # Add the XTC header
      $msg_head_after .= "X-Text-Classification: $classification$eol" if ( ( $self->{configuration}->{configuration}{xtc}          ) &&
!       																     ( $self->{parameters}{$classification}{quarantine} == 0 ) );

      # Add the XPL header
--- 852,856 ----
      # Add the XTC header
      $msg_head_after .= "X-Text-Classification: $classification$eol" if ( ( $self->{configuration}->{configuration}{xtc}          ) &&
!                                                                          ( $self->{parameters}{$classification}{quarantine} == 0 ) );

      # Add the XPL header
***************
*** 881,891 ****
                  print $client "To: $self->{parser}->{to}$eol";
                  print $client "Date: $self->{parser}->{date}$eol";
! 				if ( $self->{configuration}->{configuration}{subject} ) {
! 					# Don't add the classification unless it is not present
! 					if ( !( $msg_subject =~ /\[$classification\]/ ) && 
! 						 ( $self->{parameters}{$classification}{subject} == 1 ) ) {
! 						$msg_subject = " [$classification]$msg_subject";
! 					}
! 				}
                  print $client "Subject:$msg_subject$eol";
                  print $client "X-Text-Classification: $classification$eol" if ( $self->{configuration}->{configuration}{xtc} );
--- 881,891 ----
                  print $client "To: $self->{parser}->{to}$eol";
                  print $client "Date: $self->{parser}->{date}$eol";
!                 if ( $self->{configuration}->{configuration}{subject} ) {
!                     # Don't add the classification unless it is not present
!                     if ( !( $msg_subject =~ /\[\Q$classification\E\]/ ) && 
!                          ( $self->{parameters}{$classification}{subject} == 1 ) ) {
!                         $msg_subject = " [$classification]$msg_subject";
!                     }
!                 }
                  print $client "Subject:$msg_subject$eol";
                  print $client "X-Text-Classification: $classification$eol" if ( $self->{configuration}->{configuration}{xtc} );

2003	Jan	Feb (160)	Mar (119)	Apr (111)	May (118)	Jun (101)	Jul (304)	Aug (113)	Sep (140)	Oct (137)	Nov (87)	Dec (122)
2004	Jan (78)	Feb (125)	Mar (131)	Apr (59)	May (121)	Jun (166)	Jul (150)	Aug (137)	Sep (73)	Oct (58)	Nov (27)	Dec (60)
2005	Jan (131)	Feb (84)	Mar (36)	Apr (8)	May (28)	Jun (20)	Jul (10)	Aug (72)	Sep (76)	Oct (34)	Nov (3)	Dec (29)
2006	Jan (13)	Feb (92)	Mar (7)	Apr (1)	May (1)	Jun (2)	Jul (4)	Aug (17)	Sep (5)	Oct (2)	Nov (8)	Dec (12)
2007	Jan (28)	Feb (15)	Mar	Apr	May (8)	Jun (4)	Jul (5)	Aug (8)	Sep (20)	Oct (38)	Nov (65)	Dec (92)
2008	Jan (21)	Feb (56)	Mar (27)	Apr (174)	May (25)	Jun	Jul	Aug	Sep	Oct	Nov	Dec

popfile-commit Mailing List for POPFile - Automatic Email Classification (Page 154)

popfile-commit — CVS commit list messages