[popfile-commit] engine/Classifier MailParse.pm, 1.216.4.6, 1.216.4.7

[naoki i. <am...@us...>]

Update of /cvsroot/popfile/engine/Classifier
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv29448/Classifier

Modified Files:
      Tag: b0_22_2
	MailParse.pm 
Log Message:
Following Japanese support:
- Performance update for convert_encoding.
- Fix the problem with HTML entity.
- Clear the character set per mail to avoid using the wrong charsets.
- Fix the problem that 'uninitialized value' warnings occur on Japanese environment. 

Index: MailParse.pm
===================================================================
RCS file: /cvsroot/popfile/engine/Classifier/MailParse.pm,v
retrieving revision 1.216.4.6
retrieving revision 1.216.4.7
diff -C2 -d -r1.216.4.6 -r1.216.4.7
*** MailParse.pm	9 Feb 2006 00:25:31 -0000	1.216.4.6
--- MailParse.pm	23 Oct 2006 14:05:19 -0000	1.216.4.7
***************
*** 572,578 ****
                  if ( defined( $to ) ) {
  
!                     # HTML entities confilict with DBCS chars. Replace entities with blanks.
  
!                     if ( $self->{lang__} eq 'Korean' ) {
                              $to = ' ';
                      } else {
--- 572,578 ----
                  if ( defined( $to ) ) {
  
!                     # HTML entities confilict with DBCS and EUC-JP chars. Replace entities with blanks.
  
!                     if ( $self->{lang__} eq 'Korean' || $self->{lang__} eq 'Nihongo' ) {
                              $to = ' ';
                      } else {
***************
*** 1539,1542 ****
--- 1539,1545 ----
      $self->{colorized__} .= "<tt>" if ( $self->{color__} ne '' );
  
+     # Clear the character set to avoid using the wrong charsets
+     $self->{charset__} = '';
+ 
      # Since Text::Kakasi is not thread-safe, we use it under the
      # control of a Mutex to avoid a crash if we are running on
***************
*** 1641,1650 ****
              }
  
-             # Decode \x??
-             if ( $self->{lang__} eq 'Nihongo' && !$self->{in_headers__} ) {
-                 $line =~ s/\\x([8-9A-F][A-F0-9])/pack("C", hex($1))/eig;
-             }
- 
              if ( $self->{lang__} eq 'Nihongo' ) {
                  $line = convert_encoding( $line, $self->{charset__}, 'euc-jp', '7bit-jis', @{$encoding_candidates{$self->{lang__}}} );
                  $line = parse_line_with_kakasi( $self, $line );
--- 1644,1653 ----
              }
  
              if ( $self->{lang__} eq 'Nihongo' ) {
+                 # Decode \x??
+                 if ( !$self->{in_headers__} ) {
+                     $line =~ s/\\x([8-9A-F][A-F0-9])/pack("C", hex($1))/eig;
+                 }
+ 
                  $line = convert_encoding( $line, $self->{charset__}, 'euc-jp', '7bit-jis', @{$encoding_candidates{$self->{lang__}}} );
                  $line = parse_line_with_kakasi( $self, $line );
***************
*** 2480,2483 ****
--- 2483,2489 ----
      my ( $string, $from, $to, $default, @candidates ) = @_;
  
+     # If the string contains only ascii characters, do nothing
+     return $string if ( $string =~ /^[\r\n\t\x20-\x7E]*$/ );
+ 
      require Encode;
      require Encode::Guess;