X-Git-Url: http://dxcluster.org/gitweb/gitweb.cgi?a=blobdiff_plain;f=perl%2FBadWords.pm;h=b598c385d4e940f45716a4f2f35c303ad9b46e1e;hb=c77ea09e91a5f7c3052e3e30dfb48fcaad1e0dcd;hp=ff9dd04c8b4e3b70a634828862420fcac1ccaa8a;hpb=8e0eef80216fbb2bca3606daf5797e39b2889d7a;p=spider.git

diff --git a/perl/BadWords.pm b/perl/BadWords.pm
index ff9dd04c..b598c385 100644
--- a/perl/BadWords.pm
+++ b/perl/BadWords.pm
@@ -12,43 +12,107 @@ use strict;
 
 use DXUtil;
 use DXVars;
+use DXHash;
+use DXDebug;
+
 use IO::File;
 
-use vars qw(%badwords $fn);
+use vars qw($badword $regexcode);
+
+my $oldfn = "$main::data/badwords";
+my $regex = "$main::data/badw_regex";
+my $bwfn = "$main::data/badword";
+
+# copy issue ones across
+filecopy("$regex.issue", $regex) unless -e $regex;
+filecopy("$bwfn.issue", $bwfn) unless -e $bwfn;
+
+$badword = new DXHash "badword";
 
-$fn = "$main::data/badwords";
-%badwords = ();
+use vars qw($VERSION $BRANCH);
+$VERSION = sprintf( "%d.%03d", q$Revision$ =~ /(\d+)\.(\d+)/ );
+$BRANCH = sprintf( "%d.%03d", q$Revision$ =~ /\d+\.\d+\.(\d+)\.(\d+)/ ) || 0;
+$main::build += $VERSION;
+$main::branch += $BRANCH;
 
 # load the badwords file
 sub load
 {
 	my @out;
-	return unless -e $fn;
-	my $fh = new IO::File $fn;
+	my $fh = new IO::File $oldfn;
 	
 	if ($fh) {
-		%badwords = ();
 		while (<$fh>) {
 			chomp;
 			next if /^\s*\#/;
 			my @list = split " ";
 			for (@list) {
-				$badwords{lc $_}++;
+				$badword->add($_);
 			}
 		}
 		$fh->close;
+		$badword->put;
+		unlink $oldfn;
+	}
+	push @out, create_regex(); 
+	return @out;
+}
+
+sub create_regex
+{
+	my @out;
+	my $fh = new IO::File $regex;
+	
+	if ($fh) {
+		my $s = "sub { my \$str = shift; my \@out; \n";
+		while (<$fh>) {
+			chomp;
+			next if /^\s*\#/;
+			my @list = split " ";
+			for (@list) {
+				# create a closure for each word so that it matches stuff with spaces/punctuation
+				# and repeated characters in it
+				my $w = uc $_;
+				my @l = split //, $w;
+				my $e = join '+[\s\W]*', @l;
+				$s .= "push \@out, \$1 if \$str =~ /($e)/;\n";
+			}
+		}
+		$s .= "return \@out;\n}";
+		$regexcode = eval $s;
+		dbg($s) if isdbg('badword');
+		if ($@) {
+			@out = ($@);
+			dbg($@);
+			return @out;
+		}
+		$fh->close;
 	} else {
-		my $l = "can't open $fn $!";
-		dbg('err', $l);
+		my $l = "can't open $regex $!";
+		dbg($l);
 		push @out, $l;
 	}
+	
 	return @out;
 }
 
 # check the text against the badwords list
 sub check
 {
-	return grep { $badwords{$_} } split(/\b/, lc shift);
+	my $s = uc shift;
+	my @out;
+
+	dbg($s) if isdbg('badword');
+	push @out, &$regexcode($s) if $regexcode;
+	
+	return @out if @out;
+	
+	for (split(/\s+/, $s)) {
+		s/\'?S$//;
+		push @out, $_ if $badword->in($_);
+	}
+
+	return @out;
 }
 
 1;