package Jevix::Makeup;

# ==========================================
#
#  Jevix Version 0.9.5 (utf-8)
#
#  Developed by Igor Askarov
# 
#  Please send all suggestions to
#  Igor Askarov <juks@juks.ru>
#  http://www.jevix.ru/
#
#  Release date: 21/01/2008
#
# === Methods list==========================
#
#  new 						the constructor
#  procces 					entry sub
#  preset 					presets selector
#  makeup 					makeup the text
#  quotes 					quotes processor
#  cuttags 					tags processor
#  tagend					looking fo tag end
#  planttags 					sub to bring the tags back
#  vanish 					sub to remove all the stuff and bring the text to plain mode
#  parseTagsAllowString     			parse the tagsAllow string to hash
#  parseTagsDenyString      			parse the tagsDeny string to hash
#
# ==========================================

use strict;
use warnings;

#UTFMODE#
use Encode;
use utf8;
#UTFMODE#

my $markLength = 7;
my $strip;
my $result;
my $tags;
my $conf;

my @singleTags = qw/link input spacer img br hr/;
my @breakingTags = qw/p td div hr/;
my @spaceTags = qw/br/;
my @tagsToEat = qw/script style/;

# ==The constructor
sub new {
    my Jevix::Makeup $class = shift;

    return $class;
}

# ==Here we've got the input
sub process($$$) {
    my($class, $text, $userConfig) = @_;

    #UTFMODE#
    $$text = decode("utf8", $$text);
    #UTFMODE#

    $conf = $userConfig ? $userConfig : {presetBasic=>1};
    $class->preset();
    
    $strip = "";
    $tags = [];

    $result = {};
    $result->{error} = 0;
    $result->{errorLog} = [];

    if(!$conf->{isHTML}) { $strip = $$text; } else { $class->cuttags($text, $conf, $result); }
    if($conf->{quotes}) { $class->quotes($conf); }
    $class->makeup($conf);

    $result->{text} = "";
    if($conf->{isHTML}) { $class->planttags($result); } else { $result->{text} = $strip; }

    #UTFMODE#
    Encode::_utf8_off($result->{text});
    #UTFMODE#

    return $result;
}

# ==Choosing default setup when necessary
sub preset($$) {
    my ($class) = @_;

    if(!$conf || $conf->{presetBasic}) {
        $conf->{isHTML} = 1 if(!defined($conf->{isHTML}));                                # HTML mode
        $conf->{lineBreaks} = 1 if(!defined($conf->{lineBreaks}));                        # Linebreaks to <br/>
        $conf->{paragraphs} = 0 if(!defined($conf->{paragraphs}));                        # Paragraphs
        $conf->{dashes} = 1 if(!defined($conf->{dashes}));                                # Replace hyphens with dashes when necessary
        $conf->{dots} = 1 if(!defined($conf->{dots}));                                    # Replace 3 dots with a symbol
        $conf->{edgeSpaces} = 1 if(!defined($conf->{edgeSpaces}));                        # Wipe edge space characters
        $conf->{multiSpaces} = 1 if(!defined($conf->{multiSpaces}));                      # Wipe multispaces
        $conf->{redundantSpaces} = 1 if(!defined($conf->{redundantSpaces}));              # Wipe redundant spaces
        $conf->{compositeWordsLength} = 10 if(!defined($conf->{compositeWordsLength}));   # The maximim length of composite word to be put inside <nobr>
        $conf->{tagLf} = 1 if(!defined($conf->{tagLf}));                                  # Wipe crs and lfs after droppped tag
        $conf->{nbsp} = 1 if(!defined($conf->{nbsp}));                                    # Insert non-breaking spaces
        $conf->{quotes} = 1 if(!defined($conf->{quotes}));                                # Makeup quotes
        $conf->{qaType} = 0 if(!defined($conf->{qaType}));                                # Main quotes type
        $conf->{qbType} = 2 if(!defined($conf->{qbType}));                                # Nested quotes type
        $conf->{misc} = 1 if(!defined($conf->{misc}));                                    # Misc substitutions
        $conf->{codeMode} = 2 if(!defined($conf->{codeMode}));                            # The way jevix should represent html special characters
    }

    # If tagsAllow came as a string
    if(defined($conf->{tagsAllow}) && !ref($conf->{tagsAllow})) {
        my $tmp = $class->parseTagsAllowString($conf->{tagsAllow});
	$conf->{tagsAllow} = $tmp->{tagsAllow};
	$conf->{tagsDenyAllAttributes} = $tmp->{tagsDenyAllAttributes};
    }

    # If tagsDeny came as a string
    if(defined($conf->{tagsDeny}) && !ref($conf->{tagsDeny})) {
	$conf->{tagsDeny} = $class->parseTagsDenyString($conf->{tagsDeny});
    }
}

# ==Imposing clear text
sub makeup($$) {
    my ($class, $conf) = @_;
    
    # ==Misc
    # Prepositions
    my $prp_rus = "а|без|безо|в|вне|во|да|для|до|за|и|из|изо|или|к|как|на|над|надо|не|ни|но|о|об|обо|около|от|ото|по|под|подо|при|про|с|сквозь|со|у|через";
    my $prp_eng = "aboard|about|above|absent|across|after|against|along|alongside|amid|amidst|among|amongst|around|as|astride|at|atop|before|behind|below|beneath|beside|besides|between|beyond|but|by|despite|down|during|except|following|for|from|in|inside|into|like|mid|minus|near|nearest|notwithstanding|of|off|on|onto|opposite|out|outside|over|past|re|round|save|since|than|through|throughout|till|to|toward|towards|under|underneath|unlike|until|up|upon|via|with|within|without";
    my $prp = "$prp_rus|$prp_eng";

    my $letters = "A-Za-zА-Яа-яЁёЙй";         # Characters
    my $cap_letters = "A-ZА-ЯЁё";                 # Capital characters

    my $sp = " \xA0\t";                               # space class
    my $rt = "\r?\n";                                 # cr class

    my $br = "\x00\x0F.[\x01\x03].\x0F\x00";          # br tag
    my $pt = "\x00\x0F.[\x02].\x0F\x00";              # Paragraph tag
    my $ps = "\x00\x0F.[\x02][\x01\x03]\x0F\x00";     # Paragraph start
    my $pe = "\x00\x0F.[\x02][\x02\x00]\x0F\x00";     # Paragraph end
    my $to = "\x00\x0F..[\x03\x01]\x0F\x00";          # Opening tag
    my $tc = "\x00\x0F..[\x02\x00]\x0F\x00";          # Closing tag
    my $bb = "\x00\x0F..[\x02\x03]\x0F\x00";          # Tag where <nobr> is open
    my $nb = "\x00\x0F..[\x01\x00]\x0F\x00";          # Tag where no <nobr> open
    my $ts = "\x00\x0F";                              # Tag start
    my $te = "\x0F\x00";                              # Tag end

    my $brt = "<br *\/?>";                            # br tag in text mode
    my $pst = "<p>";
    my $pet = "</p>";

    # Codes, metasymbols or what ever?
    my ($cdash, $cnbsp, $cdots, $cfracs, $ccopy, $creg);
    if(!$conf->{codeMode}) {
        ($cdash, $cnbsp, $cdots, $ccopy, $creg) = ("—", " ", "…", "©", "®");
        $cfracs = {'1/4'=>"¼", '1/2'=>"½", '3/4'=>"¾"};
    } elsif($conf->{codeMode} == 1) {
        ($cdash, $cnbsp, $cdots, $ccopy, $creg) = ("&#151;", "&#160;", "&#133;", "&#169;", "&#174;");
        $cfracs = {'1/4'=>"&#188;", '1/2'=>"&#189;", '3/4'=>"&#190;"};
    } else {
        ($cdash, $cnbsp, $cdots, $ccopy, $creg) = ("&mdash;", "&nbsp;", "&hellip;", "&copy;", "&reg;");
        $cfracs = {'1/4'=>"&frac14;", '1/2'=>"&frac12;", '3/4'=>"&frac34;"};
    }

    # Wiping edge spaces
    if($conf->{edgeSpaces}) { $strip =~ s/^[$sp\r\n]*(.+?)[$sp\r\n]*$/$1/isg; }

    # Wiping spaces between tags (</td> </tr>)
    if($conf->{tagSpaces}) { $strip =~ s/($tc)[$sp]($tc)/$1$2/isg; }

    # Wiping multispaces
    if($conf->{multiSpaces}) { $strip =~ s/([$sp]){2,}/$1/ig; }

    # Wiping redundant spaces
    if($conf->{redundantSpaces}) { $strip =~ s{([$sp]+(?![:;]-[)(])([;:,.)?!]))|(\()(?<![:%;]-\()[$sp]+}{$1 ? $2 : $3}eig; } 

    if($conf->{nbsp}) {
        # Prepositions with &nbsp;
        $strip =~ s/(^|\x00|[$sp])($prp)[$sp]([0-9$letters])/$1$2$cnbsp$3/gm;

        # &nbsp; with digits
        $strip =~ s{($nb|^)(.*?)($bb|$)}{ my ($a, $b, $c) = ($1, $2, $3); $b =~ s/([0-9]+)([$sp]+|&nbsp;|&#160;)(?:(?=[0-9]{2,})|(?=%))/$1$cnbsp/ig; "$a$b$c"; }eisg;
    }    
    
    # Put composite words inside <nobr>
    if($conf->{compositeWords}) { $strip =~ s{($nb|^)(.*?)($bb|$)}{ my ($a, $b, $c) = ($1, $2, $3);
						$b =~ s{(^|[$sp\x00]|&nbsp;)([$letters]+(?:-[$letters]+)+)(?=$|[$sp\x00])}{
							    my $d = !defined($1) ? "" : $1; my $e = !defined($2) ? "" : $2; my $f = !defined($3) ? "" : $3;
							    if(length($e) <= $conf->{compositeWordsLength}) { "$d<nobr>$e<\/nobr>" } else {"$d$e$f"}
							}eig; "$a$b$c";
					    }eisg; }

    # Dots
    if($conf->{dots}) { $strip =~ s/\.{3}|…|&hellip;/$cdots/ig; }
   
    # Dashes
    if($conf->{dashes}) {
        # Hyphen
        $strip =~ s/([^$sp])([$sp]|&#160;|&nbsp;)(-{1,2}|—|&mdash;|&#151;)/$1$cnbsp$cdash/ig;
        # "Speech" hyphen
        $strip =~ s/((?:^|$ps|$br|$brt(?:$rt)*|[$rt]))[$sp]*(?:&nbsp;)*(-{1,2}|—|&mdash;|&#151;)[$sp]*(?:&nbsp;)*(.)/$1$cdash$cnbsp$3/ig;
    }

    # Misc stuff
    if($conf->{misc}) {
        # Fracs
        $strip =~ s{(?:(?<=[$sp\x00])|(?<=^))([13])/([24])(?:(?=[$sp\x00])|(?=$))}{if(defined($cfracs->{"$1/$2"})) { $cfracs->{"$1/$2"} } else { "$1/$2" } }esg;
        # Copyright & registered
        $strip =~ s{(?:(?<=[$sp\x00])|(?<=^))(\([cr]\)|&copy;|©)(?:(?=[$sp\x00?!;.,])|(?=$))}{ if((lc($1) eq "(c)") || (lc($1) eq "&copy;") || ($1 eq "©")) {$ccopy} elsif((lc($1) eq "(r)") || (lc($1) eq "&reg;") || ($1 eq "®")) {$creg} else { $2 } }eig;
    }

    # Paragraphs
    if($conf->{paragraphs}) { $strip =~ s{(^|$pe(?:$rt$rt)?|$rt$rt)(?!$ps)(.+?)($br)?($brt)?(?<!$pe)(?:(?=$)|(?=$rt$rt)|(?=$ps))}{ my ($a, $b, $c) = ($1,$2,$3||""); (($b =~ /^[ \r\n]+$/) || ($b =~ /^(<br *\/?>|$br)+$/)) ? "$a$b$c" : "$a<p>$b</p>";}eisg; }

    # Line break
    if($conf->{lineBreaks}) { $strip =~ s/(?<!$pt)(?<!$br)(?<!$br\r)(?<!$pe\r\n\r\n)(?<!$pe\n\n)(?<!$pe\r\n)(?<!$pe\n)(?<!$pe\r)(?<!$pe)(?<!$pet\r\n\r\n)(?<!$pet\r\n\r)(?<!$pet\n\n)(?<!$pet\r\n)(?<!$pet\n)(?<!$pet\r)(?<!$pet)(?<!$pst)($rt)(?!$brt)/<br \/>$1/isg; }
}

# ==impose quotes
sub quotes($$) {
    my ($class, $conf) = @_;

    my $i;
    my ($a_open, $b_open) = (0,0);
    my ($cp, $c, $cn, $cn_is_sp, $cp_is_sp) = ('', '', '', 0, 0);
    my ($qaStart, $qaEnd, $qbStart, $qbEnd);
    my (@qs, @qe, @qs_ansi, @qe_ansi, @qs_html, @qe_html, @qs_ent, @qe_ent,);

    # space class
    my $sp =" \t\xA0";
    # characters
    my $letters = "A-Za-zА-Яа-яЁёЙй";

    @qs_ansi = ("«", "“", "„", "‘", "‚", '"');    
    @qe_ansi = ("»", "”", "“", "’", "‘", '"');
    @qs_html = ("&#171;", "&#147;", "&#132;", "&#145;", "&#130;", "&#34;");
    @qe_html = ("&#187;", "&#148;", "&#147;", "&#146;", "&#145;", "&#34;");
    #          <<         ``         ..         `          .          "
    @qs_ent = ("&laquo;", "&ldquo;", "&bdquo;", "&lsquo;", "&sbquo;", "&quot;");    
    #          >>         ''         ''          '         `          "
    @qe_ent = ("&raquo;", "&rdquo;", "&ldquo;", "&rsquo;", "&lsquo;", "&quot;");

    # Quotes collection
    if(!$conf->{codeMode}) {
        @qs = @qs_ansi; @qe = @qe_ansi;
    } elsif ($conf->{codeMode} == 1) {
        @qs = @qs_html; @qe = @qe_html;
    } else {
        @qs = @qs_ent; @qe = @qe_ent;
    }
    
    # Getting configuration setting
    $conf->{qaType} ||= 0;
    $conf->{qbType} ||= 1;
    $conf->{qaType} = ($conf->{qaType} >= 0 && $conf->{qaType} <= 5) ? $conf->{qaType} : 0;
    $conf->{qbType} = ($conf->{qbType} >= 0 && $conf->{qbType} <= 5) ? $conf->{qbType} : 1;
   
    # Selecting quotes as requested by user
    ($qaStart, $qaEnd) = ($qs[$conf->{qaType}], $qe[$conf->{qaType}]);
    ($qbStart, $qbEnd) = ($qs[$conf->{qbType}], $qe[$conf->{qbType}]);
    
    # Resetting all the quotes inside text to <">
    my $qa = join('|', @qs_ansi) . '|' . join('|', @qe_ansi) . '|' . join('|', @qs_html) . '|' . join('|', @qe_html) . '|' . join('|', @qs_ent) . '|' . join('|', @qe_ent);
    $strip =~ s/(?:(?:(?<=[^$letters])|(?<=^))($qa))|(?:($qa)(?:(?=[^$letters])|(?=$)))/\"/ig;
    
    my $spread = 1;
    my $mv = 0;
    my $mvn = 0;
    my @st;
    $i = 0;
    my $skip = 0;
    my @space;          # Space tags flag
    my @break;          # Text break flags
    
    $st[$_] = '' foreach(0..$spread + 1);
    $space[$_] = 0 foreach(0 + 1..$spread + 1);
    $break[$_] = 0 foreach(0 + 1..$spread + 1);
    $space[0] = 1;
    $break[0] = 1;

    while(1) {
        # Skipping tags
        foreach(0..$spread) {
            do {
                $skip = 0;
                if($i + $_ + $mv <= length($strip)) {
                    if($i + $_ + $mv + 1 < length($strip)) {
                        if((substr($strip, $i + $_ + $mv, 1) eq "\x00") && (substr($strip, $i + $_ + $mv + 1, 1) eq "\x0F")) {
                            $space[$_ + 1] |= (ord(substr($strip, $i + $_ + $mv + 2, 1)) & 2) >> 1;
                            $break[$_ + 1] |= ord(substr($strip, $i + $_ + $mv + 2, 1)) & 1;
                            $mv += $markLength;
                            if(!$_) { $mvn = $mv; }
                            $st[$_ + 1] = "";
                            $skip = 1;
                        }
                    }
                    if(!$skip) { $st[$_ + 1] = substr($strip, $i + $_ + $mv, 1); }
                } 
            } while($skip);
        }
        
        $i += $mvn;
        $mv = 0;
        $mvn = 0;

	($cp, $c, $cn) = ($st[0], $st[1], $st[2]);
	$cp_is_sp = (($cp =~ /[^0-9$letters]/) || $space[0] || $space[1] || $break[0] || !$i) ? 1 : 0;
	$cn_is_sp = (($cn =~ /[^0-9$letters]/) || $space[2] || $break[2] || $cn eq '') ? 1 : 0;

        # Reset state if breaking tag appears
        if($break[1] || $i == length($strip)) {
            if($a_open || $b_open) {
                # Log quote error if appears
                if($conf->{logErrors}) {
                    my $quoteErrSampleLength = 100;
                    my $z = $i - 1;
                    my $y;
                    while(1) {
                        if(substr($strip, $z, 1) eq " " || substr($strip, $z, 1) eq "\xA0" || !$z) { if($i-$z <= $quoteErrSampleLength) {$y = $z}}
                        last if(!$z);
                        $z--;
                    }
                    my $sample = substr($strip, $y, ($i - $y));
                    $sample =~ s/\x00\x0F[^\x0F]+\x0F\x00//g;
                    $sample =~ s/<\/?[a-z]+.*?>//g;
                    push(@{$result->{errorLog}}, {type=>"Quote_error", message=>"Quote mismatch near [$sample]<--"});
                    $result->{error} = 1;
                }
            }
	    $a_open = 0;
	    $b_open = 0;
	}

        if($c eq '"') {
	    if(!$a_open) {
	        $a_open = 1;
		substr($strip, $i, 1) = $qaStart;
		$i += length($qaStart) - 1;
	    } elsif ($a_open && (($i == length($strip) - 1) || (!$b_open && $cn_is_sp))) {
	        $a_open = 0;
		substr($strip, $i, 1) = $qaEnd;
		$i += length($qaEnd) - 1;
	    } elsif ($a_open && !$b_open) {
	        $b_open = 1;
		substr($strip, $i, 1) = $qbStart;
		$i += length($qbStart) - 1;
	    } elsif ($a_open && $b_open) {
	        $b_open = 0;
		substr($strip, $i, 1) = $qbEnd;
		$i += length($qbEnd) - 1;
	    }
	}
    
	last if($i == length($strip));
	$st[0] = $st[1];
	$space[0] = $space[1];
        $break[0] = $break[1];
	$space[$_] = 0 foreach(0 + 1..$spread + 1);
        $break[$_] = 0 foreach(0 + 1..$spread + 1);
	$i++;
    }
}

# ==Cutting the tags away
sub cuttags($$$$) {
    my($class, $text, $conf, $result) = @_;
    my $i = 0;                                                                                                                    # loop counter
    my $hop;												                                                                      # Jump length
    my ($c, $cn);                                                                                                                 # current & next character
    my ($tl, $ts, $te, $cl, $tagName, $tagBody, $tagContent);                                                                     # tag length, tag dimensions, tag name, tag body text, single tag flag, content inside the tag
    my ($isTag, $isTagStart, $isSingle, $isSingleClosed, $isSpace, $isBreaking, $nobrIsOpen, $flagSet2, $flagSet1, $flagSet0);    # some useful flags
    my @tagsOpen;                                                                                                                 # an array storing the info about all the tags currently open
    
    # space class
    my $sp =" \t\xA0";

    while(1) {
        $hop = index($$text, "<", $i);

        if($hop < 0) {
            $strip .= substr($$text, $i, length($$text) - $i);
            last;
        } elsif($hop > 0) {
            $strip .= substr($$text, $i, $hop - $i);
            $i = $hop;
        }

        ($c, $cn) = unpack("aa", substr($$text, $i, 2));
	
	$isTag = 0;

        # =If tag opens
	$isTagStart = ($cn =~ /!|[a-z]/i) ? 1 : 0;
	if($isTagStart || ($cn eq "/")) { $isTag = 1; }

	if($isTag) {
	    $ts = $i;                                                                # Tag start position 
	    $te = $isTagStart ? tagend($text, $ts) : index($$text, ">", $ts);        # Tag end position

            if($te) {
                $tagBody = substr($$text, $ts, $te - $ts + 1);
		$tagName = $isTagStart ? ($tagBody =~ m/^<([a-z]+)/i)[0] : ($tagBody =~ m/^<\/\s*([a-z]+)/i)[0];
		$tagName =~ tr/A-Z/a-z/;
            }

	    if($te && $tagName) {
		$tagBody = substr($$text, $ts, $te - $ts + 1);
		$tagName = $isTagStart ? ($tagBody =~ m/^<([a-z]+)/i)[0] : ($tagBody =~ m/^<\/\s*([a-z]+)/i)[0];
		$tagName =~ tr/A-Z/a-z/;

		# =Flags
		# Detecting whether the tag is single (self-closing) or double
		$isSingleClosed = 0;
                $isSingle = 0;
		if($isTagStart) {
  		    if(grep{$tagName eq $_} @singleTags) {
		        $isSingle = 1;
 		    } elsif (substr($tagBody, length($tagBody) - 2, 1) eq "/") {
			$isSingle = 1;
			$isSingleClosed = 1;
		    }
		}

		# Detecting wether this is space tag or not
		$isSpace = (grep{$tagName eq $_} @spaceTags) ? 1 : 0;
		
		# Detecting wether this is breaking tag or not
		$isBreaking = (grep{$tagName eq $_} @breakingTags) ? 1 : 0;
		
		# Tag Length
		$tl = $te - $ts + 1;
		
		# Updating the status for tags open
		if($conf->{checkHTML} && !$isSingle) {
  		    if($isTagStart) {
		        push(@tagsOpen, $tagName);
  		    } else {
		        if($tagsOpen[$#tagsOpen] ne $tagName) {
			# HTML error
			    $result->{error} = 1;
			    if($conf->{logErrors}) { push(@{$result->{errorLog}}, {type=>"HTML_Parse", position=>$i, message=>"Found closing tag <$tagName> while waiting tag <" . $tagsOpen[$#tagsOpen] . "> to close!"}); }
			} else {
			    pop(@tagsOpen);
			}
 		    }
		}

                # Eating tag content for some tags like <script>
                $tagContent = "";
                $cl = 0;
   	        if((grep{$tagName eq $_} @tagsToEat) && $isTagStart) {
                      $cl = index($$text, "</$tagName>", $ts + $tl) - $ts - $tl;
                      if($cl > 0) {
                          $tagContent = substr($$text, $ts + $tl, $cl);
                      } else {
                          $cl = 0;
		          $result->{error} = 1;
			  if($conf->{logErrors}) { push(@{$result->{errorLog}}, {type=>"HTML_Parse", position=>$i, message=>"Can't find <$tagName> end!"}); }
                      }
                }

     	        # Should I drop all the tags by default?
		my $dropTag = 0;
		if($conf->{tagsDenyAll}) { $dropTag = 1; }
		
                # Checking deny list
                if(defined($conf->{tagsDeny}) && !$dropTag) {
                    if($conf->{tagsDeny}->{$tagName}) { $dropTag = 1; }
                }

		# Checking allow list
		if(defined($conf->{tagsAllow}) && $dropTag) {
		    if($conf->{tagsAllow}->{$tagName}) { $dropTag = 0; }
                }

                # Nobr tag status
                if($tagName eq "nobr" && $isTagStart) {
                    $nobrIsOpen = 1;
                } elsif(($tagName eq "nobr" && !$isTagStart) || (grep{$tagName eq $_} @breakingTags)) {
                    $nobrIsOpen = 0;
                }

                # =Final part
                if(!$dropTag) {
		    # =Processing tags
		    # Tag name to lower case
		    if($conf->{tagNamesToLower}) {
		        if($isTagStart) { $tagBody = "<" . $tagName . substr($tagBody, length($tagName) + 1, length($tagBody) - length($tagName) - 1); }
			else { $tagBody =~ tr/A-Z/a-z/; }
		    }                           
		    # Tag name to upper case
	            if($conf->{tagNamesToUpper}) {
	                if($isTagStart) { $tagBody = "<" . uc($tagName) . substr($tagBody, length($tagName) + 1, length($tagBody) - length($tagName) - 1); }
	                else { $tagBody =~ tr/a-z/A-Z/; }
	            }
		    # =Tag parameters to lower or upper case
		    if($isTagStart && ($conf->{tagAttributesToLower} || $conf->{tagAttributesToUpper})) {
		        # Regular parameters
		        my $tmp = "";
			while ($tagBody =~ m/([^\s]*\s*)(?:([a-z\r]+)(\s*)(?==)(=\s*))?/ig ) {
			    $tmp .= $1 if ($1); if($conf->{tagAttributesToLower}) { if($2) { $tmp .= lc($2); } } else { if($2) { $tmp .= uc($2); } } $tmp .= $3 if ($3); $tmp .= $4 if ($4); $tmp .= $5 if ($5);
			}

			# Single parameters (like <checked>)
                        if($conf->{tagAttributesToLower}) { $tagBody =~ s{(?<!=)( +([a-z]+))}{lc($1)}eig; }
                        elsif($conf->{tagAttributesToUpper}) { $tagBody =~ s{(?<!=)( +([a-z]+))}{uc($1)}eig; }
		    }
		    
		    # Simple XSS & tag attributes protection
		    if($isTagStart && ($conf->{simpleXSS} || $conf->{tagsAllow}->{$tagName}->{validAttributes} || $conf->{tagsAllow}->{$tagName}->{invalidAttributes} || $conf->{tagsAllow}->{$tagName}->{denyAllAttributes} || $conf->{tagsDenyAllAttributes})) {
			$tagBody =~ s{(?<!<)(\s*)([a-z]+)([$sp]*=[$sp]*)("[^"]+"|[^$sp/>]+)} {
			    my ($a, $b, $c, $d) = ($1||'', $2, $3, $4);
			    if($conf->{simpleXSS} && ($b =~ /^on/ig || $d =~ /javascript|expression/ig)) {
				'';
			    } elsif(($conf->{tagsDenyAllAttributes} || $conf->{tagsAllow}->{$tagName}->{denyAllAttributes} || ($conf->{tagsAllow}->{$tagName}->{invalidAttributes} && $conf->{tagsAllow}->{$tagName}->{invalidAttributes}->{$b}))
								    && !(($conf->{tagsAllow}->{$tagName}->{validAttributes} && $conf->{tagsAllow}->{$tagName}->{validAttributes}->{$b})
								    || $conf->{tagsAllow}->{$tagName}->{allowAllAttributes})
				    ) {
				'';
                            } elsif($conf->{tagsAllow}->{$tagName}->{validAttributes} && !$conf->{tagsAllow}->{$tagName}->{validAttributes}->{$b}) {
				'';
			    } else {
				$a . $b . $c . $d;
			    }
			}eig;			
		    }
		    
		    # Close single tag
		    if($conf->{tagCloseSingle} && $isSingle && !$isSingleClosed) {
		        if(substr($tagBody, length($tagBody) - 2, 1) ne "/") {
			    if(substr($tagBody, length($tagBody) - 2, 1) ne " ") { substr($tagBody, length($tagBody) - 2, 1) .= " /"; } else { substr($tagBody, length($tagBody) - 2, 1) .= "/"; }
			}
		    }
						
		    # Quote attribute values
		    if($conf->{tagQuoteValues} && $isTagStart) {
		        my $tmp = "";
		        #                   1          23  4    5       6
		        while($tagBody =~ m/([<a-z >]+)?((=)(\s*)([^ >]+)([ >]+))?/ig) {
		            $tmp .= $1 if($1);
		            if($2) {
		    		$tmp .= $3 if($3);
		    		$tmp .= $4 if($4);
		        	if($5 && substr($5, 0, 1) ne '"' && substr($5, length($5) - 1, 1) ne '"') { $tmp .= "\"$5\""; } else { $tmp .= $5; }
                                $tmp .= $6 if($6);
                            }
                        }
                        $tagBody = $tmp;
    		    }
																																														
		    # Unquote attribute values
		    if($conf->{tagUnQuoteValues}) {
		        $tagBody =~ s/([a-z]+)(\s*)(=)(\s*)"([^\=\s">]+)"/$1$2$3$4$5/ig;   #"
		    }
																																																						    
                    # Saving the tag
		    push(@$tags, {name=>$tagName, body=>$tagBody, content=>$tagContent});
		
                    # Forming flagSet
                    #
                    # |byte2: _ _ _ _ _ _ isSpace isBreaking| byte1: _ _ _ _ _ p br| byte0: _ _ _ _ nobr isTagStart
                    $flagSet2 = 0;
                    if($isSpace) { $flagSet2 |= 2; }
                    if($isBreaking) { $flagSet2 |= 1; }
                    $flagSet1 = 0;
                    if($tagName eq "br") { $flagSet1 |= 1; }
                    if($tagName eq "p") { $flagSet1 |= 2; }
                    $flagSet0 = 0;
                    if($isTagStart) { $flagSet0 |= 1; }
                    if($nobrIsOpen) { $flagSet0 |= 2; }
	  	    # Planting the marker
		    $strip .= "\x00\x0F" . chr($flagSet2) . chr($flagSet1) . chr($flagSet0) . "\x0F\x00";
		}
		
		# Moving the pointer (tag end position + content length)
		$i = $te + $cl;

                # Eating crs & lfs after dropped tag
                if($conf->{tagLf} && $dropTag) {
                    while(1) {
                        if(substr($$text, $i + 1, 1) eq "\r") { $i++; } elsif(substr($$text, $i + 1, 1) eq "\n") { $i++; last; } else { last }
                    }
                }
	    }
	} else {
        # This is not a tag, just add the "<" to result
            $strip .= $c;
        }
	
	last if($i == length($$text));
	$i++;
    }
}

# ==Find where tag ends
sub tagend($$$) {
    my ($text, $i) = @_;
    
    my $gotcha = 0;
    my $quote = 0;
    
    $i |= 0;
    
    while (1) {
        if (substr($$text, $i, 1) eq '"') { $quote ^= 1; }
	if (!$quote && substr($$text, $i, 1) eq '>') { $gotcha = $i; }
        last if ($i == length($$text) || $gotcha);
	$i++;
    }
    
    return $gotcha;
}

# ==Bring everything back to HTML
sub planttags($$) {
    my ($class, $result) = @_;
    my $i = 0;
    my $max = length($strip);
    my $ctag = 0;
    my $step;
  
    while (1) {
        if($i < $max - 2 && substr($strip, $i, 2) eq "\x00\x0F") {
	    $result->{text} .= $$tags[$ctag]->{body};
            if($$tags[$ctag]->{content}) { $result->{text} .= $$tags[$ctag]->{content}; }
	    $i += $markLength;
	    $ctag++;
	} else {
   	    if($i < $max - 2) { $step = index($strip, "\x00\x0F", $i) - $i; } else { $step = $max - $i; }
	    if($step < 0) { $step = $max - $i; }
	    
	    if($step >= 0) {
  	        $result->{text} .= substr($strip, $i, $step);
                $i += $step;
	    }
	}
	
        last if($i == $max);
    }
}

# ==Bring the text to plain mode==
sub vanish($$) {
    my($class, $text) = @_;

    $$text =~ s/&laquo;|&ldquo;|&bdquo;|&lsquo;|&sbquo;|&quot;|&raquo;|&rdquo;|&ldquo;|&rsquo;|&#171;|&#147;|&#132;|&#145;|&#130;|&#34;|&#187;|&#148;|&#146;|«|“|„|‘|‚|"|»|”|’/"/ig;
    $$text =~ s/&nbsp;|&#160;| / /ig;
    $$text =~ s/&mdash;|&ndash;|&#151;|&#150;|—|–/-/ig;
    $$text =~ s/&hellip;|&#133;|…/.../ig;
    $$text =~ s/&copy;|&#169;|©/(c)/ig;
    $$text =~ s/&reg;|&#174;|®/(r)/ig;
    $$text =~ s/&frac14;|&#188;|¼/1\/4/ig;
    $$text =~ s/&frac12;|&#189;|½/1\/2/ig;
    $$text =~ s/&frac34;|&#190;|¾/3\/4/ig;
}

# ==Parse the tagsAllow string advanced format==
sub parseTagsAllowString($$) {
    my($class, $string) = @_;

    return {tagsAllow=>{}, tagsDenyAllAttributes=>0} if(!$string);

    my $tagsAllow = {};
    my $tagsDenyAllAttributes = 0;

    # Should I deny all tag attributes by default?
    if(substr($string,0,1) eq '|') {
        $tagsDenyAllAttributes = 1;
	substr($string,0,1) = '';
    };
    
    # Parsing the Configuration String
    while($string =~ /([a-z:|]+)/ig) {
        my $tBody = $1;
	my ($tagName) = ($tBody =~ /^([a-z]+)/i)[0];
	
	last if(!$tagName);
	
	my $attrList = ();
	$tagsAllow->{$tagName}->{val}=1;
	
	if($tBody =~ /^$tagName\|$/i) {
	    $tagsAllow->{$tagName}->{denyAllAttributes}=1;
	} elsif($tBody =~ /^$tagName\:$/i) {
	    $tagsAllow->{$tagName}->{allowAllAttributes}=1;
	} else {
    	    while($tBody =~ /:([a-z]+)/ig) {
    		$tagsAllow->{$tagName}->{validAttributes}->{$1}=1;
    	    }
	
            while($tBody =~ /\|([a-z]+)/ig) {
		if(!$tagsAllow->{$tagName}->{validAttributes}->{$1}) {
        	    $tagsAllow->{$tagName}->{invalidAttributes}->{$1}=1;
		}
	    }
        }
    }

    return {tagsAllow=>$tagsAllow, tagsDenyAllAttributes=>$tagsDenyAllAttributes};
}

# ==Parse the tagsAllow string advanced format==
sub parseTagsDenyString($$) {
    my($class, $string) = @_;

    return {} if(!$string);

    my $tagsDeny = {};
    while($string =~ /([a-z]+)/ig) {
        $tagsDeny->{$1}->{val}=1;
    }

    return $tagsDeny;
}
                       
return 1;