...

"; $g_close_p = "

# and tags get encoded. # my $text = shift; # Any remaining arguments after the first are options; either a single # hashref or a list of name, value pairs. See _SanitizeOpts comments. %opt = ( # set initial defaults style_prefix => $g_style_prefix, empty_element_suffix => $g_empty_element_suffix, tab_width => _tabDefault, indent_width => $g_indent_width, abs_prefix => "", # Prefixed to absolute path URLs url_prefix => "", # Prefixed to non-absolute URLs img_prefix => "", # Prefixed to non-absolute image URLs base_prefix => "", # Prefixed to fragment-only URLs ); @autonum = (); my %args = (); if (ref($_[0]) eq "HASH") { %args = %{$_[0]}; } else { %args = @_; } while (my ($k,$v) = each %args) { $opt{$k} = $v; } _SanitizeOpts(\%opt); my $yaml; ($text, $yaml) = _PrepareInput($text, $opt{yamlmode}); _ApplyYAMLOpts($yaml, \%opt) if ref($yaml) eq "HASH" && $opt{yamlmode} > 0; my $yamltable = ""; if (ref($yaml) eq "HASH" && %$yaml && $opt{yamlmode} && $opt{yamlvis}) { if ($opt{yamlvis} > 0 || _HasUnknownYAMLOptions($yaml)) { my ($hrows, $drows) = ("", ""); foreach (sort(keys(%$yaml))) { my $v = $yaml->{$_}; my $rspn = ''; if (defined($v)) { $v =~ s/&/&/g; $v =~ s/" . $v . "\n"; } else { $rspn = " class=\"$opt{style_prefix}yaml-undef-value\" rowspan=\"2\" valign=\"top\""; } $hrows .= "" . $_ . "\n"; } $yamltable = "\n" . "\n$hrows\n\n$drows\n

\n"; $opt{firstline} = scalar(@{[$yamltable =~ /\n/g]}); } } # Clear the globals. If we don't clear these, you get conflicts # from other articles when generating a page which contains more than # one article (e.g. an index page that shows the N most recent # articles): %g_urls = (); %g_titles = (); %g_anchors = (); %g_anchors_id = (); %g_block_ids = (); %g_code_block_ids = (); %g_html_blocks = (); %g_code_blocks = (); @g_xml_comments = (); $g_list_level = 0; # Make sure $text ends with a couple of newlines: $text .= "\n\n"; # Handle backticks-delimited code blocks $text = _HashBTCodeBlocks($text); # Convert all tabs to spaces. $text = _DeTab($text); # Strip any lines consisting only of spaces. # This makes subsequent regexen easier to write, because we can # match consecutive blank lines with /\n+/ instead of something # contorted like / *\n+/ . $text =~ s/^ +$//mg; # Turn block-level HTML blocks into hash entries $text = _HashHTMLBlocks($text, 1); # Strip link definitions, store in hashes. $text = _StripLinkDefinitions($text); $text = _RunBlockGamut($text, 1); # Remove indentation markers $text =~ s/\027+//gs; # Expand auto number flags $text =~ s/\034([1-6])/_AutoHeaderNum(ord($1)&0x7)/gse if $opt{auto_number} ne '' && $opt{auto_number} > 0; # Unhashify code blocks $text =~ s/(\025\d+\026)/$g_code_blocks{$1}/g; $text = _UnescapeSpecialChars($text); $text .= "\n" unless $text eq ""; # Sanitize all '<'...'>' tags if requested $text = _SanitizeTags($text, $opt{xmlcheck}, 1) if $opt{sanitize}; utf8::encode($text); if (ref($_[0]) eq "HASH") { ${$_[0]}{anchors} = {%g_anchors_id} if exists(${$_[0]}{anchors}); if (defined($opt{h1}) && $opt{h1}) { utf8::encode($opt{h1}); ${$_[0]}{h1} = $opt{h1}; } ${$_[0]}{yaml} = $yaml if ref($yaml) eq "HASH"; } # Eliminate known named character entities $opt{keep_named_character_entities} or do { $yamltable = ConvertNamedCharacterEntities($yamltable); $text = ConvertNamedCharacterEntities($text); }; if ($opt{divwrap}) { my $id = $opt{divname}; defined($id) or $id = ""; $id eq "" or $id = ' id="'.escapeXML($id).'"'; chomp($text); return "\n".$yamltable.$text."\n\n"; } return $yamltable.$text; } sub _HashBTCodeBlocks { # # Process Markdown backticks (```) delimited code blocks # Process some (limited recognition) tilde (~~~) delimited code blocks # my $text = shift; my $less_than_indent = $opt{indent_width} - 1; $text =~ s{ (?:(?<=\n)|\A) ([ ]{0,$less_than_indent})``(`+)[ \t]*(?:([\w.+-]+[#]?)(?:[ \t][ \t\w.+-]*)?)?\n ( # $4 = the code block -- one or more lines, starting with ``` (?: .*\n )+? ) # and ending with ``` or end of document (?:(?:[ ]{0,$less_than_indent}``\2`*[ \t]*(?:\n|\Z))|\Z) }{ # $2 contains syntax highlighting to use if defined my $leadsp = length($1); my $codeblock = $4; $codeblock =~ s/[ \t]+$//mg; # trim trailing spaces on lines $codeblock = _DeTab($codeblock, 8, $leadsp); # physical tab stops are always 8 $codeblock =~ s/\A\n+//; # trim leading newlines $codeblock =~ s/\s+\z//; # trim trailing whitespace $codeblock = _EncodeCode($codeblock); # or run highlighter here $codeblock = "

"
		. $codeblock . "\n

"; my $key = block_id($codeblock); $g_html_blocks{$key} = $codeblock; "\n\n" . $key . "\n\n"; }egmx; $text =~ s{ (?:(?<=\n)|\A) ([ ]{0,$less_than_indent})~~(~)[ \t]*(?:([\w.+-]+[#]?)(?:[ \t][ \t\w.+-]*)?)?\n ( # $4 = the code block -- one or more lines, starting with ~~~ (?: .*\n )+? ) # and ending with ~~~ or end of document (?:(?:[ ]{0,$less_than_indent}~~\2~*[ \t]*(?:\n|\Z))|\Z) }{ # $2 contains syntax highlighting to use if defined my $leadsp = length($1); my $codeblock = $4; $codeblock =~ s/[ \t]+$//mg; # trim trailing spaces on lines $codeblock = _DeTab($codeblock, 8, $leadsp); # physical tab stops are always 8 $codeblock =~ s/\A\n+//; # trim leading newlines $codeblock =~ s/\s+\z//; # trim trailing whitespace $codeblock = _EncodeCode($codeblock); # or run highlighter here $codeblock = "

"
		. $codeblock . "\n

"; my $key = block_id($codeblock); $g_html_blocks{$key} = $codeblock; "\n\n" . $key . "\n\n"; }egmx; return $text; } sub _StripLinkDefinitions { # # Strips link definitions from text, stores the URLs and titles in # hash references. # my $text = shift; my $less_than_indent = $opt{indent_width} - 1; # Link defs are in the form: ^[id]: url "optional title" while ($text =~ s{ ^[ ]{0,$less_than_indent}\[(.+)\]: # id = $1 [ ]* \n? # maybe *one* newline [ ]* ? # url = $2 [ ]* \n? # maybe one newline [ ]* (?: (?<=\s) # lookbehind for whitespace (?:(['"])|($)) # title quote char (.+?) # title = $5 (?(4)$|\3) # match same quote [ ]* )? # title is optional (?:\n+|\Z) } {}mx) { my $id = _strip(lc $1); # Link IDs are case-insensitive my $url = $2; my $title = _strip($5); $url =~ s/\\\n\s*//gs; if ($id ne "") { # These values always get passed through _MakeATag or _MakeIMGTag later $g_urls{$id} = $url; if (defined($title) && $title ne "") { $g_titles{$id} = $title; } } } return $text; } my %ok_tag_name; # initialized later my ($block_tags_a, $block_tags_b, $block_tags_c); BEGIN { $block_tags_a = qr/\020|p|div|center|h[1-6]|blockquote|pre|table|dl|ol|ul|script|noscript|form|fieldset|iframe|math|ins|del/io; $block_tags_b = qr/\020|p|div|center|h[1-6]|blockquote|pre|table|dl|ol|ul|script|noscript|form|fieldset|iframe|math/io; $block_tags_c = qr/div|center|h[1-6]|blockquote|pre|table|dl|ol|ul|script|noscript|form|fieldset|iframe|math/io; } sub _HashHTMLBlocks { my ($text, $toplevel) = @_; my $less_than_indent = $opt{indent_width} - 1; my $idt = "\027" x $g_list_level; my $blkprc = $toplevel ? sub { return $ok_tag_name{$_[1]} ? _EncodeAmpsAndAngles($_[0]) : $_[0] } : sub { return $_[0] }; # Hashify HTML blocks: # We only want to do this for block-level HTML tags, such as headers, # lists, and tables. That's because we still want to wrap

s around # "paragraphs" that are wrapped in non-block-level tags, such as anchors, # phrase emphasis, and spans. The list of tags we're looking for is # hard-coded: # First, look for nested blocks, e.g.: #

# tags for inner block must be indented. #

# # The outermost tags must start at the left margin for this to match, and # the inner nested divs must be indented. # We need to do this before the next, more liberal match, because the next # match will start at the first `

` and stop at the first `

`. $text =~ s{ ( # save in $1 ^ # start of line (with /m) ((?:\Q$idt\E)?) # optional lead in = $2 <($block_tags_a) # start tag = $3 \b # word break (?:.*\n)*? # any number of lines, minimally matching \2 # the matching end tag [ ]* # trailing spaces (?=\n+|\Z) # followed by a newline or end of document ) }{ my $blk = &$blkprc($1, $3); my $key = block_id($blk); $g_html_blocks{$key} = $blk; "\n\n" . $key . "\n\n"; }eigmx; # # Now match more liberally, simply from `\n` to `\n` # $text =~ s{ ( # save in $1 ^ # start of line (with /m) (?:\Q$idt\E)? # optional lead in <($block_tags_b) # start tag = $2 \b # word break (?:.*\n)*? # any number of lines, minimally matching .* # the matching end tag [ ]* # trailing spaces (?=\n+|\Z) # followed by a newline or end of document ) }{ my $blk = &$blkprc($1, $2); my $key = block_id($blk); $g_html_blocks{$key} = $blk; "\n\n" . $key . "\n\n"; }eigmx; # # Now match any empty block tags that should have been paired # $text =~ s{ ( # save in $1 ^ # start of line (with /m) (?:\Q$idt\E)? # optional lead in <($block_tags_c) # start tag = $2 \b # word break (?:[^<>])*? # /?> # the matching end tag [ ]* # trailing spaces (?=\n+|\Z) # followed by a newline or end of document ) }{ my $key = block_id($1); $g_html_blocks{$key} = $1; "\n\n" . $key . "\n\n"; }eigmx; # Special case just for

. It was easier to make a special case than # to make the other regex more complicated. $text =~ s{ (?: (?<=\n) # Starting after end of line | # or \A # the beginning of the doc ) ( # save in $1 [ ]{0,$less_than_indent} <(?:hr) # start tag \b # word break (?:[^<>])*? # /?> # the matching end tag [ ]* (?=\n{1,}|\Z) # followed by end of line or end of document ) }{ my $key = block_id($1); $g_html_blocks{$key} = $1; "\n\n" . $key . "\n\n"; }eigx; # Special case for standalone XML comments: $opt{stripcomments} != 2 && $text =~ s{ (?: (?<=\n\n) # Starting after a blank line | # or \A\n? # the beginning of the doc ) ( # save in $1 [ ]{0,$less_than_indent} (?s: (?: (?:[ \t]*\n[ \t]*)? )* ) [ ]* (?=\n{1,}|\Z) # followed by end of line or end of document ) }{ my $key = block_id($1); push(@g_xml_comments, $key) if $opt{stripcomments} && $opt{stripcomments} < 3 && !exists($g_html_blocks{$key}); $g_html_blocks{$key} = $1; "\n\n" . $key . "\n\n"; }egx; # Special case for standalone XML-like comments: $opt{stripcomments} >= 2 && $text =~ s{ (?: (?<=\n\n) # Starting after a blank line | # or \A\n? # the beginning of the doc ) ( # save in $1 [ ]{0,$less_than_indent} (?s: (?: (?:[ \t]*\n[ \t]*)? )* ) [ ]* (?=\n{1,}|\Z) # followed by end of line or end of document ) }{ my $key = block_id($1); push(@g_xml_comments, $key) unless exists($g_html_blocks{$key}); $g_html_blocks{$key} = $1; "\n\n" . $key . "\n\n"; }egx; return $text; } sub _RunBlockGamut { # # These are all the transformations that form block-level # tags like paragraphs, headers, and list items. # my ($text, $anchors) = @_; $text = _DoHeaders($text, $anchors); # Do Horizontal Rules: $text =~ s{^ {0,3}\*(?: {0,2}\*){2,}[ ]*$}{\n tags around block-level tags. $text = _HashHTMLBlocks($text); $text = _FormParagraphs($text, $anchors); return $text; } sub _DoListBlocks { return _DoBlockQuotes(_DoCodeBlocks($_[0])) if $_[0] ne ""; } sub _RunSpanGamut { # # These are all the transformations that occur *within* block-level # tags like paragraphs, headers, and list items. # my $text = shift; $text = _DoCodeSpans($text); $text = _EscapeSpecialChars($text); # Process anchor and image tags. Images must come first, # because ![foo][f] looks like an anchor. $text = _DoImages($text); $text = _DoAnchors($text); # Make links out of things like `` # Must come after _DoAnchors(), because you can use < and > # delimiters in inline links like [this](). $text = _DoAutoLinks($text); $text = _EncodeAmpsAndAngles($text); $text = _DoItalicsAndBoldAndStrike($text); # Do hard breaks: $text =~ s/ {3,}(\n|\z)/
or

 tags.
#   my $tags_to_skip = qr!<(/?)(?:pre|code|kbd|script|math)[\s>]!;

    foreach my $cur_token (@$tokens) {
	if ($cur_token->[0] eq "tag") {
	    # Within tags, encode *, _ and ~ so they don't conflict
	    # with their use in Markdown for italics and strong.
	    # We're replacing each such character with its
	    # corresponding block id value; this is likely
	    # overkill, but it should prevent us from colliding
	    # with the escape values by accident.
	    $cur_token->[1] =~ s!([*_~])!$g_escape_table{$1}!g;
	    $text .= $cur_token->[1];
	} else {
	    my $t = $cur_token->[1];
	    $t = _EncodeBackslashEscapes($t);
	    $text .= $t;
	}
    }
    return $text;
}


sub _ProcessWikiLink {
    my ($link_text, $link_loc) = @_;
    if (defined($link_loc) &&
	($link_loc =~ m{^#\S*$} || $link_loc =~ m{^(?:http|ftp)s?://\S+$}i)) {
	# Return the new link
	return _MakeATag(_FindFragmentMatch($link_loc), $link_text);
    }
    if (!defined($link_loc)) {
	$link_loc = _RunSpanGamut($link_text);
	$link_loc = _strip(unescapeXML(_StripTags(_UnescapeSpecialChars($link_loc))));
	$link_loc =~ m{^(?:http|ftp)s?://\S+$}i and
	    # Return the new link
	    return _MakeATag($link_loc, $link_text);
    }
    return undef if $link_loc eq "" || $link_text eq "";
    if ($link_loc =~ /^[A-Za-z][A-Za-z0-9+.-]*:/os) {
	# Unrecognized scheme
	return undef;
    }
    if ($opt{wikipat}) {
	my $o = $opt{wikiopt};
	my $img_link = _strip($link_text);
	my $img = 0;
	my $qsfrag = "";
	my $base;
	my $imgopts = undef;
	if ($img_link =~ /^[^#?\s]+\.(?:png|gif|jpe?g|svgz?)$/i) {
	    $base = _wxform($img_link, 1);
	    $img = 1;
	    $imgopts = _ParseWikiImgOpts($link_loc);
	    $imgopts->{imgflag} = \$img;
	} else {
	    $base = $link_loc;
	    if ($link_loc =~ /^(.*?)([?#].*)$/os) {
		($base, $qsfrag) = ($1, $2);
	    }
	    $base = _wxform($base);
	}
	my $result = $opt{wikipat};
	$result =~ s/%\{\}.+$/%{}/os if $img;
	$result =~ s/%\{\}/$base/;
	if ($qsfrag =~ /^([^#]*)(#.+)$/os) {
	    my ($q,$f) = ($1,$2);
	    #$f = _wxform($f) if $f =~ / /;
	    $qsfrag = $q . $f;
	}
	$result .= $qsfrag;
	$result = &{$opt{wikifunc}}($result, \%opt, ($img?$img_link:$link_loc), $base, $qsfrag, $imgopts)
		if ref($opt{wikifunc}) eq 'CODE';
	{
	    use bytes;
	    $result =~ s/%(?![0-9A-Fa-f]{2})/%25/sog;
	    if ($o->{r}) {
		$result =~
		s/([\x00-\x1F <>"{}|\\^`x7F])/sprintf("%%%02X",ord($1))/soge;
	    } else {
		$result =~
		s/([\x00-\x1F <>"{}|\\^`\x7F-\xFF])/sprintf("%%%02X",ord($1))/soge;
	    }
	    $result =~ s/(%(?![0-9A-F]{2})[0-9A-Fa-f]{2})/uc($1)/soge;
	}
	# Return the new link
	return $img ? _MakeIMGTag($result, undef, undef, $imgopts) : _MakeATag($result, $link_text);
    }
    # leave it alone
    return undef;
}


sub _ParseWikiImgOpts {
    my $alts = shift;
    my %o = ();
    # alt= consumes the rest of the line, do it first
    if ($alts =~ /(?:^|,)\s*alt\s*=\s*(.*)$/ios) {
	my $atext = $1;
	$alts = substr($alts, 0, $-[0]);
	$o{alt} = _strip($atext);
    }
    foreach my $kv (split(/\s*,\s*/, lc($alts))) {
	if ($kv =~ /^\s*([^\s]+)\s*=\s*([^\s]+)\s*$/os) {
	    my ($k, $v) = ($1, $2);
	    if (($k eq "width" || $k eq "height") && $v =~ /^\d+$/) {
		$o{$k} = 0+$v if $v > 0;
		next;
	    }
	    if ($k eq "align" && ($v eq "left" || $v eq "right" || $v eq "center")) {
		$o{$k} = $v;
		next;
            }
	}
    }
    return \%o;
}


sub _wxform {
    my ($w, $img) = @_;
    my $o = $opt{wikiopt};
    my $opt_s = $o->{s};
    if (!$img && $opt_s) {
	if (ref($opt_s)) {
	    if ($w =~ m{^(.*)[.]([^./]*)$}) {
		my ($base, $ext) = ($1, $2);
		$w = $base if $opt_s->{lc($ext)};
	    }
	} else {
	    $w =~ s{[.][^./]*$}{};
	}
    }
    $w = uc($w) if $o->{u};
    $w = lc($w) if $o->{l};
    $w =~ s{/+}{%252F}gos if $o->{"%"};
    $w =~ s/ +/%20/gos if $o->{b};
    $w =~ tr{/}{ } if $o->{f};
    $w =~ s{/+}{/}gos if !$o->{f} && !$o->{v};
    if ($o->{d}) {
	$w =~ tr{ }{-};
	$w =~ s/-+/-/gos unless $o->{v};
    } else {
	$w =~ tr{ }{_};
	$w =~ s/_+/_/gos unless $o->{v};
    }
    return $w;
}


# Return a suitably encoded  tag string
# On input NONE of $url, $text or $title should be xmlencoded
# but $url should already be url-encoded if needed, but NOT g_escape_table'd
sub _MakeATag {
    my ($url, $text, $title) = @_;
    defined($url) or $url="";
    defined($text) or $text="";
    defined($title) or $title="";

    $url =~ m"^#" && ref($opt{base_prefix}) eq 'CODE' and $url = &{$opt{base_prefix}}($url);
    my $result = $g_escape_table{'<'}."a href=\"" . _EncodeAttText($url) . "\"";
    $title = _strip($title);
    $text =~ s{<(/?a)}{<$1}sogi;
    $text = _DoItalicsAndBoldAndStrike($text);
    # We've got to encode any of these remaining to avoid
    # conflicting with other italics, bold and strike through and links.
    $text =~ s!([]*_~[])!$g_escape_table{$1}!g;
    $result .= " title=\"" . _EncodeAttText($title) . "\"" if $title ne "";
    return $result . $g_escape_table{'>'} .
	$text . $g_escape_table{'<'}."/a".$g_escape_table{'>'};
}


sub _DoAnchors {
#
# Turn Markdown link shortcuts into XHTML

` blocks. # my $text = shift; $text =~ s{ (?:\n\n|\A\n?) ( # $1 = the code block -- one or more lines, starting with indent_width spaces (?: (?:[ ]{$opt{indent_width}}) # Lines must start with indent_width of spaces .*\n+ )+ ) ((?=^[ ]{0,$opt{indent_width}}\S)|\Z) # Lookahead for non-space at line-start, or end of doc }{ my $codeblock = $1; $codeblock =~ s/\n\n\n/\n\n/g; # undo "paragraph for last list item" change $codeblock = _EncodeCode(_Outdent($codeblock)); $codeblock =~ s/\A\n+//; # trim leading newlines $codeblock =~ s/\s+\z//; # trim trailing whitespace my $result = " " . $codeblock . "\n"; my $key = block_id($result, 2); $g_code_blocks{$key} = $result; "\n\n" . $key . "\n\n"; }egmx; return $text; } sub _DoCodeSpans { # # * Backtick quotes are used for spans. # # * You can use multiple backticks as the delimiters if you want to # include literal backticks in the code span. So, this input: # # Just type ``foo `bar` baz`` at the prompt. # # Will translate to: # # Just type foo `bar` baz at the prompt. # # There's no arbitrary limit to the number of backticks you # can use as delimters. If you need three consecutive backticks # in your code, use four for delimiters, etc. # # * You can use spaces to get literal backticks at the edges: # # ... type `` `bar` `` ... # # Turns to: # # ... type `bar` ... # my $text = shift; $text =~ s@ (`+) # $1 = Opening run of ` (.+?) # $2 = The code block (?$c"; @egsx; return $text; } sub _EncodeCode { # # Encode/escape certain characters inside Markdown code runs. # The point is that in code, these characters are literals, # and lose their special Markdown meanings. # local $_ = shift; # Encode all ampersands; HTML entities are not # entities within a Markdown code span. s/&/&/g; # Encode $'s, but only if we're running under Blosxom. # (Blosxom interpolates Perl variables in article bodies.) s/\$/$/g if $_haveBX; # Do the angle bracket song and dance: s! < !<!gx; s! > !>!gx; # Now, escape characters that are magic in Markdown: s!([*_~{}\[\]\\])!$g_escape_table{$1}!g; return $_; } sub _DoItalicsAndBoldAndStrike { my $text = shift; my $doital1 = sub { my $text = shift; $text =~ s{ \* (?=\S) (.+?) (?<=\S) \* } {$1}gsx; # We've got to encode any of these remaining to # avoid conflicting with other italics and bold. $text =~ s!([*])!$g_escape_table{$1}!g; $text; }; my $doital2 = sub { my $text = shift; $text =~ s{ (?$1}gsx; # We've got to encode any of these remaining to # avoid conflicting with other italics and bold. $text =~ s!([_])!$g_escape_table{$1}!g; $text; }; # must go first: $text =~ s{ \*\* (?=\S) (.+?[*_]*) (?<=\S) \*\* } {"".&$doital1($1).""}gsex; $text =~ s{ (?".&$doital2($1).""}gsex; $text =~ s{ ~~ (?=\S) (.+?[*_]*) (?<=\S) ~~ } {$1}gsx; $text =~ s{ \* (?=\S) (.+?) (?<=\S) \* } {$1}gsx; $text =~ s{ (?$1}gsx; return $text; } sub _DoBlockQuotes { my $text = shift; $text =~ s{ ( # Wrap whole match in $1 ( ^[ ]*>[ ]? # '>' at the start of a line .*\n # rest of the first line (.+\n)* # subsequent consecutive lines \n* # blanks )+ ) }{ my $bq = $1; $bq =~ s/^[ ]*>[ ]?//gm; # trim one level of quoting $bq =~ s/^[ ]+$//mg; # trim whitespace-only lines $bq = _RunBlockGamut($bq); # recurse $bq =~ s/^/\027/mg; "

\n$bq\n

\n\n"; }egmx; return $text; } my ($LEAD, $TRAIL, $LEADBAR, $LEADSP, $COLPL, $SEP); BEGIN { $LEAD = qr/(?>[ ]*(?:\|[ ]*)?)/o; $TRAIL = qr/[ ]*(?[ ]*\|[ ]*)/o; $LEADSP = qr/(?>[ ]*)/o; $COLPL = qr/(?:[^\n|\\]|\\(?:(?>[^\n])|(?=\n|$)))+/o; $SEP = qr/[ ]*:?-+:?[ ]*/o; } sub _DoTables { my $text = shift; $text =~ s{ ( # Wrap whole thing to avoid $& (?: (?<=\n\n) | \A\n? ) # Preceded by blank line or beginning of string ^( # Header line $LEADBAR \| [^\n]* | $LEADBAR $COLPL [^\n]* | $LEADSP $COLPL \| [^\n]* )\n ( # Separator line $LEADBAR $SEP (?: \| $SEP )* (?: \| [ ]*)? | $SEP (?: \| $SEP )+ (?: \| [ ]*)? | $SEP \| [ ]* )\n ((?: # Rows (0+) $LEADBAR \| [^\n]* \n | $LEADBAR $COLPL [^\n]* \n | $LEADSP $COLPL \| [^\n]* \n )*) ) } { my ($w, $h, $s, $rows) = ($1, $2, $3, $4); my @heads = _SplitTableRow($h); my @seps = _SplitTableRow($s); if (@heads == @seps) { my @align = map { if (/^:-+:$/) {" align=\"center\""} elsif (/^:/) {" align=\"left\""} elsif (/:$/) {" align=\"right\""} else {""} } @seps; my $nohdr = ""; $nohdr = " $opt{style_prefix}table-nohdr" if join("", @heads) eq ""; my $tab ="\n\n"; $tab .= " " . _MakeTableRow("th", \@align, @heads) . "\n" unless $nohdr; my $cnt = 0; my @classes = ("class=\"$opt{style_prefix}row-even\"", "class=\"$opt{style_prefix}row-odd\""); $tab .= " " . _MakeTableRow("td", \@align, @$_) . "\n" foreach (_SplitMergeRows($rows)); $tab .= "

\n\n"; } else { $w; } }egmx; return $text; } sub _SplitMergeRows { my @rows = (); my ($mergeprev, $mergenext) = (0,0); foreach (split(/\n/, $_[0])) { $mergeprev = $mergenext; $mergenext = 0; my @cols = _SplitTableRow($_); if (_endswithbareslash($cols[$#cols])) { my $last = $cols[$#cols]; substr($last, -1, 1) = ""; $last =~ s/[ ]+$//; $cols[$#cols] = $last; $mergenext = 1; } if ($mergeprev) { for (my $i = 0; $i <= $#cols; ++$i) { my $cell = $rows[$#rows]->[$i]; defined($cell) or $cell = ""; $rows[$#rows]->[$i] = _MergeCells($cell, $cols[$i]); } } else { push(@rows, [@cols]); } } return @rows; } sub _endswithbareslash { return 0 unless substr($_[0], -1, 1) eq "\\"; my @parts = split(/\\\\/, $_[0], -1); return substr($parts[$#parts], -1, 1) eq "\\"; } sub _MergeCells { my ($c1, $c2) = @_; return $c1 if $c2 eq ""; return $c2 if $c1 eq ""; return $c1 . " " . $c2; } sub _SplitTableRow { my $row = shift; $row =~ s/^$LEAD//; $row =~ s/$TRAIL$//; $row =~ s!\\\\!$g_escape_table{'\\'}!go; # Must process escaped backslashes first. $row =~ s!\\\|!$g_escape_table{'|'}!go; # Then do \| my @elems = map { s!$g_escape_table{'|'}!|!go; s!$g_escape_table{'\\'}!\\\\!go; s/^[ ]+//; s/[ ]+$//; $_; } split(/[ ]*\|[ ]*/, $row, -1); @elems or push(@elems, ""); return @elems; } sub _MakeTableRow { my $etype = shift; my $align = shift; my $row = ""; for (my $i = 0; $i < @$align; ++$i) { my $data = $_[$i]; defined($data) or $data = ""; $row .= "<" . $etype . $$align[$i] . ">" . _RunSpanGamut($data) . ""; } return $row; } sub _FormParagraphs { # # Params: # $text - string to process with html

tags # my ($text, $anchors) = @_; # Strip leading and trailing lines: $text =~ s/\A\n+//; $text =~ s/\n+\z//; my @grafs = split(/\n{2,}/, $text); # # Wrap

tags. # foreach (@grafs) { unless (defined($g_html_blocks{$_}) || defined($g_code_blocks{$_})) { $_ = _RunSpanGamut($_); s/^([ ]*)/$g_start_p/; $_ .= $g_close_p; } } # # Strip standalone XML comments if requested # if ($anchors && $opt{stripcomments} && @g_xml_comments) { my %xml_comment = (); $xml_comment{$_} = 1 foreach @g_xml_comments; my @grafs2 = (); do { push(@grafs2, $_) unless $xml_comment{$_} } foreach @grafs; @grafs = @grafs2; } # # Unhashify HTML blocks # foreach (@grafs) { if (defined( $g_html_blocks{$_} )) { $_ = $g_html_blocks{$_}; } } return join "\n\n", @grafs; } # %ok_tag_name declared previously my $g_possible_tag_name; BEGIN { # note: length("blockquote") == 10 $g_possible_tag_name = qr/(?i:[a-z]{1,10}|h[1-6]|\020)/o; %ok_tag_name = map({$_ => 1} "\20", qw( a abbr acronym address area b basefont bdo big blockquote br caption center cite code col colgroup dd del dfn div dl dt em font h1 h2 h3 h4 h5 h6 hr i img ins kbd li map ol p pre q s samp small span strike strong sub sup table tbody td tfoot th thead tr tt u ul var )); $ok_tag_name{$_} = 0 foreach (qw( dir menu )); } sub _SetAllowedTag { my ($tag, $forbid) = @_; $ok_tag_name{$tag} = $forbid ? 0 : 1 if defined($tag) && exists($ok_tag_name{$tag}); } # Encode leading '<' of any non-tags # However, "]|/>$)} || $tag =~ m{^}) && $ok_tag_name{lc($1)}) { return _ProcessURLTag("href", $tag, 1) if $tag =~ /^]|/>$)} || $tag =~ m{^}) && $ok_tag_name{lc($1)}) { return ""; # strip it out } return $tag; }; $text =~ s{(<[^>]*>)}{&$_StripTag($1)}ige; return $text; } my %univatt; # universally allowed attribute names my %tagatt; # per-element allowed attribute names my %tagmt; # empty element tags my %tagocl; # non-empty elements with optional closing tag my %tagacl; # which %tagocl an opening %tagocl will close my %tagblk; # block elements my %taginl; # inline markup tags which trigger an auto

reopen my %taga1p; # open tags which require at least one attribute my %lcattval; # names of attribute values to lowercase my %impatt; # names of "implied" attributes BEGIN { %univatt = map({$_ => 1} qw(class dir id lang style title xml:lang)); %tagatt = ( 'a' => { map({$_ => 1} qw(href name rel target)) }, 'area' => { map({$_ => 1} qw(alt coords href nohref shape)) }, 'basefont' => { map({$_ => 1} qw(color face size)) }, 'br' => { map({$_ => 1} qw(clear)) }, 'caption' => { map({$_ => 1} qw(align)) }, 'col' => { map({$_ => 1} qw(align char charoff span width valign)) }, 'colgroup' => { map({$_ => 1} qw(align char charoff span width valign)) }, 'dir' => { map({$_ => 1} qw(compact)) }, 'div' => { map({$_ => 1} qw(align)) }, 'dl' => { map({$_ => 1} qw(compact)) }, 'font' => { map({$_ => 1} qw(color face size)) }, 'h1' => { map({$_ => 1} qw(align)) }, 'h2' => { map({$_ => 1} qw(align)) }, 'h3' => { map({$_ => 1} qw(align)) }, 'h4' => { map({$_ => 1} qw(align)) }, 'h5' => { map({$_ => 1} qw(align)) }, 'h6' => { map({$_ => 1} qw(align)) }, 'hr' => { map({$_ => 1} qw(align noshade size width)) }, # NO server-side image maps, therefore NOT ismap ! 'img' => { map({$_ => 1} qw(align alt border height hspace src usemap vspace width)) }, 'li' => { map({$_ => 1} qw(compact type value)) }, 'map' => { map({$_ => 1} qw(name)) }, 'menu' => { map({$_ => 1} qw(compact)) }, 'ol' => { map({$_ => 1} qw(compact start type)) }, 'p' => { map({$_ => 1} qw(align)) }, 'pre' => { map({$_ => 1} qw(width)) }, 'table' => { map({$_ => 1} qw(align bgcolor border cellpadding cellspacing frame rules summary width)) }, 'tbody' => { map({$_ => 1} qw(align char charoff valign)) }, 'tfoot' => { map({$_ => 1} qw(align char charoff valign)) }, 'thead' => { map({$_ => 1} qw(align char charoff valign)) }, 'td' => { map({$_ => 1} qw(align bgcolor char charoff colspan height nowrap rowspan valign width)) }, 'th' => { map({$_ => 1} qw(align bgcolor char charoff colspan height nowrap rowspan valign width)) }, 'tr' => { map({$_ => 1} qw(align bgcolor char charoff valign)) }, 'ul' => { map({$_ => 1} qw(compact type)) } ); %tagmt = map({$_ => 1} qw(area basefont br col hr img)); %tagocl = map({$_ => 1} qw(colgroup dd dt li p tbody td tfoot th thead tr)); %tagacl = ( 'colgroup' => \%tagocl, 'dd' => { map({$_ => 1} qw(colgroup dd dt li p)) }, 'dt' => { map({$_ => 1} qw(colgroup dd dt li p)) }, 'li' => { map({$_ => 1} qw(colgroup dd dt li p)) }, 'tbody' => \%tagocl, 'td' => { map({$_ => 1} qw(colgroup dd dt li p td th)) }, 'tfoot' => \%tagocl, 'th' => { map({$_ => 1} qw(colgroup dd dt li p td th)) }, 'thead' => \%tagocl, 'tr' => { map({$_ => 1} qw(colgroup dd dt li p td th tr)) }, ); %tagblk = map({$_ => 1} qw(address blockquote center div dl h1 h2 h3 h4 h5 h6 hr ol p pre table ul)); %taginl = map({$_ => 1} qw(a abbr acronym b basefont bdo big br cite code dfn em font i img kbd map q s samp small span strike strong sub sup tt u var)); %impatt = map({$_ => 1} qw(checked compact ismap nohref noshade nowrap)); %lcattval = map({$_ => 1} qw( align border cellpadding cellspacing checked clear color colspan compact coords height hspace ismap nohref noshade nowrap rowspan size span shape valign vspace width )); %taga1p = map({$_ => 1} qw(a area bdo img map)); } # _SanitizeTags # # Inspect all '<'...'>' tags in the input and HTML encode those things # that cannot possibly be tags and at the same time sanitize them. # # $1 => text to process # <= sanitized text sub _SanitizeTags { my ($text, $validate, $htmlauto) = @_; $text =~ s/\s+$//; $text ne "" or return ""; my @stack = (); my $ans = ""; my $end = length($text); pos($text) = 0; my ($autoclose, $autoclopen); my $lastmt = ""; my $reopenp = 0; $autoclose = $htmlauto ? sub { my $s = $_[0] || ""; while (@stack && ($stack[$#stack]->[0] ne $s || $_[1] && !$stack[$#stack]->[2]) && $tagocl{$stack[$#stack]->[0]}) { $ans .= "[0] . ">"; pop(@stack); } } : sub {} if $validate; $autoclopen = $htmlauto ? sub { my $s = $_[0] || ""; my $c; if ($tagblk{$s}) {$c = {p=>1}} elsif ($tagocl{$s}) {$c = $tagacl{$s}} else {return} my $clp = 0; while (@stack && $c->{$stack[$#stack]->[0]}) { $clp = 0; if ($stack[$#stack]->[2] && $stack[$#stack]->[1]+3 eq $_[1]) { $ans .= ""; } else { $ans .= "[0] . ">"; } if ($stack[$#stack]->[2]) { $stack[$#stack]->[0] = "\20"; } else { $clp = $s ne "p" && $stack[$#stack]->[0] eq "p"; pop(@stack); } } $clp; } : sub {} if $validate; while (pos($text) < $end) { if ($text =~ /\G(\s+)/gc) { $ans .= $1; next; } if ($text =~ /\G([^<]+)/gc) { if ($validate && @stack && $stack[$#stack]->[0] eq "\20") { push(@stack,["p",pos($text)-length($1)]); $reopenp = 0; $ans .= "

"; } $reopenp && do { push(@stack,["p",pos($text)-length($1)]); $reopenp = 0; $ans .= "

"; }; $ans .= _EncodeAmps($1); $lastmt = ""; next; } my $tstart = pos($text); if ($opt{stripcomments} != 2 && $text =~ /\G()/gc) { # pass "comments" through unless stripping them if ($opt{stripcomments} && $opt{stripcomments} < 3) { # strip any trailing whitespace + \n after comment if present $text =~ /\G[ \t]*\n/gc; } else { # pass the "comment" on through $ans .= $1; } next; } if ($opt{stripcomments} >= 2 && $text =~ /\G()/gc) { # strip any trailing whitespace + \n after lax comment if present $text =~ /\G[ \t]*\n/gc; next; } if ($text =~ /\G(<[^>]*>)/gc) { my $tag = $1; my $tt; if (($tag =~ m{^<($g_possible_tag_name)(?:[\s>]|/>$)} || $tag =~ m{^}) && $ok_tag_name{$tt=lc($1)}) { my ($stag, $styp, $autocloseflag) = _Sanitize($tag); if ($styp == 2 && $lastmt eq $tt) { $lastmt = ""; next; } $lastmt = $styp == -3 ? $tt : ""; $tt = "p" if $autocloseflag; if ($validate && $styp) { my $clp = &$autoclopen($tt, $tstart) if $styp != 2; if ($styp == 1) { $reopenp && $taginl{$tt} and do { push(@stack,["p",$tstart]); $ans .= "

"; }; push(@stack,[$tt,$tstart,$autocloseflag,$clp]); $reopenp = 0; } elsif ($styp == 2) { $reopenp && ($tt eq "p" || $tt eq "\20") and do { $reopenp = 0; next; }; &$autoclose($tt, $autocloseflag); my $mtstkchk = sub { !@stack and _xmlfail("closing tag $tt without matching open at " . _linecol($tstart, $text)); }; &$mtstkchk; if ($autocloseflag && $stack[$#stack]->[0] eq "\20") { pop(@stack); $stag = ""; } elsif ($stack[$#stack]->[0] eq $tt) { $stack[$#stack]->[3] and $reopenp = 1; pop(@stack); } else { pop(@stack) while @stack && $stack[$#stack]->[0] eq "\20"; &$mtstkchk; my @i = @{$stack[$#stack]}; _xmlfail("opening tag $i[0] at " . _linecol($i[1], $text) . " mismatch with closing tag $tt at " . _linecol($tstart, $text)); } } } $ans .= $stag; next; } else { $tag =~ s/^

sections $ans =~ s{

}{}gs if $validate; return $ans."\n"; } sub _linecol { my ($pos, $txt) = @_; pos($txt) = 0; my ($l, $p); $l = 1 + $opt{firstline}; ++$l while ($p = pos($txt)), $txt =~ /\G[^\n]*\n/gc && pos($txt) <= $pos; return "line $l col " . (1 + ($pos - $p)); } sub _xmlfail { die join("", map("$_\n", @_)); } sub _Sanitize { my $tag = shift; my $seenatt = {}; if ($tag =~ m{^" if $tag eq ""; $tag =~ tr/\t\n\f\r //d; # remove whitespace return (lc($tag),2,$autocloseflag); } if ($tag =~ /^<([^\s<\/>]+)\s+/gs) { my $tt = lc($1); my $autocloseflag = undef; $autocloseflag = 1, $tt="p" if $tt eq "\20"; my $out = "<" . $tt . " "; my $ok = $tagatt{$tt}; ref($ok) eq "HASH" or $ok = {}; my $atc = 0; while ($tag =~ /\G\s*([^\s\042\047<\/>=]+)((?>=)|\s*)/gcs) { my ($a,$s) = ($1, $2); if ($s eq "" && substr($tag, pos($tag), 1) =~ /^[\042\047]/) { # pretend the "=" sign wasn't overlooked $s = "="; } if (substr($s,0,1) ne "=") { # it's one of "those" attributes (e.g. compact) or not # _SanitizeAtt will fix it up if it is $out .= _SanitizeAtt($a, '""', $ok, $seenatt, $tt); ++$atc; next; } if ($tag =~ /\G([\042\047])((?:(?!\1)(?!<).)*)\1\s*/gcs) { $out .= _SanitizeAtt($a, $1.$2.$1, $ok, $seenatt, $tt); ++$atc; next; } if ($tag =~ /\G([\042\047])((?:(?!\1)(?![<>])(?![\/][>]).)*)/gcs) { # what to do what to do what to do # trim trailing \s+ and magically add the missing quote my ($q, $v) = ($1, $2); $v =~ s/\s+$//; $out .= _SanitizeAtt($a, $q.$v.$q, $ok, $seenatt, $tt); ++$atc; next; } if ($tag =~ /\G([^\s<\/>]+)\s*/gcs) { # auto quote it my $v = $1; $v =~ s/\042/"/go; $out .= _SanitizeAtt($a, '"'.$v.'"', $ok, $seenatt, $tt); ++$atc; next; } # give it an empty value $out .= _SanitizeAtt($a, '""', $ok, $seenatt, $tt); ++$atc; } my $sfx = substr($tag, pos($tag)); $out =~ s/\s+$//; my $typ = 1; if ($tagmt{$tt}) { $typ = ($tag =~ m,/>$,) ? 3 : -3; $out .= $opt{empty_element_suffix}; return ("<" . substr($tag,1), 0) if !$atc && $taga1p{$tt}; } else { if ($tag =~ m,/>$,) { return ("<" . substr($tag,1), 0) if !$atc && $taga1p{$tt}; $typ = 3; } else { return ("<" . substr($tag,1), 0) if !$atc && $taga1p{$tt}; } $out .= ">"; $out .= "" if $typ == 3; } return ($out,$typ,$autocloseflag); } elsif ($tag =~ /^<([^\s<\/>]+)/s) { my $tt = lc($1); return ("<" . substr($tag,1), 0) if $taga1p{$tt}; if ($tagmt{$tt}) { my $typ = ($tag =~ m,/>$,) ? 3 : -3; return ("<" . $tt . $opt{empty_element_suffix}, $typ); } elsif ($tag =~ m,/>$,) { return ("<" . $tt . ">", 3); } else { return ("<" . $tt . ">", 1) unless $tt eq "\20"; return ("

", 1, 1); } } return (lc($tag),0); } sub _SanitizeAtt { my $att = lc($_[0]); return "" unless $att =~ /^[_a-z:][_a-z:0-9.-]*$/; # no weirdo char att names return "" unless $univatt{$att} || $_[2]->{$att}; return "" if $_[3]->{$att}; # no repeats $_[3]->{$att} = 1; $impatt{$att} and return $att."=".'"'.$att.'" '; (($_[4] eq "a" && $att eq "href") || ($_[4] eq "img" && $att eq "src")) && $_[1] =~ /^\s*[\047\042]\s*javascript:/io and do { $_[1] = '"#"'; ref($opt{base_prefix}) eq 'CODE' and $_[1] = '"' . escapeXML(&{$opt{base_prefix}}("#")) . '"'; }; if ($_[4] eq "a") { $att eq "target" and return $_[1] =~ /^([\042\047])\s*_blank\s*\1$/io ? 'target="_blank" ' : ""; $att eq "rel" and return $_[1] =~ /^([\042\047])\s*nofollow\s*\1$/io ? 'rel="nofollow" ' : ""; } if ($lcattval{$att}) { return $att."="._SanitizeAttValue(lc($_[1]))." "; } else { my $satt = _SanitizeAttValue($_[1]); if (ref($opt{urlfunc}) eq 'CODE' && (($_[4] eq "a" && $att eq "href") || ($_[4] eq "img" && $att eq "src")) ) { my ($lq,$v,$rq); $lq = substr($satt, 0, 1); $rq = substr($satt, -1, 1); $v = unescapeXML(substr($satt, 1, length($satt)-2)); my ($uhost, $upath, $uq, $uf) = SplitURL($v); $v = &{$opt{urlfunc}}($v, \%opt, $_[4], $uhost, $upath, $uq, $uf); $satt = $lq . escapeXML($v) . $rq; } return $att."=".$satt." "; } } sub _SanitizeAttValue { my $v = shift; if ($v =~ /^([\042\047])(.*?)\1$/s) { return $1.escapeXML($2).$1; } else { return '"'.escapeXML($v).'"'; } } sub _ProcessURLTag { my ($att, $tag, $dofrag) = @_; $att = lc($att) . "="; if ($tag =~ /^(<[^\s>]+\s+)/g) { my $out = $1; while ($tag =~ /\G([^\s\042\047<\/>=]+=)([\042\047])((?:(?!\2)(?!<).)*)(\2\s*)/gcs) { my ($p, $q, $v, $s) = ($1, $2, $3, $4); if (lc($p) eq $att && $v ne "") { if ($dofrag && $v =~ m"^#") { $v = _FindFragmentMatch($v); my $bpcr; if (ref($bpcr = $opt{base_prefix}) eq 'CODE') { $v = "\2\3" . &$bpcr($v); } } else { $v = _PrefixURL($v); } $v = _EncodeAttText($v); } $out .= $p . $q . $v . $s; } $out .= substr($tag, pos($tag)); substr($out,0,1) = $g_escape_table{'<'}; substr($out,-1,1) = $g_escape_table{'>'}; return $out; } return $tag; } my $oops_entities; BEGIN { $oops_entities = qr/(?:lt|gt|amp|quot|apos|nbsp)/io; } # $_[0] => the value to XML escape # returns the XML escaped value # Encodes the five required entites (amp,lt,gt,quot,apos) # while preserving any pre-existing entities which means that # calling this repeatedly on already-escaped text should return # it unchanged (i.e. it's idempotent). # sub escapeXML { my $text = shift; # Treat these accidents as though they had the needed ';' $text =~ s/&($oops_entities)(?![A-Za-z0-9=;])/&$1;/go; # Ampersand-encoding based entirely on Nat Irons's Amputator MT plugin: # http://bumppo.net/projects/amputator/ $text =~ s/&(?!#?[xX]?(?:[0-9a-fA-F]+|\w+);)/&/g; # Remaining entities now $text =~ s/\042/"/g; $text =~ s/\047/'/g; # Some older browsers do not grok ' $text =~ s//>/g; return $text; } # $_[0] => value to be unescaped # returns unescaped value # The five required XML entities (amp,lt,gt,quot,apos) plus nbsp # are decoded as well as decimal (#d+) and hexadecimal (#xh+). # # While the escapeXML function tries to be idempotent when presented # with an already-escaped string, this function is NOT necessarily # idempotent when presented with an already decoded string unless it's # been decoded to the point there are no more recognizable entities left. # In other words given a string such as: # # &amp;amp; # # Each call will only decode one layer of escaping and it will take four # successive calls to finally end up with just "&". # sub unescapeXML { my $text = shift; # Treat these accidents as though they had the needed ';' $text =~ s/&($oops_entities)(?![A-Za-z0-9=;])/&$1;/go; $text =~ s/&[qQ][uU][oO][tT];/\042/gso; $text =~ s/&[aA][pP][oO][sS];/\047/gso; $text =~ s/&[gG][tT];/>/gso; $text =~ s/&[lL][tT];/ the input URL to split # $_[1] => if a true value, call unescapeXML before splitting # returns array: # [0] => scheme, name+password, host, port ("" if not present) # [1] => path in url (starts with "/" if absolute otherwise relative) # [2] => query string ("" if not present otherwise starts with "?") # [3] => fragment ("" if not present otherwise starts with "#") # The returned value recovers the (possibly unescapeXML'd) input # string by simply concatenating the returned array elements. # sub SplitURL { my ($url, $unesc) = @_; $unesc and $url = unescapeXML($url); my ($sh, $p, $q, $f) = ("", "", "", ""); if ($url =~ m{^([A-Za-z][A-Za-z0-9.+-]*:)(//.*)$}os) { $sh = $1; $url = $2; } if ($url =~ m{^(//[^/?#]*)((?:[/?#].*)?)$}os) { $sh .= $1; $url = $2; } ($p, $q, $f) = $url =~ m{^([^?#]*)((?:[?][^#]*)?)((?:[#].*)?)$}os; return ($sh, $p, $q, $f); } my $_replacesub; BEGIN { $_replacesub = sub { my $x = $named_character_entity{$_[1]}; $x ? '&#'.$x.';' : $_[0]; } } # $_[0] => the input text to process # returns text with all known named character entities replaced # with their equivalent numerical entity sub ConvertNamedCharacterEntities { use bytes; my $text = shift; defined($text) or return undef; $text =~ s/(&([A-Za-z]{3,8}[1-4]{0,2});)/&$_replacesub($1,$2)/goes; return $text; } sub _EncodeAmps { my $text = shift; # Treat these accidents as though they had the needed ';' $text =~ s/&($oops_entities)(?![A-Za-z0-9=;])/&$1;/go; # Ampersand-encoding based entirely on Nat Irons's Amputator MT plugin: # http://bumppo.net/projects/amputator/ $text =~ s/&(?!#?[xX]?(?:[0-9a-fA-F]+|\w+);)/&/g; return $text; } sub _EncodeAmpsAndAngles { # Smart processing for ampersands and angle brackets that need to be encoded. my $text = shift; # Treat these accidents as though they had the needed ';' $text =~ s/&($oops_entities)(?![A-Za-z0-9=;])/&$1;/go; # Ampersand-encoding based entirely on Nat Irons's Amputator MT plugin: # http://bumppo.net/projects/amputator/ $text =~ s/&(?!#?[xX]?(?:[0-9a-fA-F]+|\w+);)/&/g; # Encode naked <'s $text =~ s{<(?![\020a-z/?\$!])}{<}gi; $text =~ s{<(?=[^>]*$)}{<}g; # Encode <'s that cannot possibly be a start or end tag $text =~ s{(<[^>]*>)}{_DoTag($1)}ige; return $text; } sub _EncodeBackslashEscapes { # # Parameter: String. # Returns: String after processing the following backslash escape sequences. # local $_ = shift; s!\\\\!$g_escape_table{'\\'}!go; # Must process escaped backslashes first. s{\\([`*_~{}\[\]()>#+\-.!`])}{$g_escape_table{$1}}g; return $_; } sub _DoAutoLinks { local $_ = shift; s{<((https?|ftps?):[^'\042>\s]+)>(?!\s*)}{_MakeATag($1, "<".$1.">")}gise; # Email addresses: s{ < (?:mailto:)? ( [-.\w]+ \@ [-a-z0-9]+(\.[-a-z0-9]+)*\.[a-z]+ ) > }{ _EncodeEmailAddress(_UnescapeSpecialChars($1), "<", ">"); }egix; # (kjm) I don't do "x" patterns s{(?:^|(?<=\s))((?:https?|ftps?)://(?:[-a-zA-Z0-9./?\&\%=_~!*;:\@+\$,\x23](?:(?foo # @example.com # # Based on a filter by Matthew Wickline, posted to the BBEdit-Talk # mailing list: # my ($addr, $prefix, $suffix) = @_; $prefix = "" unless defined($prefix); $suffix = "" unless defined($suffix); srand(unpack('N',md5($addr))); my @encode = ( sub { '&#' . ord(shift) . ';' }, sub { '&#x' . sprintf( "%X", ord(shift) ) . ';' }, sub { shift }, ); $addr = "mailto:" . $addr; $addr =~ s{(.)}{ my $char = $1; if ( $char eq '@' ) { # this *must* be encoded. I insist. $char = $encode[int rand 1]->($char); } elsif ( $char ne ':' ) { # leave ':' alone (to spot mailto: later) my $r = rand; # roughly 10% raw, 45% hex, 45% dec $char = ( $r > .9 ? $encode[2]->($char) : $r < .45 ? $encode[1]->($char) : $encode[0]->($char) ); } $char; }gex; # strip the mailto: from the visible part (my $bareaddr = $addr) =~ s/^.+?://; $addr = _MakeATag("$addr", $prefix.$bareaddr.$suffix); return $addr; } sub _UnescapeSpecialChars { # # Swap back in all the special characters we've hidden. # my $text = shift; while( my($char, $hash) = each(%g_escape_table) ) { $text =~ s/$hash/$char/g; } return $text; } sub _TokenizeHTML { # # Parameter: String containing HTML markup. # Returns: Reference to an array of the tokens comprising the input # string. Each token is either a tag (possibly with nested, # tags contained therein, such as , or a # run of text between tags. Each element of the array is a # two-element array; the first is either 'tag' or 'text'; # the second is the actual value. # # # Derived from the _tokenize() subroutine from Brad Choate's MTRegex plugin. # # my $str = shift; my $pos = 0; my $len = length $str; my @tokens; my $depth = 6; my $nested_tags = join('|', ('(?:<[a-z/!$](?:[^<>]') x $depth) . (')*>)' x $depth); my $match = qr/(?s: ) | # comment (?s: <\? .*? \?> ) | # processing instruction $nested_tags/iox; # nested tags while ($str =~ m/($match)/g) { my $whole_tag = $1; my $sec_start = pos $str; my $tag_start = $sec_start - length $whole_tag; if ($pos < $tag_start) { push @tokens, ['text', substr($str, $pos, $tag_start - $pos)]; } push @tokens, ['tag', $whole_tag]; $pos = pos $str; } push @tokens, ['text', substr($str, $pos, $len - $pos)] if $pos < $len; \@tokens; } sub _Outdent { # # Remove one level of line-leading indent_width of spaces # my $text = shift; $text =~ s/^ {1,$opt{indent_width}}//gm; return $text; } # _DeTab # # $1 => input text # $2 => optional tab width (default is $opt{tab_width}) # $3 => leading spaces to strip off each line first (default is 0 aka none) # <= result with tabs expanded sub _DeTab { my $text = shift; my $ts = shift || $opt{tab_width}; my $leadsp = shift || 0; my $spr = qr/^ {1,$leadsp}/ if $leadsp; pos($text) = 0; my $end = length($text); my $ans = ""; while (pos($text) < $end) { my $line; if ($text =~ /\G(.*?\n)/gcs) { $line = $1; } else { $line = substr($text, pos($text)); pos($text) = $end; } $line =~ s/$spr// if $leadsp; # From the Perl camel book section "Fluent Perl" but modified a bit $line =~ s/(.*?)(\t+)/$1 . ' ' x (length($2) * $ts - length($1) % $ts)/ges; $ans .= $line; } return $ans; } sub _PrefixURL { # # Add URL prefix if needed # my $url = shift; $url =~ s/^\s+//; $url =~ s/\s+$//; $url = "#" unless $url ne ""; return $url unless ref($opt{abs_prefix}) eq 'CODE' || ref($opt{url_prefix}) eq 'CODE' || ref($opt{img_prefix}) eq 'CODE' ; return $url if $url =~ m"^\002\003" || $url =~ m"^#" || $url =~ m,^//,; $url = &{$opt{abs_prefix}}($url) if $url =~ m,^/, && ref($opt{abs_prefix}) eq 'CODE'; return $url if $url =~ /^[A-Za-z][A-Za-z0-9+.-]*:/ || $url =~ m,^//, || ($opt{keepabs} && $url =~ m,^/,); my $cr = $opt{url_prefix}; $cr = $opt{img_prefix} if ref($opt{img_prefix}) eq 'CODE' && $url =~ m"^[^#?]*\.(?:png|gif|jpe?g|svgz?)(?:[#?]|$)"i; return $url unless ref($cr) eq 'CODE'; return "\2\3".&$cr(substr($url, 0, 1) eq '/' ? substr($url, 1) : $url); } BEGIN { $g_style_sheet = <<'STYLESHEET'; STYLESHEET $g_style_sheet =~ s/^\s+//g; $g_style_sheet =~ s/\s+$//g; $g_style_sheet .= "\n"; } 1; __DATA__ =head1 NAME Markdown.pl - convert Markdown format text files to HTML =head1 SYNOPSIS B [B<--help>] [B<--html4tags>] [B<--htmlroot>=I] [B<--imageroot>=I] [B<--version>] [B<--shortversion>] [B<--tabwidth>=I] [B<--stylesheet>] [B<--stub>] [--] [I...] Options: -h show short usage help --help show long detailed help --html4tags use
instead of
--deprecated allow

and tags --sanitize sanitize tag attributes --no-sanitize do not sanitize tag attributes --validate-xml check if output is valid XML --validate-xml-internal fast basic check if output is valid XML --no-validate-xml do not check output for valid XML --strip-comments remove XML comments from output --strip-comments-lax remove XML-like comments from output --strip-comments-lax-only remove only invalid XML-like comments --no-strip-comments do not remove XML comments (default) --tabwidth=num expand tabs to num instead of 8 --auto-number automatically number h1-h6 headers -k | --keep-abs keep abspath URLs despite -r/-i -a prefix | --absroot=prefix append abspath URLs to prefix -b prefix | --base=prefix prepend prefix to fragment-only URLs -r prefix | --htmlroot=prefix append relative non-img URLs to prefix -i prefix | --imageroot=prefix append relative img URLs to prefix -w [wikipat] | --wiki[=wikipat] activate wiki links using wikipat --yaml[=(enable|disable|strip|...)] select YAML front matter processing -V | --version show version, authors, license and copyright -s | --shortversion show just the version number --raw | --raw-xml input contains only raw xhtml --raw-html input contains only raw html --div[=id] wrap body in div with given id --stylesheet output the fancy style sheet --no-stylesheet do not output fancy style sheet --keep-named-character-entities do not convert named character entities --stub wrap output in stub document implies --stylesheet -- end options and treat next argument as file =head1 DESCRIPTION Markdown is a text-to-HTML filter; it translates an easy-to-read / easy-to-write structured text format into HTML. Markdown's text format is most similar to that of plain text email, and supports features such as headers, *emphasis*, code blocks, blockquotes, and links. Markdown's syntax is designed not as a generic markup language, but specifically to serve as a front-end to (X)HTML. You can use span-level HTML tags anywhere in a Markdown document, and you can use block level HTML tags (like

and as well). For more information about Markdown's syntax, see the F and F files included with F. Input (auto-detected) may be either ISO-8859-1 or UTF-8. Output is always converted to the UTF-8 character set. =head1 OPTIONS Use "--" to end switch parsing. For example, to open a file named "-z", use: Markdown.pl -- -z =over =item B<--html4tags> Use HTML 4 style for empty element tags, e.g.:
instead of Markdown's default XHTML style tags, e.g.:
This option is I with the B<--validate-xml> option and will produce an immediate error if both are given. =item B<--deprecated> Both "" and "" are normally taken as literal text and the leading "<" will be automatically escaped. If this option is used, they are recognized as valid tags and passed through without being escaped. When dealing with program argument descriptions "" can be particularly problematic therefore use of this option is not recommended. Other deprecated tags (such as "" and "" for example) continue to be recognized and passed through even without using this option. =item B<--sanitize> Removes troublesome tag attributes from embedded tags. Only a very strictly limited set of tag attributes will be permitted, other attributes will be silently discarded. The set of allowed attributes varies by tag. Splits empty minimized elements that are not one of the HTML allowed empty elements (C C C
CC

C) into separate begin and end tags. For example, C<<

>> or C<<

>> will be split into C<<

>>. Combines adjacent (whitespace separated only) opening and closing tags for the same HTML empty element into a single minimized tag. For example, C<<

>> will become C<<
>>. Tags that require at least one attribute to be present to be meaningful (e.g. C, C, C, C) but have none will be treated as non-tags potentially creating unexpected errors. For example, the sequence C<< text here >> will be sanitized to C<< <a>text here >> since an C tag without any attributes is meaningless, but then the trailing close tag C<< >> will become an error because it has no matching open C<< >> tag. The point of this check is not to cause undue frustration, but to allow such constructs to be used as text without the need for escaping since they are meaningless as tags. For example, C<< >> works just fine as plain text and so does C<< >> because the C<< >>/C<< >> will be treated as a non-tag automatically. In fact, they can even appear inside links too such as C<< Link to article >>. Problematic C<&> characters are fixed up such as standalone C<&>s (or those not part of a valid entity reference) are turned into C<&>. Within attribute values, single and double quotes are turned into C<&> entity refs. This is enabled by default. =item B<--no-sanitize> Do not sanitize tag attributes. This option does not allow any tags that would not be allowed without this option, but it does completely suppress the attribute sanitation process. If this option is specified, no attributes will be removed from any tag (although C and C tags will still be affected by B<--imageroot>, B<--htmlroot>, B<--absroot> and/or B<--base> options). Use of this option is I. =item B<--validate-xml> Perform XML validation on the output before it's output and die if it fails validation. This requires the C or C module be present (one is only required if this option is given). Any errors are reported to STDERR and the exit status will be non-zero on XML validation failure. Note that all line and column numbers in the error output refer to the entire output that would have been produced. Re-run with B<--no-validate-xml> to see what's actually present at those line and column positions. If the B<--stub> option has also been given, then the entire output is validated as-is. Without the B<--stub> option, the output will be wrapped in C<<

...

>> for validation purposes but that extra "div" added for validation will not be added to the final output. This option is I. This option is I with the B<--html4tags> option and will produce an immediate error if both are given. =item B<--validate-xml-internal> Perform XML validation on the output before it's output and die if it fails validation. This uses a simple internal consistency checker that finds unmatched and mismatched open/close tags. Non-empty elements that in HTML have optional closing tags (CC

C C C) will automatically have any omitted end tags inserted during the `--validate-xml-internal` process. Any errors are reported to STDERR and the exit status will be non-zero on XML validation failure. Note that all line and column numbers in the error output refer to the entire output that would have been produced before sanitization without any B<--stub> or B<--stylesheet> options. Re-run with B<--no-sanitize> and B<--no-validate-xml> and I any B<--stub> or B<--stylesheet> options to see what's actually present at those line and column positions. This option validates the output I adding any requested B<--stub> or B<--stylesheet>. As the built-in stub and stylesheet have already been validated that speeds things up. The output is I wrapped (in a C<<

...

>>) for validation as that's not required for the internal checker. This option is I unless B<--no-sanitize> is active. This option is I with the B<--html4tags> option. This option requires the B<--sanitize> option and will produce an immediate error if both B<--no-sanitize> and B<--validate-xml-internal> are given. Note that B<--validate-xml-internal> is I than B<--validate-xml> and I require any extra XML modules to be present. =item B<--no-validate-xml> Do not perform XML validation on the output. Markdown.pl itself will normally generate valid XML sequences (unless B<--html4tags> has been used). However, any raw tags in the input (that are on the "approved" list), could potentially result in invalid XML output (i.e. mismatched start and end tags, missing start or end tag etc.). Markdown.pl will I for these issues itself. But with the B<--validate-xml> option will use C or C to do so. Note that B<--validate-xml-internal> is the default option unless B<--no-sanitize> is used in which case B<--no-validate-xml> is the default option. =item B<--strip-comments> Strip XML comments from the output. Any XML comments encountered will be omitted from the output if this option is given. This option requires the B<--sanitize> option to be used (which is the default). However, note that the XML standard section 2.5 specifically prohibits a C<--> sequence within an XML comment (i.e. C<--> cannot occur after the comment start tag C<< >>). In other words, S >>>, S >>>, S >>>, and S >>> are all valid XML comments, but S >>> and S >>> are not! As part of the "sanitation" process (triggered by the B<--sanitize> option), any invalid tags have their leading C<< < >> escaped (to C<< &#lt; >>) thus making them ordinary text and this I invalid XML comments. What this means is that the B<--strip-comments> option I remove invalid XML comments (such as S >>>)! But see the B<--strip-comments-lax> option for a solution. =item B<--strip-comments-lax> Strip XML-like comments from the output. Any XML-like comments encountered will be omitted from the output if this option is given. Supersedes the B<--strip-comments> option if both are given. While the syntax of XML comments cannot be relaxed (that would require altering the XML standard), if they are being stripped out anyway, then the standard isn't quite so relevant since they will not be present in the output. The B<--strip-comments-lax> option acts just like the B<--strip-comments> option EXCEPT that the content between the starting comment tag S>> and then ending comment tag S >>> is I restricted since it will be stripped out of the final result which will therefore remain XML compliant. The only restriction, of course, is that the content between the XML comment start tag and the XML comment end tag cannot contain the XML comment end tag itself. With the B<--strip-comments-lax> option, strictly invalid XML comments (such as S >>>) I be stripped as well as all strictly valid XML comments. =item B<--strip-comments-lax-only> This is a compromise option. It works just like B<--strip-comments-lax>, but I on strictly invalid XML-like comments. Supersedes the B<--strip-comments> option if both are given. In other words, if a strictly valid XML comment is present, it will be retained in the output. If a strictly invalid XML comment is present which would have been stripped by B<--strip-comments-lax> but would have had its leading C<< < >> escaped automatically by the default B<--no-strip-comments> mode (because it's not a strictly valid XML comment), then it will be stripped by this mode. This option avoids ugly invalid XML comments from slipping through into the output as escaped plain text while still passing through valid XML comments without stripping them. =item B<--no-strip-comments> Do not strip XML comments from the output. This is the default. =item B<--tabwidth>=I Expand tabs to I character wide tab stop positions instead of the default 8. Don't use this; physical tabs should always be expanded to 8-character positions. This option does I affect the number of spaces needed to start a new "indent level". That will always be 4 no matter what value is used (or implied by default) with this option. Also note that tabs inside backticks-delimited code blocks will always be expanded to 8-character tab stop positions no matter what value is used for this option. The value must be S<2 <= I <= 32>. =item B<--auto-number> Automatically number all h1-h6 headings generated from Markdown markup. Explicit C<<

>> ... C<<

>> tag content remains unmolested. If this option is given, any YAML C setting will be ignored. =item B<-k>, B<--keep-abs> Normally any absolute path URLs (i.e. URLs without a scheme starting with "/" but not "//") are subject to modification by any B<-r>/B<--htmlroot> or B<-i>/B<--imageroot> option. If the B<-a>/B<--absroot> option is used and it transforms these absolute path URLs into a full absolute URL (i.e. starts with a scheme or "//") then any subsequent B<-r>/B<--htmlroot> or B<-i>/B<--imageroot> processing will be skipped because the URL is no longer relative. If the B<--keep-abs> option is given, then (after applying any B<-a>/B<--absroot> option if present) absolute path URLs will be kept as-is and will not be processed further by any B<-r>/B<--htmlroot> or B<-i>/B<--imageroot> option. Note that if the B<-a>/B<--absroot> option transforms an absolute path URL into a relative PATH URL it I be subject to subsequent B<-r>/B<--htmlroot> or B<-i>/B<--imageroot> processing regardless of the B<-k>/B<--keep-abs> option. =item B<-a> I, B<--absroot>=I Any absolute path URLs (i.e. URLs without a scheme starting with "/" but not "//") have I prepended which prevents them from being acted upon by the B<--htmlroot> and/or B<--imageroot> options provided the result is a full absolute URL. The default is to prepend nothing and leave them as absolute path URLs which will allow them to be processed by any B<--htmlroot> and/or B<--imageroot> options. This option can be helpful when documents are being formatted for display on a different system and the absolute path URLs need to be "fixed up". =item B<-b> I, B<--base>=I Any fragment-only URLs have I prepended. The default is to prepend nothing and leave them as bare fragment URLs. Use of this option may be necessary when embedding the output of Markdown.pl into a document that makes use of the C<< >> tag in order for intra-document fragment URL links to work properly in such a document. =item B<-r> I, B<--htmlroot>=I Any non-absolute URLs have I prepended. =item B<-i> I, B<--imageroot>=I Any non-absolute URLs have I prepended (overriding the B<-r> prefix if any) but only if they end in an image suffix. =item B<-w> [I], B<--wiki>[=I] Activate wiki links. Any link enclosed in double brackets (e.g. "[[link]]") is considered a wiki link. By default only absolute URL and fragment links are allowed in the "wiki link style" format. Any other double-bracketed strings are left unmolested. If this option is given, all other wiki links are enabled as well. Any non-absolute URL or fragment links will be transformed into a link using I where the default I if none is given is C<%{s(:md)}.html>. If the given I does not contain a C<%{...}> placeholder sequence then it will automatically have C<%{s(:md)}.html> suffixed to it. The C<...> part of the C<%{...}> sequence specifies zero or more case-insensitive single-letter options with the following effects: =over =item B Retain blanks (aka spaces) in the output. They will become C<%20> in the final URL. Because spaces are always trimmed before processing wiki links, runs of multiple spaces will be collapsed into a single space and any leading or trailing spaces will be removed. =item B Convert spaces to dashes (ASCII 0x2D) instead of underscore (ASCII 0x5F). Note that if this option is given then runs of multiple dashes will be converted to a single dash I but runs of multiple underscores will be left untouched. =item B Flatten the resulting name by replacing forward slashes (ASCII 0x2F) as well. They will be converted to underscores unless the C option is given (in which case they will be converted to dashes). This conversion takes place before applying the runs-of-multiple reduction. This option is incompatible with the B<%> option. =item B<%> Flatten the resulting name by replacing runs of one or more forward slashes (ASCII 0x2F) with C<%2F>. Note that when encoded into a URL the C<%2F> actually becomes C<%252F>. This option is incompatible with the B option. =item B Convert link target (excluding any query string and/or fragment) to lowercase. Takes precedence over any C option, but specifically excludes C<%>-escapes which are always UPPERCASE hexadecimal. =item B Leave raw UTF-8 characters in the result. Normally anything not allowed directly in a URL ends up URL-encoded. With this option, raw valid UTF-8 sequences will be left untouched. Use with care. =item B or BI<< >>[B<,>I<< >>]...B<)> After (temporarily) removing any query string and/or fragment, strip any final "dot" suffix so long as it occurs after the last slash (if any slash was present before applying the C option). The "dot" (ASCII 0x2E) and all following characters (if any) are removed. If the optional C<< (,...) >> part is present then only strip the extension if it consists of a "dot" followed by one of the case-insensitive I<< >> values. As a special case, using the value C<:md> for one of the I<< >> values causes that value to be expanded to all known markdown extensions. When processing wiki image links, this option is ignored. =item B Convert link target (excluding any query string and/or fragment) to UPPERCASE. =item B Leave runs-of-multiple characters alone (aka "verbatim"). Does not affect any of the other options except by eliminating the runs-of-multple reduction step. Also does I inhibit the initial whitespace trimming. Does not affect the runs-of-multiple "/" replacement performed by the B<%> option. =back The URL target of the wiki link is created by first trimming whitespace (starting and ending whitespace is removed and all other runs of consecutive whitespace are replaced with a single space) from the wiki link target, removing (temporarily) any query string and/or fragment, if no options are present, spaces are converted to underscores (C<_>) and runs of multiple consecutive underscores are replaced with a single underscore (ASCII 0x5F). Finally, the I string gets its first placeholder (the C<%{...}> sequence) replaced with this computed value and the original query string and/or fragment is re-appended (if any were originally present) and URL-encoding is applied as needed to produce the actual final target URL. Note that when processing wiki image links, no extension stripping ever takes place (i.e. the "s" option is ignored) and anything after the placeholder (the C<%{...}> sequence) in the pattern is omitted from the result. See above option descriptions for possible available modifications. One of the commonly used hosting platforms does something substantially similar to using C<%{dfv}> as the placeholder. One of the commonly used wiki platforms does something similar to using C<%{%}> as the placeholder. =item B<--yaml>[=I] Select YAML front matter processing. The optional I value must be one of the following: =over =item B Recognize any YAML front matter and apply any options specified therein. If any unrecognized options are present, the options will also be shown in the formatted output. This is the default I if omitted. =item B No YAML front matter processing at all takes place. If YAML front matter is present, it will be treated as regular non-YAML markup text to be processed. =item B If YAML front matter is present, it will be stripped and completely ignored before beginning to process the rest of the input. In this mode, any options in the YAML front matter that would have otherwise been recognized will I =item B If YAML front matter is present and contains anything other than comments, the non-comments parts will be shown in the formatted output. In this mode, any options in the YAML front matter that would have otherwise been recognized will I This is a show-only mode. =item B This mode works just like the B mode except that if the YAML front matter contains anything other than comments, then I of the non-comments parts will be shown in the formatted output. In this mode, any recognized options in the YAML front matter I processed the same way they would be in the B mode except that any option to suppress the B mode is ignored. =item B This mode works just like the B mode except that no options are ever shown in the formatted output regardless of whether or not there are any unrecognized options present. In this mode, any recognized options in the YAML front matter I processed the same way they would be in the B mode except that any option to suppress the B mode is ignored. =item B This mode works just like the B mode if any unrecognized YAML front matter options are present. Otherwise it works like the B mode. In this mode, any options in the YAML front matter that would have otherwise been recognized will I =back If B<--raw>, B<--raw-xml> or B<--raw-html> has been specified then the default if no B<--yaml> option has been given is B<--yaml=disable>. Otherwise the default if no B<--yaml> option has been given is B<--yaml=enable>. Note that only a limited subset of YAML is recognized. Specifically only comments, and top-level single-line S> items where key must be plain (i.e. non-quoted), start with a letter or underscore and contain only letters, underscores, hyphens (C<->), periods (C<.>) and digits. Keys are case-insensitive (i.e. converted to lowercase). As with YAML, at least one whitespace is required between the ":" and the value (unless it's the empty value). Values may be either plain or double-quoted (single-quoted is not recognized). The double-quoted style may use C-style character escape codes but may not extend past the end of the line. For YAML front matter to be recognized, the very first line of the document must be exactly three hyphens (C<--->). The YAML terminates when a line of three hyphens (C<--->) or a line of three periods (C<...>) or the end of the file is encountered. Of course the YAML mode must also be something I than B<--yaml=disable>. =item B<-V>, B<--version> Display Markdown's version number and copyright information. =item B<-s>, B<--shortversion> Display the short-form version number. =item B<--raw>, B<--raw-xml> Input contains only raw XHTML. All options other than B<--html4tags>, B<--deprecated>, B<--sanitize> (on by default), B<--strip-comments>, B<--div>, B<--keep-named-character-entities>, B<--validate-xml> and B<--validate-xml-internal> (and their B<--no-...> variants) are ignored. With this option, arbitrary XHTML input can be passed through the sanitizer and/or validator. If sanitation is requested (the default), input must only contain the contents of the "" section (i.e. no "" or ""). Output I be converted to UTF-8 regardless of the input encoding. All line endings will be normalized to C<\n> and input encodings other than UTF-8 or ISO-8859-1 or US-ASCII will end up mangled. Remember that any B<--stub> and/or B<--stylesheet> options are I when B<--raw> is given. =item B<--raw-html> Input contains only raw HTML. All options other than B<--html4tags>, B<--deprecated>, B<--sanitize> (on by default), B<--strip-comments>, and B<--validate-xml-internal> (and their B<--no-...> variants) are ignored. Requires the (possibly implicit) B<--validate-xml-internal> option. Works just like B<--raw-xml> except that HTML auto closing and optional closing tag semantics are activated during the validation causing missing closing tags to be inserted where required by the standard. Non-raw mode always enables these semantics. This will transform HTML into valid XHTML or fail with an error message. Unfortunately, it will also fail to accept some documents that the plain B<--raw-xml> option will. For example, this document:
a
Will be rejected because upon encountering the C<<
>> open tag a closing C<< >> will automatically be inserted resulting in this document:
a
Which, of course, no longer validates. Since C
blocks cannot directly be nested within C
blocks (according to the standard), the input document is not strictly correct. Remember that any B<--stub> and/or B<--stylesheet> options are I when B<--raw-html> is given. =item B<--div>[=I] Wrap the output contents in a C
tag. If I is given the tag will have that C attribute value. If the B<--stub> option and/or the B<--stylesheet> option are active, they are applied I wrapping the output contents in the C
. Note that if a YAML table ends up being generated, it I be included I the C
wrapper. In contrast to the B<--stylesheet> and B<--stub> options, this option I allowed with the B<--raw-xml> and B<--raw-html> options. =item B<--stylesheet> Include the fancy style sheet at the beginning of the output (or in the C section with B<--stub>). This style sheet makes fancy checkboxes and makes a right parenthesis C<)> show instead of a C<.> for ordered lists that use them. Without it things will still look fine except that the fancy stuff won't be there. Use this option with no other arguments and redirect standard input to /dev/null to get just the style sheet and nothing else. =item B<--no-stylesheet> Overrides a previous B<--stylesheet> and disables implicit inclusion of the style sheet by the B<--stub> option. =item B<--keep-named-character-entities> Do not convert named character entities to their equivalent numerical character entity. Normally any occurrence of a named character entity such as C<…> would be converted to its equivalent character entity such as C<…>. If this option is given, that conversion is suppressed. The only always-valid named entities as far as XML is concerned are the five entities C<&>, C<<>, C<>>, C<"> and C<'>. Even that last one (C<'>) may not be universally supported in XHTML user agents (and it is converted to C<'> for that reason unless this option is given). Regardless of this option, C<&>, C<<>, C<>> and C<"> are always left alone since they are universally supported. Use of this option is I. =item B<--stub> Wrap the output in a full document stub (i.e. has C, C and C tags). The style sheet I be included in the C section unless the B<--no-stylesheet> option is also used. The C<< >> value for a document produced with the B<--stub> option comes from the first markdown markup C<h1> that's generated unless YAML processing has been enabled (the default) and a C<title> YAML value has been set in which case that always takes precedence. =item B<-h>, B<--help> Display Markdown's help. With B<--help> full help is shown, with B<-h> only the usage and options are shown. =back =head1 HTML CONTENT Markdown format documents are intended to be human readable without the use of XML-like markup. Nevertheless, html content can be included verbatim provided that the tags used are limited to those of the HTML 4 specification and only those tags that represent body content -- scripting tags and attributes are not allowed. The final version of the HTML specification (including a DTD) can be found here: =over =over =item L<https://www.w3.org/TR/1999/REC-html401-19991224/> =back =back Note that attempts to use any of the new tags from the "HTML Living Standard" will simply result in them being escaped into literal text. Stick to markdown-format text or HTML 4 tags to avoid unexpected output. =head1 PERL MODULE Markdown can be used as a Perl module and can be "use"d like so: use Markdown qw(...); Or like so: BEGIN {require "Markdown.pl" && Markdown->import(qw(...))} where the C<...> part is the list of desired imports. The Markdown module does not export any functions by default. The C<Markdown.pm> file is a symbolic link to C<Markdown.pl>. =head2 Markdown module functions Any of these functions may be imported, but none of them are imported by default. =over =item * $result = Markdown::Markdown($string[, options...]) Converts Markdown-format C<$string> to UTF-8 encoded XHTML and returns it. The C<options...> may be either a single HASH ref or one or more pairs of C<< key => value >>. See the comments for the C<_SanitizeOpts> function for a list of possible option keys. =item * $result = Markdown::ProcessRaw($string[, options...]) Converts raw XHTML in C<$string> to XHTML and returns it. The C<options...> may be either a single HASH ref or one or more pairs of C<< key => value >>. See the comments for the C<_SanitizeOpts> function for a list of possible option keys. This function provides the ability to apply the internal XML validation and sanitation functionality to arbitrary XHTML without performing any of the Markdown format interpretation. =item * $stylesheet = Markdown::GenerateStyleSheet([$prefix]) Returns an XHTML style sheet that supports the fancy Markdown styles such as checkboxes and right parenthesis lists. All of the style class names have C<$prefix> prepended. If C<$prefix> is omitted or C<undef> then the default S<"_markdown-"> prefix will be used which is the same default prefix that the C<Markdown> function uses. The returned string value consists of a C<< <style type="text/css"> >> tag, the contents of the style sheet and ends with a C<< </style> >> tag. =item * Markdown::SetWikiOpts($hashref, $wikioption) The value of C<$wikioption> should be the value of the C<wikipat> value from the B<--wiki> option. Use the empty string S<""> to enable wiki links using the defaults and use C<undef> to disable wiki links. The C<wikipat> and C<wikiopt> keys in C<$hashref> will both be affected by this call and they should be passed in to the Markdown function as options to enable processing of wiki links. The simplest way to do this is simply to pass a HASH ref as the second argument to the Markdown function after having used this function on it to properly set the C<wikipat> and C<wikiopt> keys and values. =back =head2 Example This rudimentary example approximates running S<C<Markdown.pl --stub --wiki>> on the input (files if given, standard input if not). use Markdown qw(Markdown SetWikiOpts GenerateStyleSheet escapeXML); my $string; {local $/; $string = <>;} my %opts = ( h1 => "default title" ); SetWikiOpts(\%opts, ""); # enable default --wiki processing my $xhtml = Markdown($string, \%opts); print "<html xmlns=\"http://www.w3.org/1999/xhtml\">\n", "<head>\n<title>".escapeXML($opts{h1})."\n", GenerateStyleSheet(),"\n", "\n", "
\n", $xhtml, "
\n"; =head1 VERSION HISTORY Z<> See the F file for detailed release notes for this version. =over =item Z<> 1.1.11 - 05 Mar 2021 =item Z<> 1.1.10 - 08 Jul 2020 =item Z<> 1.1.9 - 15 Dec 2019 =item Z<> 1.1.8 - 22 Nov 2019 =item Z<> 1.1.7 - 14 Feb 2018 =item Z<> 1.1.6 - 03 Jan 2018 =item Z<> 1.1.5 - 07 Dec 2017 =item Z<> 1.1.4 - 24 Jun 2017 =item Z<> 1.1.3 - 13 Feb 2017 =item Z<> 1.1.2 - 19 Jan 2017 =item Z<> 1.1.1 - 12 Jan 2017 =item Z<> 1.1.0 - 11 Jan 2017 =item Z<> 1.0.4 - 05 Jun 2016 =item Z<> 1.0.3 - 06 Sep 2015 =item Z<> 1.0.2 - 03 Sep 2015 =item Z<> 1.0.1 - 14 Dec 2004 =item Z<> 1.0.0 - 28 Aug 2004 =back =head1 AUTHORS =over =item John Gruber =item L =item L =item E<160> =back =over =item PHP port and other contributions by Michel Fortin =item L =item E<160> =back =over =item Additional enhancements and tweaks by Kyle J. McKay =item mackylegmail.com =item L =back =head1 COPYRIGHT AND LICENSE =over =item Copyright (C) 2003-2004 John Gruber =item Copyright (C) 2015-2021 Kyle J. McKay =item All rights reserved. =back Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: =over =item * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. =item * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. =item * Neither the name "Markdown" nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission. =back THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. =cut

C
C