<?php
$er = error_reporting(E_ALL & ~E_NOTICE);
#StructuredText parser, by Keith Devens, version .99g.
#http://www.keithdevens.com/software/
#See release notes and To Do at http://www.keithdevens.com/software/markup_history.php
#
#This code is not being developed anymore. It still works and is still in use on my site,
# but I've begun development of version 2, which will replace this version when its released.
# Because this code is monolithic and not all that great, I don't even necessarily want my name
# associated with it - so, I declare it public domain. Do what you want with it, remove my name
# if you release it with something else (please) :), etc.

if(isset($old_error_reporting)) error_reporting($old_error_reporting);

define("MARKUP_STATE_NOTHING",    0);
define("MARKUP_STATE_LIST",       1);
define("MARKUP_STATE_BLOCKQUOTE", 2);
define("MARKUP_STATE_PARA",       3);
define("MARKUP_STATE_TABLE",      4);
define("MARKUP_STATE_HR",         5);
define("MARKUP_STATE_DIRECTIVE",  6);
define("MARKUP_STATE_HEADING",    7);
define("MARKUP_STATE_PRE",        8);
define("MARKUP_STATE_RAW",        9);

global $markup_list_types, $markup_list_subtypes, $markup_escape_chars;

$markup_list_types    = array("*" => "ul", "+" => "ul", "#" => "ol");
$markup_list_subtypes = array("1" => "decimal", "a" => "lower-alpha", "A" => "upper-alpha", "i" => "lower-roman", "I" => "upper-roman", "~" => "none");
$markup_escape_chars = array('\\*'=>"*", '\\_'=>'_', '\\='=>'=', '\\-'=>'-', "\\&quot;"=>"&quot;", "\\]" => "]","\\[" => "[",'\\]'=>']', "\\`"=>"`");
function make_image($prefix, $image, $alt, $alignment = ""){
    $output .= '<img ';
    if($alignment){
        $alignment = strtolower($alignment);
        if($alignment == "r"){
            $output .= 'style="float: right" ';
        }elseif($alignment == "l"){
            $output .= 'style="float: left" ';
        }elseif($alignment == 'c'){
            $output .= 'style="display: block; margin-left: auto; margin-right: auto" ';
        }
    }
    return $output."src=\"$prefix$image\" alt=\"$alt\" />";
}
function long_url_cut($escaped_url, $max_len){
    $url = html_entity_decode($escaped_url);
    return '<a href="'.$escaped_url.'">'.
        (($len = strlen($url)) > $max_len ?
            htmlspecialchars(substr($url,0,ceil($max_len/2)).'...'.substr($url,$len-(floor($max_len/2)-3))) :
            $escaped_url)
        .'</a>';
}
function & markup($str, $flags = array()){
    $er = error_reporting(E_ALL & ~E_NOTICE);
    global $markup_list_types, $markup_escape_chars;
    static $regexes, $styles, $searches, $replaces, $extensions;
    $previous_list_stack = array();
    $footnotes = array();
    $heading_count = array(0,0,0,0,0,0);
    $heading_master_count = 0;
    $heading_current_count = 0;
    $footnote_number = 1;
    $debug = 0;

    $extensions = array("code"=>"markup_code");

    #set missing flags:
    if(!isset($flags['only_get_title']))    {$flags['only_get_title']    = false;}
    if(!isset($flags['page_number']))       {$flags['page_number']       = 0;} #zero means get all pages
    if(!isset($flags['get_page_info']))     {$flags['get_page_info']     = false;}
    #Location offset is the directory offset from your root. This allows you to not have to be at the root of your
    # web space for you to use absolute links. By default it is set to "", which means the root.
    if(!isset($flags['location_offset']))   {$flags['location_offset']   = '';}
    #Main location is the permanent location of this document.
    #It's used for lead-ins.
    if(!isset($flags['main_location']))     {$flags['main_location']     = '';}
    if(!isset($flags['footnotes_heading'])) {$flags['footnotes_heading'] = "Footnotes";} #Maybe someone would want "References"
    if(!isset($flags['leadin_cont_text']))  {$flags['leadin_cont_text']  = "Read more";}
    if(!isset($flags["output_directly"]))   {$flags['output_directly']   = false;}

    if(!isset($flags['autonumber_headings'])){$flags['autonumber_headings'] = false;}
    if(!isset($flags['enable_raw_html']))   {$flags['enable_raw_html']   = true;}
    if(!isset($flags['enable_abs_links']))  {$flags['enable_abs_links']  = false;}
    if(!isset($flags['enable_TOC']))        {$flags['enable_TOC']        = true;}
    if(!isset($flags['enable_headings']))   {$flags['enable_headings']   = true;}
    if(!isset($flags['enable_images']))     {$flags['enable_images']     = true;}
    if(!isset($flags['enable_lists']))      {$flags['enable_lists']      = true;}
    if(!isset($flags['enable_tables']))     {$flags['enable_tables']     = true;}
    if(!isset($flags['enable_styles']))     {$flags['enable_styles']     = true;}
    if(!isset($flags['enable_mono']))       {$flags['enable_mono']       = true;}
    if(!isset($flags['enable_pre']))        {$flags['enable_pre']        = true;}
    if(!isset($flags['enable_hr']))         {$flags['enable_hr']         = true;}
    if(!isset($flags['enable_smileys']))    {$flags['enable_smileys']    = true;}
    if(!isset($flags['enable_comments']))   {$flags['enable_comments']   = true;}
    if(!isset($flags['enable_leadin']))     {$flags['enable_leadin']     = true;}
    if(!isset($flags['enable_footnotes']))  {$flags['enable_footnotes']  = true;}
    if(!isset($flags['enable_extensions'])) {$flags['enable_extensions'] = true;}
    if(!isset($flags['enable_settings']))   {$flags['enable_settings']   = true;}

    if(!isset($flags['show_leadin']))       {$flags['show_leadin']       = false;}

    if(!isset($flags['max_url_length']))    {$flags['max_url_length']    = 80;}

#    if(!isset($flags['link_back_headers'])) {$flags['link_back_headers'] = true;}

    if($flags["enable_abs_links"]){
        $url_prefix = "http://$_SERVER[HTTP_HOST]$flags[location_offset]";
        $flags["main_location"] = $url_prefix.$flags["main_location"];
    }else{
        $url_prefix = $flags['location_offset'];
    }

    #----------------------------------------------------------------------------------
    # Inline Styles
    #----------------------------------------------------------------------------------
    if(!$regexes){
        $t = '\/\\\\.:&;!?%0-9A-Za-z=+_^\\[\\]\\\'\\"\\$'; #characters to match
        $e = '&*=`_\s-.:;,?!)(<>\\[\\]'; #characters to exclude
        $regexes = array();
        $searches = array();
        $regexes[] = "/(?<=^|[>$e])(?<!\\\\)`([$t<-].*?[$t>-]?)`/"; #kbd
        $styles[] = '<kbd>$1</kbd>';
        $regexes[] = "/(?<=^|[>$e])(?<!\\\\)\*{2}([$t<-].*?[$t>-]?)\*{2}(?=[<$e]|$)/"; #strong
        $styles[] = '<strong>$1</strong>';
        $regexes[] = "/(?<=^|\*{2}|[>$e])(?<!\\\\)\*([$t<-].*?[$t>-]?)\*(?=\*{2}|[<$e]|$)/"; #emphasis
        $styles[] = '<em>$1</em>';
        $regexes[] = "/(?<=^|[>$e])(?<!\\\\)_([$t<*-].*?[$t>*-]?)_(?=[<$e]|$)/"; #underline
        $styles[] = '<em style="font-style: normal; text-decoration: underline">$1</em>';

    #    $regexes[] = "/(?<=^|[>$e])(?<!\\\\)(-1,2})([$t<*].*?[$t>*])\\1(?=[<$e]|$)/"; #strikeout
        $regexes[] = "/(?<=^|[>$e])(?<!\\\\)-{2}([$t<*].*?[$t>*])-{2}(?=[<$e]|$)/"; #strikeout
        $styles[] = '<strike>$1</strike>';

#        if($flags["enable_mono"]){
#            $regexes[] = "/(?<=^|[>$e])(?<!\\\\)=([$t<*-].*?[$t>*-])=(?=[<$e]|$)/"; #monospace
#            $styles[] = '<tt>$1</tt>';
#        }
        if($flags["enable_styles"]){
    #        $regexes[] = '/(?<=\\W|)(?<!\\\\)\$([A-Za-z<].*?[>A-Za-z0-9;])\$([A-Za-z<].*?[>A-Za-z])\$(?=\\W|)/'; #any style
            $regexes[] = "/(?<=\\W|)(?<!\\\\)\\$([A-Za-z<].*?[>A-Za-z0-9;])(?<!\\\\)\\$([<$t].*?[$t>])(?<!\\\\)\\$(?=\\W|)/"; #any style
            $styles[] = '<span style="$1; display: inline; position: static">$2</span>';
        }
        if($flags["enable_images"]){
            $regexes[] = "/img([lLrRcC]?)[:=](?<!\\\\)&quot;(.+?)(?<!\\\\)&quot;(?:[:=]?(?<!\\\\)&quot;(.+?)(?<!\\\\)&quot;)?/e"; #images in the format img:"url":"alttext"
            $styles[] = "make_image('$url_prefix', '$2', '$3', '$1')";
        }

#        $regexes[] = "/(?:(?<=^|[\s({]))(\w+:\/\/[\w.,\/?+~&=_:;#$%{}-]+[^\s.?!:,()<>[\]'\"&]+?)/e"; #urls
#        $styles[] = '<a href="$1">$1</a>';

        $regexes[] = "/(?<=^|[\s({])(\w+:\/\/[\w.,\/?+~&=_:;#$%{}()'-|]+[^\s.?!:,()<>[\]'\"&]+?)/e"; #urls
        $styles[] = 'long_url_cut(\'$1\','.$flags['max_url_length'].');';

#        #email addresses
#        $regexes[] = "/(([a-zA-Z0-9_\-\.]+)@((\[[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.)|(([a-zA-Z0-9\-]+\.)+))([a-zA-Z]{2,4}|[0-9]{1,3})(\]?))/e";
#        $styles[] = '"<a href=\"mailto:" . cryptemail("$1") . "\">" . cryptemail("$1") . "</a>"';
    }

    #----------------------------------------------------------------------------------
    # End Inline Styles
    #----------------------------------------------------------------------------------
    #----------------------------------------------------------------------------------
    # Simple string replacements
    #----------------------------------------------------------------------------------
    if(!$searches){
        $searches = $replaces = array();
        if($flags["enable_smileys"]){
            $searches[] = ':)';
            $replaces[] = '<img class="smiley" src="/images/smiley_side.gif" alt="Smiley" />';
            $searches[] = ':-D';
            $replaces[] = '<img class="smiley" src="/images/smiley_happy.gif" alt="Smiley (big smile)" />';
            $searches[] = ':D';
            $replaces[] = '<img class="smiley" src="/images/smiley_happy.gif" alt="Smiley (big smile)" />';
            $searches[] = ' :P';
            $replaces[] = '<img class="smiley" src="/images/smiley_tongue.gif" alt="Smiley sticking out its tongue" />';
            $searches[] = ':(';
            $replaces[] = '<img class="smiley" src="/images/smiley_frown.gif" alt="Smiley frowning" />';
            $searches[] = '&gt;:|';
            $replaces[] = '<img class="smiley" src="/images/smiley_mad.gif" alt="Smiley mad" />';
            $searches[] = '&gt;:(';
            $replaces[] = '<img class="smiley" src="/images/smiley_mad.gif" alt="Smiley mad" />';
            $searches[] = ':|';
            $replaces[] = '<img class="smiley" src="/images/smiley_indifferent.gif" alt="Smiley indifferent" />';
            $searches[] = ' ;)';
            $replaces[] = ' <img class="smiley" src="/images/smiley_wink_animated.gif" alt="Smiley winking" />';
            $searches[] = ' :/';
            $replaces[] = ' <img class="smiley" src="/images/smiley_ohwell.gif" alt="Smiley \'oh well\'" />';
            $searches[] = ' :\\';
            $replaces[] = ' <img class="smiley" src="/images/smiley_ohwell.gif" alt="Smiley \'oh well\'" />';
        }
    }
    #----------------------------------------------------------------------------------
    # End Simple string replacements
    #----------------------------------------------------------------------------------

    #split the $str into separate lines if it's not already
    #find out if it's faster to replace all crlf's with \n's first, then explode
    #rather than split according to a regex
    if(!is_array($str)){
        $str_arr = preg_split('/\n|(?:\r\n?)/', $str);
    }else{
        $str_arr = &$str;
        for($n = 0; $n<count($str_arr);$n++)
            $str_arr[$n] = chop($str_arr[$n]); #make sure there are no line returns
    }

    $line_count = count($str_arr);
    if(!$line_count) return ''; #noop
    $n = 0;

    #----------------------------------------------------------------------------------
    # Preprocessing of directives
    #----------------------------------------------------------------------------------
    if($flags["get_page_info"]){ #if you just want page information
     #improve this to a generic "get structure" flag that will return title, all page info, and all headings in each page
        $pages = array();
        for($n = 0; $n<$line_count; $n++)
            if(substr($str_arr[$n], 0, 5) == "?page")
                $pages[] = trim(substr($str_arr[$n], 5));
        return $pages;
    }
    if(substr($str_arr[0], 0, 6) == "?title"){ #the title has to be the first line in the entire document
        $title = htmlspecialchars(trim(substr($str_arr[0], 6)));
        $str_arr[0] = "<h3 class=\"st-markup\">$title</h3>\n\n";
        $n++;
    }
    if($flags["only_get_title"]) return $title;
    if($flags["page_number"] > 0){
        #get the title for the first page
        $page_count = 1;
        if($flags["page_number"] == 1){
            if(substr($str_arr[$n], 0, 5) == "?page"){ #check the first line to see if it has a page
                $page_title = trim(substr($str_arr[$n], 5));
                $n++;
            }
        }else{
            $n++; #skip past the first line of the file, because whether it's a ?page directive or not, it's in the first page.
            while($n < $line_count and $page_count < $flags["page_number"]){
                if(substr($str_arr[$n], 0, 5) == "?page") $page_count++; #found a new page.
                $n++;
            }
            $page_title = trim(substr($str_arr[$n], 5));
        }
        $m = $n;
        #I have to go to the end to find out where to stop
        while($m < $line_count and substr($str_arr[$m], 0, 5) != "?page") $m++;
        #slice the array from $n (the first line of the content after the page declaration)
        # to $line_count - $m spaces before the end. $line_count - $m will be the remaining spaces in the array
        $str_arr = array_slice($str_arr, $n, -($line_count-$m));
        $line_count = $m-$n;
        $n = 0;
    }
    #----------------------------------------------------------------------------------
    # End Preprocessing of directives
    #----------------------------------------------------------------------------------
    $list_level = 0;
    for(; $n<$line_count; $n++){
        $line = $str_arr[$n];
        $previous_state = $current_state;
        $current_state = $prepend = $append = '';

        if($line{0} == ";" and $flags["enable_raw_html"]){
            $line = substr($line, 1)."\n";
            $current_state = 'raw_html';
        }else{
            if(substr($line, 0,2) != '?!' or !$flags['enable_extensions']){ #if it's not an extension
                $line = &htmlspecialchars($line);
                #first, simple string replacements, preferred over regexes if possible (they're faster)
                $line = &str_replace($searches, $replaces, $line);
                $line = &preg_replace($regexes, $styles, $line);
            }
        }

        #First, processing to be done regardless of the other contents of the line.
        # This ensures that even if the line is a blockquote, for instance, links will still
        # be linked.

/*        $quote_position = 6; # no sense in starting within the first seven characters
        #from the beginning of the line: &quot;A&quot;=&quot;url&quot; is the minimal link.
        #Count the characters and you'll see why I want to start at position seven
        while(($quote_position = strpos($line, "&quot;=&quot;", $quote_position+1)) !=== false){
            if($line{$quote_position-1} == "\\"){
                continue;
            }else{
            }
*/

        if(($quote_position = strpos($line, '&quot;')) !== false){ #if there's a quote on the line
            $m = 0;
            #there's got to be a better way to do this....
            # Find the "=", split on the equals, and look for a quote forward and back from there.
            # Try that later, see if the algorithm is simpler
            while($quote_position !== false){
                if($quote_position == 0 or ($quote_position > 0 and $line{$quote_position-1} != '\\')){
                    #handle escaped quotes
                    $quotes[$m++] = $quote_position;
                }
                $quote_position = strpos($line, '&quot;', $quote_position+6);
            }
            $num_quotes = $m;
            $begin = 0;
            $line_copy = '';
            for($m=1; $m<$num_quotes; $m++){
                if(substr($line, $quotes[$m]+6, 7) == '=&quot;' or substr($line, $quotes[$m]+6, 7) == ':&quot;'){
                    #if the character after the quote was an equals sign and another quote
                    #there'll always be a quote before it, so as long as there's another quote after it...
                    if($m+2 != $num_quotes){ #there have to be two quotes left!
                        #make a link
                        $url      = substr($line, $quotes[$m+1]+6, ($quotes[$m+2] - $quotes[$m+1])-6);
                        if($url_prefix and $url and $url{0} != '/' and !strstr($url, "://"))
                            $url = $url_prefix . $url;
                        $linktext = substr($line, $quotes[$m-1]+6, ($quotes[$m]   - $quotes[$m-1])-6);
                        $quotestring = "<a href=\"$url\">$linktext</a>";
                        $line_copy .= substr($line, $begin, $quotes[$m-1] - $begin) . $quotestring;
                        $begin = $quotes[$m+2] + 6;
                    }
                }
            }
            $line = $line_copy.substr($line, $begin);
            unset($quotes);
        }

        #look for footnotes
        if($flags['enable_footnotes']){
            $footnote_positions = array();
            $close_footnote_position = 0;
            $line_length = strlen($line);
            while(
                $close_footnote_position + 1 < $line_length
                and (($open_footnote_position = strpos($line, "[#", ($close_footnote_position + 1))) !== false)){
                if($line{$open_footnote_position-1} == "\\"){
                    #it's escaped
                    $close_footnote_position = $open_footnote_position+2;
                    continue;
                }else{
                    $temp_pos = $open_footnote_position+2;
                    while(($close_footnote_position = strpos($line, "]", $temp_pos)) !== false){
                        if($line{$close_footnote_position-1} == "\\"){$temp_pos = $close_footnote_position+1; continue;} #escaped

                        $footnote_length = $close_footnote_position - ($open_footnote_position+2);
                        $full_footnote = substr($line, $open_footnote_position+2, $footnote_length);
                        list($footnote, $footnote_text) = explode(" ", $full_footnote, 2);
                        if(!$footnote_text) break; #if it's not a valid footnote

                        #autonumber footnotes
                        $footnote = !$footnote ? $footnote_number++ : normalize_ids($footnote, "fn_");
                        $footnotes[$footnote] = $footnote_text;
                        $replacement_text = "<sup><a id=\"fnp$footnote\" href=\"#fn$footnote\">[$footnote]</a></sup>";
                        $replacement_length = strlen($replacement_text);
                        $line = substr_replace($line, $replacement_text, $open_footnote_position, $footnote_length + 3); # the 3 is plus the opening bracket and # and the closing bracket
                        $close_footnote_position = $open_footnote_position + $replacement_length - 1;
                        $line_length = $line_length - ($footnote_length - $replacement_length);
                        break; #we want to stop once we've found the close footnote position that isn't escaped
                    }
                }
            }
        }
        #----------------------------------------------------------------------------------
        # Determine our state
        #----------------------------------------------------------------------------------
        if($current_state == 'raw_html'){}
        elseif(preg_match('/^---(?:(\d{1,2})([lLrR])?)?$/', $line, $matches)){
            $current_state = MARKUP_STATE_HR;
        }elseif(preg_match('/^((?:(?:[*+]~?|#[1aAiI]?)\s*)+)\s?(.*)$/', $line, $matches)){
            $current_state = MARKUP_STATE_LIST;
            preg_match_all('/[*+]~?|#[1aAiI]?/', $matches[1], $list_begins);
        }elseif(substr($line,0,4) == '&gt;'){
            $current_state = MARKUP_STATE_BLOCKQUOTE;
            $prepend = $previous_state == MARKUP_STATE_BLOCKQUOTE ? "<br />\n" : '<blockquote class="st-markup"><p>';
        }elseif(substr($line,0,1) == '|' and substr($line,-1) == '|'){
            $current_state = MARKUP_STATE_TABLE;
            if($previous_state != MARKUP_STATE_TABLE){$prepend = '<table border="1" class="st-markup">'."\n";}
        }elseif($line{0} == '!'){
            $current_state = MARKUP_STATE_HEADING;
        }elseif($line{0} == "="){
            $current_state = MARKUP_STATE_PRE;
            if($previous_state != MARKUP_STATE_PRE){$prepend = '<pre class="st-markup">';}
        }elseif($line{0} == '?'){
            $current_state = MARKUP_STATE_DIRECTIVE;
        }elseif(!$line){
            $current_state = MARKUP_STATE_NOTHING;
        }else{
            $current_state = MARKUP_STATE_PARA;
            $prepend = $previous_state == MARKUP_STATE_PARA ? "<br />\n" : '<p class="st-markup">';
        }

        if($debug){if(!$previous_state){echo "<table border=\"1\"><tr><th>Line Number</th><th>Current State</th><th>Previous State</th></tr>\n";}
        echo "<tr><td>".($n + 1)."</td><td>$current_state</td><td>$previous_state</td></tr>\n";}

        #these are all states that can potentially continue until the next line.
        #Make sure you correctly close any of the preceding states
        if(    $previous_state == MARKUP_STATE_LIST  and $current_state != MARKUP_STATE_LIST) {$prepend = end_lists(count($list_begins[0]), $list_begins[0]).$prepend; $previous_list_stack = array();}
        elseif($previous_state == MARKUP_STATE_TABLE and $current_state != MARKUP_STATE_TABLE){$prepend = "</table>\n\n$prepend";}
        elseif($previous_state == MARKUP_STATE_PRE   and $current_state != MARKUP_STATE_PRE)  {$prepend = "</pre>\n\n$prepend";}
        elseif($previous_state == MARKUP_STATE_PARA  and $current_state != MARKUP_STATE_PARA) {$prepend = "</p>\n\n$prepend";}
        elseif($previous_state == MARKUP_STATE_BLOCKQUOTE and $current_state != MARKUP_STATE_BLOCKQUOTE){$prepend = "</p></blockquote>\n\n$prepend";}
        #----------------------------------------------------------------------------------
        # Finish determining our state
        #----------------------------------------------------------------------------------

        if($current_state == MARKUP_STATE_LIST and $flags['enable_lists']){ #lists
            $line = list_item($previous_list_stack, $list_begins[0], $matches[2]);
        }elseif($current_state == MARKUP_STATE_TABLE and $flags["enable_tables"]){ #tables
            $cells = explode("|", substr($line, 1,-1));
            $line = "\t<tr>";
            for($m = 0; $m < count($cells); $m++){
                $colspan = 1;
                $cell = $cells[$m];
                while(isset($cells[$m+1]) and $cells[$m+1] == ''){ #if the next cell is empty
                    $colspan++;
                    $m++;
                }
                $line .= "<td".($colspan ? " colspan=\"$colspan\"" : '').">$cell</td>";
            }
            $line .= "</tr>\n";
        }elseif($current_state == MARKUP_STATE_HEADING and $flags["enable_headings"]){ #heading
            if(!$headings[$heading_current_count]){
                $headings[$heading_current_count] = extract_heading_info($line, $flags["autonumber_headings"], $heading_count, $heading_current_count);
            }
            $line = "<h{$headings[$heading_current_count][0]} id=\"{$headings[$heading_current_count][2]}\" class=\"st-markup\">{$headings[$heading_current_count][3]} {$headings[$heading_current_count][1]}</h{$headings[$heading_current_count][0]}>\n\n";
            $heading_current_count++;
        }elseif($current_state == MARKUP_STATE_HR and $flags["enable_hr"]){
            $num = $matches[1]; #left over from the test above
            if($num){ #if there are one or two numeric characters characters
                if(strlen($num) == 1) $num .= "0";
                $line = "<hr class=\"st-markup\" style=\"width: $num%";
                if($matches[2])
                    $line .= strtoupper($matches[2]) == 'L' ? '; float: left' : '; float:right';
                $line .= "\" />\n\n";
            }else
                $line = "<hr class=\"st-markup\" />\n\n";
        }elseif($current_state