/includes/render.php

<?php

/*
 * Enano - an open-source CMS capable of wiki functions, Drupal-like sidebar blocks, and everything in between
 * Copyright (C) 2006-2009 Dan Fuhry
 * render.php - handles fetching pages and parsing them into HTML
 *
 * This program is Free Software; you can redistribute and/or modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied
 * warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for details.
 */
 
class RenderMan {
	
	public static function strToPageID($string)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		$k = array_keys($paths->nslist);
		$proj_alt = 'Project:';
		if ( substr($string, 0, (strlen($proj_alt))) == $proj_alt )
		{
			$ns = 'Project';
			$pg = substr($string, strlen($proj_alt), strlen($string));
			return Array($pg, $ns);
		}
		for($i=0;$i<sizeof($paths->nslist);$i++)
		{
			$ln = strlen($paths->nslist[$k[$i]]);
			if(substr($string, 0, $ln) == $paths->nslist[$k[$i]])
			{
				$ns = $k[$i];
				$pg = substr($string, strlen($paths->nslist[$ns]), strlen($string));
			}
		}
		return Array($pg, $ns);
	}
	
	public static function getPage($page_id, $namespace, $wiki = 1, $smilies = true, $filter_links = true, $redir = true, $render = true)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		
		$page = new PageProcessor($page_id, $namespace);
		$text = $page->fetch_text();
		
		if ( !$render )
		{
			$code = $plugins->setHook('render_getpage_norender');
			foreach ( $code as $cmd )
			{
				eval($cmd);
			}
			return $text;
		}
		
		$text = self::render($text, $wiki, $smilies, $filter_links);
		return $text;
	}
	
	public static function getTemplate($id, $parms)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		
		// Verify target exists -- if not, double-bracket it to convert it to a redlink
		if ( !isPage($paths->get_pathskey($id, 'Template')) ) 
		{
			return '[['.$paths->nslist['Template'].$id.']]';
		}
		
		// fetch from DB or template cache
		if(isset($paths->template_cache[$id]))
		{
			$text = $paths->template_cache[$id];
		}
		else
		{
			$page = new PageProcessor($id, 'Template');
			$text = $page->fetch_text();
			$paths->template_cache[$id] = $text;
		}
		
		// noinclude is not shown within the included template, only on the template's page when you load it
		// nodisplay is not shown on the template's page, only in the included template
		$text = preg_replace('/<noinclude>(.*?)<\/noinclude>/is', '', $text);
		$text = preg_replace('/<nodisplay>(.*?)<\/nodisplay>/is', '\\1', $text);
		
		preg_match_all('#\(_([0-9]+)_\)#', $text, $matchlist);
		
		foreach($matchlist[1] as $m)
		{
			if(isset($parms[((int)$m)+1])) 
			{
				$p = $parms[((int)$m)+1];
			}
			else
			{
				$p = '<b>Notice:</b> RenderMan::getTemplate(): Parameter '.$m.' is not set';
			}
			$text = str_replace('(_'.$m.'_)', $p, $text);
			$text = str_replace('{{' . ( $m + 1 ) . '}}', $p, $text);
		}
		$text = RenderMan::include_templates($text);
		return $text;
	}
	
	public static function fetch_template_text($id, $params)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		$fetch_ns = 'Template';
		if ( !isPage($paths->get_pathskey($id, 'Template')) ) 
		{
			// Transclusion of another page
			// 1.1.5: Now You, Too, Can Be A Template, Even If You're Just A Plain Old Article! (TM)
			$nssep = substr($paths->nslist['Special'], -1);
			$nslist = $paths->nslist;
			foreach ( $nslist as &$ns )
			{
				if ( $ns == '' )
					$ns = $nssep;
			}
			$prefixlist = array_flip($nslist);
			foreach ( $nslist as &$ns )
			{
				$ns = preg_quote($ns);
			}
			$nslist = implode('|', $nslist);
			if ( preg_match("/^($nslist)(.*?)$/", $id, $match) )
			{
				// in practice this should always be true but just to be safe...
				if ( isset($prefixlist[$match[1]]) )
				{
					$new_id = $paths->nslist[ $prefixlist[$match[1]] ] . sanitize_page_id($match[2]);
					if ( !isPage($new_id) )
					{
						return "[[$new_id]]";
					}
					$fetch_ns = $prefixlist[$match[1]];
					$id = sanitize_page_id($match[2]);
				}
			}
			else
			{
				return '[['.$paths->nslist['Template'].$id.']]';
			}
		}
		
		$template->context_push();
		
		$template->assign_vars($params);
		
		if(isset($paths->template_cache[$id]))
		{
			$text = $paths->template_cache[$id];
		}
		else
		{
			$text = RenderMan::getPage($id, $fetch_ns, 0, false, false, false, false);
			$paths->template_cache[$id] = $text;
		}
		
		$template->context_pop();
		
		if ( is_string($text) )
		{
			$text = preg_replace('/<noinclude>(.*?)<\/noinclude>/is', '', $text);
			$text = preg_replace('/<nodisplay>(.*?)<\/nodisplay>/is', '\\1', $text);
		}
		
		return $text;
	}
	
	/**
 	* Renders a glob of text. Note that this is PHP-safe, so if returned text (or rather, "?>" . $returned) has PHP it can be eval'ed.
 	* @param string Text to render
 	* @param int Render parameters - see constants.php
 	* @return string Rendered text
 	*/
	
	public static function render($text, $flags = RENDER_WIKI_DEFAULT, $smilies = true)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		
		if ( !$smilies )
			$flags |= RENDER_NOSMILIES;
		
		if ( !($flags & RENDER_NOSMILIES) )
		{
			$text = RenderMan::smilieyize($text);
		}
		if ( $flags & RENDER_WIKI_DEFAULT )
		{
			$text = RenderMan::next_gen_wiki_format($text, $flags);
		}
		else if ( $flags & RENDER_WIKI_TEMPLATE )
		{
			$text = $template->tplWikiFormat($text);
		}           
		return $text;
	}
	
	private static function next_gen_wiki_format($text, $flags = 0)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		global $lang;
		
		profiler_log("RenderMan: starting wikitext render");
		require_once( ENANO_ROOT . '/includes/wikiformat.php' );
		require_once( ENANO_ROOT . '/includes/wikiengine/TagSanitizer.php' );
		require_once( ENANO_ROOT . '/includes/wikiengine/Tables.php' );
		
		// this is still needed by parser plugins
		$random_id = md5( time() . mt_rand() );
		
		// Strip out <nowiki> sections
		self::nowiki_strip($text, $nowiki_stripped);
		
		// Run early parsing plugins
		$code = $plugins->setHook('render_wikiformat_veryearly');
		foreach ( $code as $cmd )
		{
			eval($cmd);
		}
		
		// Strip out embedded PHP
		self::php_strip($text, $php_stripped);
		
		// Convert newlines for the parser
		$text = str_replace("\r\n", "\n", $text);
		
		// Perform render through the engine
		$carpenter = new Carpenter();
		$carpenter->flags = $flags;
		$carpenter->hook(array(__CLASS__, 'hook_pre'), PO_AFTER, 'lang');
		$carpenter->hook(array(__CLASS__, 'hook_posttemplates'), PO_AFTER, 'templates');
		if ( $flags & RENDER_WIKI_TEMPLATE )
		{
			// FIXME: Where is noinclude/nodisplay being processed in the pipeline? (Seems to be processed, but not here)
		}
		
		//
		// Set rules for the rendering process
		//
		
		if ( $flags & RENDER_BLOCK && !($flags & RENDER_INLINE) )
		{
			// block only
			$carpenter->disable_all_rules();
			foreach ( array('blockquote', 'tables', 'heading', 'hr', 'multilist', 'bold', 'italic', 'underline', 'paragraph', 'blockquotepost') as $rule )
			{
				$carpenter->enable_rule($rule);
			}
			
			$code = $plugins->setHook('render_block_only');
			foreach ( $code as $cmd )
			{
				eval($cmd);
			}
		}
		else if ( $flags & RENDER_INLINE && !($flags & RENDER_BLOCK) )
		{
			// inline only
			$carpenter->disable_all_rules();
			foreach ( array('heading', 'bold', 'italic', 'underline', 'externalwithtext', 'externalnotext', 'image', 'internallink') as $rule )
			{
				$carpenter->enable_rule($rule);
			}
			
			$code = $plugins->setHook('render_inline_only');
			foreach ( $code as $cmd )
			{
				eval($cmd);
			}
		}
		else
		{
			// full render
			$code = $plugins->setHook('render_full');
			foreach ( $code as $cmd )
			{
				eval($cmd);
			}
		}
		$text = $carpenter->render($text);
		
		// For plugin compat
		$result =& $text;
		
		// Post processing hook
		$code = $plugins->setHook('render_wikiformat_post');
		foreach ( $code as $cmd )
		{
			eval($cmd);
		}
		
		// Add PHP and nowiki back in
		self::nowiki_unstrip($text, $nowiki_stripped);
		self::php_unstrip($text, $php_stripped);
		
		profiler_log("RenderMan: finished wikitext render");
		
		return $text;
	}
	
	public static function hook_pre($text)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		
		$code = $plugins->setHook('render_wikiformat_pre');
		foreach ( $code as $cmd )
		{
			eval($cmd);
		}
		
		return $text;
	}
	
	public static function hook_posttemplates($text)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		
		$code = $plugins->setHook('render_wikiformat_posttemplates');
		foreach ( $code as $cmd )
		{
			eval($cmd);
		}
		
		return $text;
	}
	
	/**
 	* Strip out <nowiki> tags (to bypass parsing on them)
 	* @access private
 	*/
	
	private static function nowiki_strip(&$text, &$stripdata)
	{
		self::tag_strip('nowiki', $text, $stripdata);
	}
	
	/**
 	* Restore stripped <nowiki> tags.
 	* @access private
 	*/
	
	public static function nowiki_unstrip(&$text, &$stripdata)
	{
		self::tag_unstrip('nowiki', $text, $stripdata);
	}
	
	/**
 	* Strip out an arbitrary HTML tag.
 	* @access private
 	*/
	
	public static function tag_strip($tag, &$text, &$stripdata)
	{
		$random_id = md5( time() . mt_rand() );
		
		preg_match_all("#<$tag>(.*?)</$tag>#is", $text, $blocks);
		
		foreach ( $blocks[0] as $i => $match )
		{
			$text = str_replace($match, "{{$tag}:{$random_id}:{$i}}", $text);
		}
		
		$stripdata = array(
				'random_id' => $random_id,
				'blocks' => $blocks[1]
			);
	}
	
	/**
 	* Strip out an arbitrary HTML tag, pushing on to the existing list of stripped data.
 	* @access private
 	*/
	
	public static function tag_strip_push($tag, &$text, &$stripdata)
	{
		if ( !is_array($stripdata) )
		{
			$stripdata = array(
					'random_id' => md5( time() . mt_rand() ),
					'blocks' => array()
				);
		}
		$random_id =& $stripdata['random_id'];
		
		preg_match_all("#<$tag>(.*?)</$tag>#is", $text, $blocks);
		
		foreach ( $blocks[0] as $i => $match )
		{
			$j = count($stripdata['blocks']);
			$stripdata['blocks'][] = $blocks[1][$i];
			$text = str_replace($match, "{{$tag}:{$random_id}:{$j}}", $text);
		}
	}
	
	/**
 	* Restore stripped <nowiki> tags.
 	* @access private
 	*/
	
	public static function tag_unstrip($tag, &$text, &$stripdata, $keep = false)
	{
		$random_id = $stripdata['random_id'];
		
		foreach ( $stripdata['blocks'] as $i => $block )
		{
			$block = $keep ? "<$tag>$block</$tag>" : $block;
			$text = str_replace("{{$tag}:{$random_id}:{$i}}", $block, $text);
		}
		
		$stripdata = array();
	}
	
	/**
 	* Strip out PHP code (to prevent it from being sent through the parser). Private because it does not do what you think it does. (The method you are looking for is strip_php.)
 	* @access private
 	*/
	
	private static function php_strip(&$text, &$stripdata)
	{
		$random_id = md5( time() . mt_rand() );
		
		preg_match_all('#<\?(?:php)?[\s=].+?\?>#is', $text, $blocks);
		
		foreach ( $blocks[0] as $i => $match )
		{
			$text = str_replace($match, "{PHP:$random_id:$i}", $text);
		}
		
		$stripdata = array(
				'random_id' => $random_id,
				'blocks' => $blocks[0]
			);
	}
	
	/**
 	* Restore stripped PHP code
 	* @access private
 	*/
	
	private static function php_unstrip(&$text, &$stripdata)
	{
		$random_id = $stripdata['random_id'];
		
		foreach ( $stripdata['blocks'] as $i => $block )
		{
			$text = str_replace("{PHP:$random_id:$i}", $block, $text);
		}
		
		$stripdata = array();
	}
	
	/**
 	* Deprecated.
 	*/
	
	public static function wikiFormat($message, $filter_links = true, $do_params = false, $plaintext = false)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		
		return RenderMan::next_gen_wiki_format($message, $plaintext, $filter_links, $do_params);
	}
	
	public static function destroy_javascript($message, $_php = false)
	{
		$message = preg_replace('#<(script|object|applet|embed|iframe|frame|form|input|select)(.*?)>#is', '&lt;\\1\\2&gt;', $message);
		$message = preg_replace('#</(script|object|applet|embed|iframe|frame|form|input|select)(.*?)>#is', '&lt;/\\1\\2&gt;', $message);
		$message = preg_replace('#(javascript|script|activex|chrome|about|applet):#is', '\\1&#058;', $message);
		if ( $_php )
		{
			// Left in only for compatibility
			$message = preg_replace('#&lt;(.*?)>#is', '<\\1>', $message);
			$message = preg_replace('#<(.*?)&gt;#is', '<\\1>', $message);
			$message = preg_replace('#<(\?|\?php|%)(.*?)(\?|%)>#is', '&lt;\\1\\2\\3&gt;', $message);
			// strip <a href="foo" onclick="bar();">-type attacks
			$message = preg_replace('#<([a-zA-Z:\-]+) (.*?)on([A-Za-z]*)=(.*?)>#is', '&lt;\\1\\2on\\3=\\4&gt;', $message);
		}
		return $message;
	}
	
	public static function strip_php($message)
	{
		return RenderMan::destroy_javascript($message, true);
	}
	
	public static function sanitize_html($text)
	{
		$text = htmlspecialchars($text);
		$allowed_tags = Array('b', 'i', 'u', 'pre', 'code', 'tt', 'br', 'p', 'nowiki', '!--([\w\W]+)--');
		foreach($allowed_tags as $t)
		{
			$text = preg_replace('#&lt;'.$t.'&gt;(.*?)&lt;/'.$t.'&gt;#is', '<'.$t.'>\\1</'.$t.'>', $text);
			$text = preg_replace('#&lt;'.$t.' /&gt;#is', '<'.$t.' />', $text);
			$text = preg_replace('#&lt;'.$t.'&gt;#is', '<'.$t.'>', $text);
		}
		return $text;
	}
	
	/**
 	* Reverse-renders a blob of text (converts it from XHTML back to wikitext) by using parser hints and educated guesses.
 	* @param string XHTML
 	* @return string Wikitext
 	*/
	
	public static function reverse_render($text)
	{
		// convert \r\n to \n
		$text = str_replace("\r\n", "\n", $text);
		
		// Separate certain block level elements onto their own lines. This tidies up the tag
		// soup that TinyMCE sometimes produces.
		$block_elements = array('h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'p', 'div', 'table', 'ul', 'pre');
		$block_elements = implode('|', $block_elements);
		$regex = "#(</(?:$block_elements)>)\n?<($block_elements)(>| .+?>)#i";
		$text = preg_replace($regex, "$1\n\n<$2$3", $text);
		
		$text = self::reverse_process_parser_hints($text);
		$text = self::reverse_process_headings($text);
		$text = self::reverse_process_lists($text);
		$text = self::reverse_process_tables($text);
		
		// Lastly, strip out paragraph tags.
		$text = preg_replace('|^ *<p>(.+?)</p> *$|m', "\\1", $text);
		
		return $text;
	}
	
	public static function reverse_process_parser_hints($text)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		
		if ( !preg_match_all('|<!--#([a-z0-9_]+)(?: (.+?))?-->([\w\W]*?)<!--#/\\1-->|s', $text, $matches) )
			return $text;
		
		foreach ( $matches[0] as $i => $match )
		{
			$tag =& $matches[1][$i];
			$attribs =& $matches[2][$i];
			$inner =& $matches[3][$i];
			
			$attribs = self::reverse_process_hint_attribs($attribs);
			switch($tag)
			{
				case 'smiley':
				case 'internallink':
				case 'imagelink':
					if ( isset($attribs['code']) )
					{
						$text = str_replace($match, $attribs['code'], $text);
					}
					else if ( isset($attribs['src']) )
					{
						$text = str_replace($match, $attribs['src'], $text);
					}
					break;
			}
		}
		
		return $text;
	}
	
	public static function reverse_process_hint_attribs($attribs)
	{
		$return = array();
		if ( !preg_match_all('/([a-z0-9_-]+)="([^"]+?)"/', $attribs, $matches) )
			return array();
		
		foreach ( $matches[0] as $i => $match )
		{
			$name =& $matches[1][$i];
			$value =& $matches[2][$i];
			
			$value = base64_decode($value);
			
			$return[$name] = $value;
		}
		
		return $return;
	}
	
	/**
 	* Escapes a string so that it's safe to use as an attribute in a parser hint.
 	* @param string
 	* @return string
 	*/
	
	public static function escape_parser_hint_attrib($text)
	{
		return base64_encode($text);
	}
	
	public static function reverse_process_headings($text)
	{
		if ( !preg_match_all('|^<h([1-6])(?: id="head:[\w\.\/:;\(\)@\[\]=_-]+")?>(.*?)</h\\1>$|m', $text, $matches) )
			return $text;
		
		foreach ( $matches[0] as $i => $match )
		{
			// generate heading tag
			$heading_size = intval($matches[1][$i]);
			$eq = '';
			for ( $j = 0; $j < $heading_size; $j++ )
				$eq .= '=';
			
			$heading =& $matches[2][$i];
			
			$tag = "$eq $heading $eq";
			$text = str_replace($match, $tag, $text);
		}
		
		return $text;
	}
	
	public static function reverse_process_lists($text)
	{
		if ( !preg_match('!(</?(?:ul|ol|li)>)!', $text) )
			return $text;
		
		$split = preg_split('!(</?(?:ul|ol|li)>)!', $text, -1, PREG_SPLIT_DELIM_CAPTURE);
		
		$stack_height = 0;
		$current_list = '';
		$old_current_list = '';
		$spaces = '';
		$marker = '*';
		$list_id = 0;
		$just_terminated = false;
		foreach ( $split as $tag )
		{
			switch($tag) 
			{
				case '<ul>':
				case '<ol>':
					$stack_height++;
					$just_terminated = false;
					if ( $stack_height > 1 )
						$spaces .= $marker;
					
					$marker = ( $tag == 'ol' ) ? '#' : '*';
					if ( $stack_height > 1 )
						$current_list .= "\n";
					
					break;
				case '</ul>':
				case '</ol>':
					$stack_height--;
					$spaces = substr($spaces, 1);
					
					if ( $stack_height == 0 )
					{
						// rotate
						$text = str_replace_once("{$old_current_list}{$tag}", trim($current_list), $text);
						$current_list = '';
						$old_current_list = '';
					}
					$just_terminated = true;
					break;
				case '<li>':
					if ( $stack_height < 1 )
						break;
					
					$current_list .= "{$spaces}{$marker} ";
					break;
				case '</li>':
					if ( $stack_height < 1 )
						break;
					
					if ( !$just_terminated )
						$current_list .= "\n";
					
					$just_terminated = false;
					break;
				default:
					if ( $stack_height > 0 )
					{
						$current_list .= trim($tag);
					}
					break;
			}
			if ( $stack_height > 0 )
			{
				$old_current_list .= $tag;
			}
		}
		
		return $text;
	}
	
	public static function reverse_process_tables($text)
	{
		return $text;
	}
	
	/**
 	* Parses internal links (wikilinks) in a block of text.
 	* @param string Text to process
 	* @param string Optional. If included will be used as a template instead of using the default syntax.
 	* @param bool If false, does not add wikilink-nonexistent or check for exsistence of pages. Can reduce DB queries; defualts to true.
 	* @param string Page ID. If specified, class="currentpage" will be added to links if they match the given page ID and namespace
 	* @param string Namespace. If specified, class="currentpage" will be added to links if they match the given page ID and namespace
 	* @return string
 	*/
	
	public static function parse_internal_links($text, $tplcode = false, $do_exist_check = true, $match_page_id = false, $match_namespace = false)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
	
		$parser = is_string($tplcode) ? $template->makeParserText($tplcode) : false;
		
		// allow blank urlname?
		$repeater = have_blank_urlname_page() ? '*' : '+';
		
		// stage 1 - links with alternate text
		preg_match_all('/\[\[([^\[\]<>\{\}\|]' . $repeater . ')\|(.+?)\]\]/', $text, $matches);
		foreach ( $matches[0] as $i => $match )
		{
			list($page_id, $namespace) = RenderMan::strToPageID($matches[1][$i]);
			$link = self::generate_internal_link($namespace, $page_id, $matches[2][$i], $match, $parser, $do_exist_check, $match_page_id, $match_namespace);
			$text = str_replace($match, $link, $text);
		}
		
		// stage 2 - links with no alternate text
		preg_match_all('/\[\[([^\[\]<>\{\}\|]' . $repeater . ')\]\]/', $text, $matches);
		foreach ( $matches[0] as $i => $match )
		{
			list($page_id, $namespace) = RenderMan::strToPageID($matches[1][$i]);
			$pid_clean = $paths->nslist[$namespace] . sanitize_page_id($page_id);
			$inner_text = ( isPage($pid_clean) ) ? htmlspecialchars(get_page_title($pid_clean)) : htmlspecialchars($matches[1][$i]);
			
			$link = self::generate_internal_link($namespace, $page_id, $inner_text, $match, $parser, $do_exist_check, $match_page_id, $match_namespace);
			
			$text = str_replace($match, $link, $text);
		}
		
		return $text;
	}
	
	/**
 	* Internal link generation function
 	* @access private
 	* @return string HTML
 	*/
	
	private static function generate_internal_link($namespace, $page_id, $inner_text, $match, $parser = false, $do_exist_check = true, $match_page_id = false, $match_namespace = false)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		
		if ( ($pos = strrpos($page_id, '#')) !== false )
		{
			$hash = substr($page_id, $pos);
			$page_id = substr($page_id, 0, $pos);
		}
		else
		{
			$hash = '';
		}
		
		if ( $namespace == 'Admin' )
		{
			// No linking directly to Admin pages!
			$get = 'module=' . $paths->nslist[$namespace] . sanitize_page_id($page_id);
			$pid_clean = $paths->nslist['Special'] . 'Administration';
			$onclick = ' onclick="ajaxLoginNavTo(\'Special\', \'Administration\', USER_LEVEL_ADMIN, \'' . addslashes($get) . '\'); return false;"';
		}
		else
		{
			$get = false;
			$onclick = '';
			$pid_clean = $paths->nslist[$namespace] . sanitize_page_id($page_id);
		}
		
		$url = makeUrl($pid_clean, $get, true) . $hash;
		$quot = '"';
		$exists = ( ($do_exist_check && isPage($pid_clean)) || !$do_exist_check ) ? '' : ' class="wikilink-nonexistent"';
		
		if ( $match_page_id && $match_namespace && $pid_clean === $paths->get_pathskey($match_page_id, $match_namespace) )
			$exists .= ' class="currentpage"';
		
		if ( $parser )
		{
			$parser->assign_vars(array(
					'HREF' => $url,
					'FLAGS' => $exists,
					'TEXT' => $inner_text
				));
			$link = $parser->run();
		}
		else
		{
			$omatch = self::escape_parser_hint_attrib($match);
			$link = "<!--#internallink src=\"$omatch\" --><a{$onclick} href={$quot}{$url}{$quot}{$exists}>{$inner_text}</a><!--#/internallink-->";
		}
		
		return $link;
	}
	
	/**
 	* Parses a partial template tag in wikitext, and return an array with the parameters.
 	* @param string The portion of the template tag that contains the parameters.
 	* @example
 	* <code>
 	foo = lorem ipsum
 	bar = dolor sit amet
 	* </code>
 	* @return array Example:
 	* [foo] => lorem ipsum
 	* [bar] => dolor sit amet
 	*/
	
	public static function parse_template_vars($input, $newlinemode = true)
	{
		$parms = array();
		$input = trim($input);
		if ( $newlinemode )
		{
			// we're going by newlines.
			// split by parameter, then parse each one individually
			$input = explode("\n", str_replace("\r", '', $input));
			$lastparam = '';
			$i = 0;
			foreach ( $input as $line )
			{
				if ( preg_match('/^ *\|? *([A-z0-9_]+) *= *(.+)$/', $line, $match) )
				{
					// new parameter, named
					$parms[ $match[1] ] = $match[2];
					$lastparam = $match[1];
				}
				else if ( preg_match('/^ *\| *(.+)$/', $line, $match) || $lastparam === '' )
				{
					$parms[ $i ] = $match[1];
					$lastparam = $i;
					$i++;
				}
				else
				{
					$parms[ $lastparam ] .= "\n$line";
				}
			}
		}
		else
		{
			$result = preg_match_all('/
 																(?:^|[ ]*)\|         # start of parameter - string start or series of spaces
 																[ ]*
 																(?:
 																	([A-z0-9_]+)       # variable name
 																	[ ]* = [ ]*        # assignment
 																)?                   # name section is optional - if the parameter name is not given, a numerical index is assigned
 																([^\|]+|.+?\n[ ]*\|) # value
 															/x', trim($input), $matches);
			if ( $result )
			{
				$pi = 0;
				for ( $i = 0; $i < count($matches[0]); $i++ )
				{
					$matches[1][$i] = trim($matches[1][$i]);
					$parmname = !empty($matches[1][$i]) ? $matches[1][$i] : strval(++$pi);
					$parms[ $parmname ] = $matches[2][$i];
				}
			}
		}
		// die('<pre>' . print_r($parms, true) . '</pre>');
		return $parms;
	}
	
	/**
 	* Processes all template tags within a block of wikitext.
 	* Updated in 1.0.2 to also parse template tags in the format of {{Foo |a = b |b = c |c = therefore, a}}
 	* @param string The text to process
 	* @return string Formatted text
 	* @example
 	* <code>
 	$text = '{{Template
 			| parm1 = Foo
 			| parm2 = Bar
 		}}';
 	$text = RenderMan::include_templates($text);
 	* </code>
 	*/
	
	public static function include_templates($text)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		// $template_regex = "/\{\{([^\]]+?)((\n([ ]*?)[A-z0-9]+([ ]*?)=([ ]*?)(.+?))*)\}\}/is";
		// matches:
		//  1 - template name
		//  2 - parameter section
		$template_regex = "/
 												\{\{                     # opening
 													([^\n\t\a\r]+)         # template name
 													((?:(?:[\s]+\|?)[ ]*(?:[A-z0-9_]+)[ ]*=[ ]*?(?:.+))*) # parameters
 												\}\}                     # closing
 											/isxU";
		if ( $count = preg_match_all($template_regex, $text, $matches) )
		{
			// die('<pre>' . print_r($matches, true) . '</pre>');
			for ( $i = 0; $i < $count; $i++ )
			{
				$matches[1][$i] = sanitize_page_id($matches[1][$i]);
				$newlinemode = ( substr($matches[2][$i], 0, 1) == "\n" );
				$parmsection = trim($matches[2][$i]);
				if ( !empty($parmsection) )
				{
					$parms = RenderMan::parse_template_vars($parmsection, $newlinemode);
					if ( !is_array($parms) )
						// Syntax error
						$parms = array();
				}
				else
				{
					$parms = Array();
				}
				if ( $tpl_code = RenderMan::fetch_template_text($matches[1][$i], $parms) )
				{
					// Intelligent paragraphs.
					// If:
					//   * A line is fully wrapped in a <p> tag
					//   * The line contains a variable
					//   * The variable contains newlines
					// Then:
					//   * Drop the <p> tag, replace it fully paragraph-ized by newlines
					
					if ( preg_match_all('/^( *)<p(\s.+)?>(.*?\{([A-z0-9]+)\}.*?)<\/p> *$/m', $tpl_code, $paramatch) )
					{
						$parser = new Carpenter();
						$parser->exclusive_rule('paragraph');
						
						foreach ( $paramatch[0] as $j => $match )
						{
							// $line is trimmed (the <p> is gone)
							$spacing =& $paramatch[1][$i];
							$para_attrs =& $paramatch[2][$j];
							$para_attrs = str_replace(array('$', '\\'), array('\$', '\\\\'), $para_attrs);
							$line =& $paramatch[3][$j];
							$varname =& $paramatch[4][$j];
							if ( isset($parms[$varname]) && strstr($parms[$varname], "\n") )
							{
								$newline = str_replace('{' . $varname . '}', $parms[$varname], $line);
								$paraized = $parser->render($newline);
								$paraized = preg_replace('/^<p>/m', "$spacing<p{$para_attrs}>", $paraized);
								$paraized = $spacing . trim($paraized);
								$tpl_code = str_replace_once($match, $paraized, $tpl_code);
							}
						}
					}
					
					$parser = $template->makeParserText($tpl_code);
					$parser->assign_vars($parms);
					$text = str_replace($matches[0][$i], $parser->run(), $text);
				}
			}
		}
		return $text;
	}
	
	/**
 	* Preprocesses an HTML text string prior to being sent to MySQL.
 	* @param string $text
 	* @param bool $strip_all_php - if true, strips all PHP regardless of user permissions. Else, strips PHP only if user level < USER_LEVEL_ADMIN. Defaults to true.
 	* @param bool $sqlescape - if true, sends text through $db->escape(). Otherwise returns unescaped text. Defaults to true.
 	* @param bool $reduceheadings - if true, finds HTML headings and replaces them with wikitext. Else, does not touch headings. Defaults to true.
 	* @param Session_ACLPageInfo Optional permissions instance to check against, $session is used if not provided
 	*/
	public static function preprocess_text($text, $strip_all_php = true, $sqlescape = true, $reduceheadings = true, $perms = false)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		$random_id = md5( time() . mt_rand() );
		
		$code = $plugins->setHook('render_sanitize_pre');
		foreach ( $code as $cmd )
		{
			eval($cmd);
		}
		
		if ( !is_object($perms) )
		{
			$namespace = $paths->namespace;
			$perms =& $session;
		}
		else
		{
			$namespace = $perms->namespace;
		}
		
		$can_do_php = ( !$strip_all_php && $perms->get_permissions('php_in_pages') );
		$can_do_html = $session->check_acl_scope('html_in_pages', $namespace) && $perms->get_permissions('html_in_pages');
		
		if ( $can_do_html && !$can_do_php )
		{
			$text = preg_replace('#<(\?|\?php|%)(.*?)(\?|%)>#is', '&lt;\\1\\2\\3&gt;', $text);
		}
		else if ( !$can_do_html && !$can_do_php )
		{
			$text = sanitize_html($text, true);
			// If we can't do PHP, we can't do Javascript either.
			$text = RenderMan::destroy_javascript($text);
		}
		
		// Strip out <nowiki> sections and PHP code
		
		$php = preg_match_all('#(<|&lt;)\?php(.*?)\?(>|&gt;)#is', $text, $phpsec);
		
		//die('<pre>'.htmlspecialchars(print_r($phpsec, true))."\n".htmlspecialchars(print_r($text, true)).'</pre>');
		
		for($i=0;$i<sizeof($phpsec[1]);$i++)
		{
			$text = str_replace($phpsec[0][$i], '{PHP:'.$random_id.':'.$i.'}', $text);
		}
		
		$nw = preg_match_all('#<nowiki>(.*?)<\/nowiki>#is', $text, $nowiki);
		
		for($i=0;$i<sizeof($nowiki[1]);$i++)
		{
			$text = str_replace('<nowiki>'.$nowiki[1][$i].'</nowiki>', '{NOWIKI:'.$random_id.':'.$i.'}', $text);
		}
		
		$text = str_replace('~~~~~', enano_date('G:i, j F Y (T)'), $text);
		$text = str_replace('~~~~', "[[User:$session->username|$session->username]] ".enano_date('G:i, j F Y (T)'), $text);
		$text = str_replace('~~~', "[[User:$session->username|$session->username]] ", $text);
		
		$code = $plugins->setHook('render_sanitize_post');
		foreach ( $code as $cmd )
		{
			eval($cmd);
		}
		
		// gently apply some reverse-processing to allow the parser to do magic with TOCs and stuff
		if ( $reduceheadings )
			$text = self::reverse_process_headings($text);
		
		// Reinsert <nowiki> sections
		for($i=0;$i<$nw;$i++)
		{
			$text = str_replace('{NOWIKI:'.$random_id.':'.$i.'}', '<nowiki>'.$nowiki[1][$i].'</nowiki>', $text);
		}
		// Reinsert PHP
		for($i=0;$i<$php;$i++)
		{
			$phsec = ''.$phpsec[1][$i].'?php'.$phpsec[2][$i].'?'.$phpsec[3][$i].'';
			if ( $strip_all_php )
				$phsec = htmlspecialchars($phsec);
			$text = str_replace('{PHP:'.$random_id.':'.$i.'}', $phsec, $text);
		}
		
		$text = ( $sqlescape ) ? $db->escape($text) : $text;
		
		return $text;
	}
	
	public static function smilieyize($text, $complete_urls = false)
	{
		
		$random_id = md5( time() . mt_rand() );
		
		// Smileys array - eventually this will be fetched from the database by
		// RenderMan::initSmileys during initialization, but it will all be hardcoded for beta 2
		
		$smileys = Array(
			'O:-)'    => 'face-angel.png',
			'O:)'     => 'face-angel.png',
			'O=)'     => 'face-angel.png',
			':-)'     => 'face-smile.png',
			':)'      => 'face-smile.png',
			'=)'      => 'face-smile-big.png',
			':-('     => 'face-sad.png',
			':('      => 'face-sad.png',
			';('      => 'face-sad.png',
			':-O'     => 'face-surprise.png',
			';-)'     => 'face-wink.png',
			';)'      => 'face-wink.png',
			'8-)'     => 'face-glasses.png',
			'8)'      => 'face-glasses.png',
			':-D'     => 'face-grin.png',
			':D'      => 'face-grin.png',
			'=D'      => 'face-grin.png',
			':-*'     => 'face-kiss.png',
			':*'      => 'face-kiss.png',
			'=*'      => 'face-kiss.png',
			':\'('    => 'face-crying.png',
			':-|'     => 'face-plain.png',
			':-\\'    => 'face-plain.png',
			':-/'     => 'face-plain.png',
			':joke:'  => 'face-plain.png',
			']:-&gt;' => 'face-devil-grin.png',
			']:->'    => 'face-devil-grin.png',
			':kiss:'  => 'face-kiss.png',
			':-P'     => 'face-tongue-out.png',
			':P'      => 'face-tongue-out.png',
			':-p'     => 'face-tongue-out.png',
			':p'      => 'face-tongue-out.png',
			':-X'     => 'face-sick.png',
			':X'      => 'face-sick.png',
			':sick:'  => 'face-sick.png',
			':-]'     => 'face-oops.png',
			':]'      => 'face-oops.png',
			':oops:'  => 'face-oops.png',
			':-['     => 'face-embarassed.png',
			':['      => 'face-embarassed.png'
			);
		
		// Strip out <nowiki> sections
		//return '<pre>'.htmlspecialchars($text).'</pre>';
		$nw = preg_match_all('#<nowiki>(.*?)<\/nowiki>#is', $text, $nowiki);
		
		for ( $i = 0; $i < count($nowiki[1]); $i++ )
		{
			$text = str_replace('<nowiki>' . $nowiki[1][$i] . '</nowiki>', '{NOWIKI:'.$random_id.':'.$i.'}', $text);
		}
		
		foreach ( $smileys as $smiley => $smiley_path )
		{
			$hex_smiley = hexencode($smiley, '&#x', ';');
			$pfx = ( $complete_urls ) ? get_server_url() : '';
			$text = str_replace(' ' . $smiley,
					' <!--#smiley code="' . self::escape_parser_hint_attrib($smiley) . '"--><nowiki>
 					<!-- The above is a reverse-parser hint -->
 						<img title="' . $hex_smiley . '" alt="' . $hex_smiley . '" src="' . $pfx . scriptPath . '/images/smilies/' . $smiley_path . '"
							style="border: 0;" />
 					</nowiki><!--#/smiley-->', $text);
		}
		//*/
		
		// Reinsert <nowiki> sections
		for ( $i = 0; $i < $nw; $i++ )
		{
			$text = str_replace('{NOWIKI:'.$random_id.':'.$i.'}', '<nowiki>'.$nowiki[1][$i].'</nowiki>', $text);
		}
		
		return $text;
	}
	
	/**
 	* Generates a summary of the differences between two texts, and formats it as XHTML.
 	* @param $str1 string the first block of text
 	* @param $str2 string the second block of text
 	* @return string
 	*/
	public static function diff($str1, $str2)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		require_once(ENANO_ROOT.'/includes/diff.php');
		$str1 = explode("\n", $str1);
		$str2 = explode("\n", $str2);
		$diff = new Diff($str1, $str2);
		$renderer = new TableDiffFormatter();
		return '<table class="diff">'.$renderer->format($diff).'</table>';
	}
	
	/**
 	* Changes wikitext image tags to HTML.
 	* @param string The wikitext to process
 	* @param array Will be overwritten with the list of HTML tags (the system uses tokens for TextWiki compatibility)
 	* @return string
 	*/
	
	public static function process_image_tags($text, &$taglist)
	{
		global $db, $session, $paths, $template, $plugins; // Common objects
		
		$s_delim = "\xFF";
		$f_delim = "\xFF";
		$taglist = array();
		
		// Wicked huh?
		$ns_file = str_replace('/', '\\/', preg_quote($paths->nslist['File']));
		$regex = '/
 					\[\[                                                                  # starting delimiter 
 					:' . $ns_file . '([\w\s0-9_\(\)!@%\^\+\|\.-]+?\.(?:png|gif|jpg|jpeg)) # image filename
 					(?:(?:\|(?:.+?))*)                                                    # parameters
 					\]\]                                                                  # ending delimiter
 					/ix';
		
		preg_match_all($regex, $text, $matches);
		
		foreach ( $matches[0] as $i => $match )
		{
			
			$full_tag   =& $matches[0][$i];
			$filename   =& $matches[1][$i];
			
			// apply recursion (hack? @todo could this be done with (?R) in PCRE?)
			// this allows other elements such as internal/external links to be embedded in image captions
			$tag_pos = strpos($text, $full_tag);
			$tag_end_pos = $tag_pos + strlen($full_tag);
			while ( get_char_count($full_tag, ']') < get_char_count($full_tag, '[') && $tag_end_pos < strlen($text) )
			{
				$full_tag .= substr($text, $tag_end_pos, 1);
				$tag_end_pos++;
			}
			if ( $tag_end_pos > strlen($text) )
			{
				// discard tag, not closed fully
				continue;
			}
			
			// init the various image parameters
			$width = null;
			$height = null;
			$scale_type = null;
			$clear = null;
			$caption = null;
			$display_type = 'inline';
			
			// trim tag and parse particles
			$tag_trim = rtrim(ltrim($full_tag, '['), ']');
			// trim off the filename from the start of the tag
			$filepart_len = 1 + strlen($paths->nslist['File']) + strlen($filename) + 1;
			$tag_trim = substr($tag_trim, $filepart_len);
			// explode and we should have parameters
			$tag_parts = explode('|', $tag_trim);
			
			// for each of the parameters, see if it matches a known option. If so, apply it;
			// otherwise, see if a plugin reserved that parameter and if not treat it as the caption
			foreach ( $tag_parts as $param )
			{
				switch($param)
				{
					case 'left':
					case 'right':
						$clear = $param;
						$display_type = 'framed';
						break;
					case 'thumb':
						$scale_type = 'thumb';
						break;
					case 'raw':
						$display_type = 'raw';
						break;
					default:
						// height specification
						if ( preg_match('/^([0-9]+)x([0-9]+)$/', $param, $dims) )
						{
							$width = intval($dims[1]);
							$height = intval($dims[2]);
							break;
						}
						// not the height, so see if a plugin took this over
						// this hook requires plugins to return true if they modified anything
						$code = $plugins->setHook('img_tag_parse_params', true);
						foreach ( $code as $cmd )
						{
							if ( eval($cmd) )
								break 2;
						}
						// we would have broken out by now if a plugin properly handled this,
						// so just set the caption now.
						$caption = $param;
						break;
				}
			}
			
			if ( !isPage( $paths->nslist['File'] . $filename ) )
			{
				$text = str_replace($full_tag, '[[' . $paths->nslist['File'] . $filename . ']]', $text);
				continue;
			}
			
			if ( $scale_type == 'thumb' )
			{
				$r_width  = 225;
				$r_height = 225;
				
				$url = makeUrlNS('Special', 'DownloadFile/' . $filename, 'preview&width=' . $r_width . '&height=' . $r_height, true);
			}
			else if ( !empty($width) && !empty($height) )
			{
				$r_width = $width;
				$r_height = $height;
				
				$url = makeUrlNS('Special', 'DownloadFile/' . $filename, 'preview&width=' . $r_width . '&height=' . $r_height, true);
			}
			else
			{
				$url = makeUrlNS('Special', 'DownloadFile/' . $filename);
			}
			
			$img_tag = '<img src="' . $url . '" ';
			
			// if ( isset($r_width) && isset($r_height) && $scale_type != '|thumb' )
			// {
			//   $img_tag .= 'width="' . $r_width . '" height="' . $r_height . '" ';
			// }
			
			$img_tag .= 'style="border-width: 0px;" ';
			
			$code = $plugins->setHook('img_tag_parse_img');
			foreach ( $code as $cmd )
			{
				eval($cmd);
			}
			
			$img_tag .= '/>';
			
			$s_full_tag = self::escape_parser_hint_attrib($full_tag);
			$complete_tag = '<!--#imagelink src="' . $s_full_tag . '" -->';
			
			if ( $display_type == 'framed' )
			{
				$complete_tag .= '<div class="thumbnail" ';
				$clear_text = '';
				if ( !empty($clear) )
				{
					$side = ( $clear == 'left' ) ? 'left' : 'right';
					$opposite = ( $clear == 'left' ) ? 'right' : 'left';
					$clear_text .= "float: $side; margin-$opposite: 20px; width: {$r_width}px;";
					$complete_tag .= 'style="' . $clear_text . '" ';
				}
				$complete_tag .= '>';
				
				$complete_tag .= '<a href="' . makeUrlNS('File', $filename) . '" style="display: block;">';
				$complete_tag .= $img_tag;
				$complete_tag .= '</a>';
				
				$mag_button = '<a href="' . makeUrlNS('File', $filename) . '" style="display: block; float: right; clear: right; margin: 0 0 10px 10px;"><img alt="[ + ]" src="' . scriptPath . '/images/thumbnail.png" style="border-width: 0px;" /></a>';
			
				if ( !empty($caption) )
				{
					$complete_tag .= $mag_button . $caption;
				}
				
				$complete_tag .= '</div>';
			}
			else if ( $display_type == 'raw' )
			{
				$complete_tag .= "$img_tag";
				$taglist[$i] = $complete_tag;
				
				$repl = "{$s_delim}e_img_{$i}{$f_delim}";
				$text = str_replace($full_tag, $repl, $text);
				continue;
			}
			else
			{
				$complete_tag .= '<a href="' . makeUrlNS('File', $filename) . '" style="display: block;"';
				$code = $plugins->setHook('img_tag_parse_link');
				foreach ( $code as $cmd )
				{
					eval($cmd);
				}
				$complete_tag .= '>';
				$complete_tag .= $img_tag;
				$complete_tag .= '</a>';
			}
			
			$complete_tag .= "<!--#/imagelink-->";
			$taglist[$i] = $complete_tag;
			
			/*
			$pos = strpos($text, $full_tag);
			
			while(true)
			{
				$check1 = substr($text, $pos, 3);
				$check2 = substr($text, $pos, 1);
				if ( $check1 == '<p>' || $pos == 0 || $check2 == "\n" )
				{
					// die('found at pos '.$pos);
					break;
				}
				$pos--;
			}
			*/
			
			/*
			$repl = "{$s_delim}e_img_{$i}{$f_delim}";
			$text = substr($text, 0, $pos) . $repl . substr($text, $pos);
			
			$text = str_replace($full_tag, '', $text);
			*/
			$text = str_replace_once($full_tag, $complete_tag, $text);
			
			unset($full_tag, $filename, $scale_type, $width, $height, $clear, $caption, $r_width, $r_height);
			
		}
		
		// if ( count($matches[0]) > 0 )
		//   die('<pre>' . htmlspecialchars($text) . '</pre>');
		
		return $text;
	}
	
	/**
 	* Finalizes processing of image tags.
 	* @param string The preprocessed text
 	* @param array The list of image tags created by RenderMan::process_image_tags()
 	*/
 	
	public static function process_imgtags_stage2($text, $taglist)
	{
		$s_delim = "\xFF";
		$f_delim = "\xFF";
		foreach ( $taglist as $i => $tag )
		{
			$repl = "{$s_delim}e_img_{$i}{$f_delim}";
			$text = str_replace($repl, $tag, $text);
		}               
		return $text;
	}
	
}
 
?>