* Removed the troublesome regular expression in isValidEmailAddr which returns
[lhc/web/wiklou.git] / includes / ParserXML.php
index a4c30b1..3998479 100644 (file)
-<?php
-require_once ( "Parser.php" ) ;
+<?php 
+/**
+ * 
+ * @package MediaWiki
+ * @subpackage Experimental
+ */
+
+/** */
+require_once ('Parser.php');
 
 /**
  * This should one day become the XML->(X)HTML parser
  * Based on work by Jan Hidders and Magnus Manske
+ * To use, set
+ *    $wgUseXMLparser = true ;
+ *    $wgEnableParserCache = false ;
+ *    $wgWiki2xml to the path and executable of the command line version (cli)
+ * in LocalSettings.php
  * @package MediaWiki
  * @subpackage Experimental
  */
 
 /**
  * the base class for an element
+ * @package MediaWiki
+ * @subpackage Experimental
  */
 class element {
-    var $name = '';
-    var $attrs = array();
-    var $children = array();
-
-    function sub_makeXHTML ( &$parser , $tag = "" , $attr = "" )
-       {
-       $ret = "" ;
-       if ( $tag != "" )
-               {
-               $ret .= "<" . $tag ;
-               if ( $attr != "" ) $ret .= " " . $attr ;
-               $ret .= ">" ;
-               }
-      foreach ($this->children as $child) {
-            if ( is_string($child) ) {
-                $ret .= $child ;
-            } else {
-                $ret .= $child->makeXHTML ( $parser );
-            }
-           }
-       if ( $tag != "" )
-               $ret .= "</" . $tag . ">\n" ;
-      return $ret ;
-       }
-       
-    function createInternalLink ( &$parser , $target , $display_title , $options )
-       {
-       $tp = explode ( ":" , $target ) ; # tp = target parts
-       $title = "" ;     # The plain title
-       $language = "" ;  # The language/meta/etc. part
-       $namespace = "" ; # The namespace, if any
-       $subtarget = "" ; # The '#' thingy
-       if ( count ( $tp ) == 1 ) $title = $target ; # Plain and simple case
-       else
-               {
-               # To be implemented
-               }
-       
-       if ( $language != "" ) # External link within the WikiMedia project
-               {
-               return "{language link}" ;
-               }
-       else if ( $namespace != "" ) # Link to another namespace, check for image/media stuff
-               {
-               return "{namespace link}" ;
-               }
-       else
-               {
-               return "{internal link}" ;
-               }
-       }
-    
-    function makeInternalLink ( &$parser )
-       {
-       $target = "" ;
-       $option = array () ;
-      foreach ($this->children as $child) {
-            if ( is_string($child) ) {
-               # This shouldn't be the case!
-            } else {
-                if ( $child->name == "LINKTARGET" )
-                       $target = trim ( $child->makeXHTML ( $parser ) ) ;
-                   else
-                       $option[] = trim ( $child->makeXHTML ( $parser ) ) ;
-            }
-           }
-           
-      if ( count ( $option ) == 0 ) $option[] = $target ; # Create dummy display title
-      $display_title = array_pop ( $option ) ;
-      return $this->createInternalLink ( $parser , $target , $display_title , $option ) ;
-       }
-       
-    function makeXHTML ( &$parser )
-       {
-       $ret = "" ;
-       $n = $this->name ; # Shortcut
-       if ( $n == "ARTICLE" )
-               $ret .= $this->sub_makeXHTML ( $parser ) ;
-       else if ( $n == "HEADING" )
-               $ret .= $this->sub_makeXHTML ( $parser , "h" . $this->attrs["LEVEL"] ) ;
-       else if ( $n == "PARAGRAPH" )
-               $ret .= $this->sub_makeXHTML ( $parser , "p" ) ;
-       else if ( $n == "BOLD" )
-               $ret .= $this->sub_makeXHTML ( $parser , "strong" ) ;
-       else if ( $n == "ITALICS" )
-               $ret .= $this->sub_makeXHTML ( $parser , "em" ) ;
-
-       else if ( $n == "LINK" )
-               $ret .= $this->makeInternalLink ( $parser ) ;
-       else if ( $n == "LINKTARGET" )
-               $ret .= $this->sub_makeXHTML ( $parser ) ;
-       else if ( $n == "LINKOPTION" )
-               $ret .= $this->sub_makeXHTML ( $parser ) ;
-               
-       else if ( $n == "EXTENSION" ) # This is currently a dummy!!!
-               {
-               $ext = $this->attrs["NAME"] ;
-               
-               $ret .= "&lt;" . $ext . "&gt;" ;
-               $ret .= $this->sub_makeXHTML ( $parser ) ;
-               $ret .= "&lt;/" . $ext . "&gt; " ;
-               }
-
-       else if ( $n == "TABLE" )
-               {
-               $ret .= $this->sub_makeXHTML ( $parser , "table" ) ;
-               }
-       else if ( $n == "TABLEROW" )
-               {
-               $ret .= $this->sub_makeXHTML ( $parser , "tr" ) ;
-               }
-       else if ( $n == "TABLECELL" )
-               {
-               $ret .= $this->sub_makeXHTML ( $parser , "td" ) ;
-               }
-
-
-       else if ( $n == "LISTITEM" )
-               $ret .= $this->sub_makeXHTML ( $parser , "li" ) ;
-       else if ( $n == "LIST" )
-               {
-               $type = "ol" ; # Default
-               if ( $this->attrs["TYPE"] == "bullet" ) $type = "ul" ;
-               $ret .= $this->sub_makeXHTML ( $parser , $type ) ;
-               }
-               
-       else
-               {
-               $ret .= "&lt;" . $n . "&gt;" ;
-               $ret .= $this->sub_makeXHTML ( $parser ) ;
-               $ret .= "&lt;/" . $n . "&gt; " ;
-               }
-       $ret = "\n{$ret}\n" ;
-       $ret = str_replace ( "\n\n" , "\n" , $ret ) ;
-       return $ret ;
-       }
-
-    function myPrint() {
-        $ret = "<ul>\n";
-        $ret .= "<li> <b> Name: </b> $this->name </li>\n";
-        // print attributes
-        $ret .= '<li> <b> Attributes: </b>';
-        foreach ($this->attrs as $name => $value) {
-            $ret .= "$name => $value; " ;
-        }
-        $ret .= " </li>\n";
-        // print children
-        foreach ($this->children as $child) {
-            if ( is_string($child) ) {
-                $ret .= "<li> $child </li>\n";
-            } else {
-                $ret .= $child->myPrint();
-            }
-        }
-        $ret .= "</ul>\n";
-        return $ret;
-    }
+       var $name = '';
+       var $attrs = array ();
+       var $children = array ();
+
+       /**
+       * This finds the ATTRS element and returns the ATTR sub-children as a single string
+       */
+       function getSourceAttrs() {
+               $ret = '';
+               foreach ($this->children as $child) {
+                       if (!is_string($child) AND $child->name == 'ATTRS') {
+                               $ret = $child->makeXHTML($parser);
+                       }
+               }
+               return $ret;
+       }
+
+       /**
+        * This collects the ATTR thingies for getSourceAttrs()
+        */
+       function getTheseAttrs() {
+               $ret = array ();
+               foreach ($this->children as $child) {
+                       if (!is_string($child) AND $child->name == 'ATTR') {
+                               $ret[] = $child->attrs["NAME"]."='".$child->children[0]."'";
+                       }
+               }
+               return implode(' ', $ret);
+       }
+
+       function fixLinkTails(& $parser, $key) {
+               $k2 = $key +1;
+               if (!isset ($this->children[$k2]))
+                       return;
+               if (!is_string($this->children[$k2]))
+                       return;
+               if (is_string($this->children[$key]))
+                       return;
+               if ($this->children[$key]->name != "LINK")
+                       return;
+
+               $n = $this->children[$k2];
+               $s = '';
+               while ($n != '' AND (($n[0] >= 'a' AND $n[0] <= 'z') OR $n[0] == 'ä' OR $n[0] == 'ö' OR $n[0] == 'ü' OR $n[0] == 'ß')) {
+                       $s .= $n[0];
+                       $n = substr($n, 1);
+               }
+               $this->children[$k2] = $n;
+
+               if (count($this->children[$key]->children) > 1) {
+                       $kl = array_keys($this->children[$key]->children);
+                       $kl = array_pop($kl);
+                       $this->children[$key]->children[$kl]->children[] = $s;
+               } else {
+                       $e = new element;
+                       $e->name = "LINKOPTION";
+                       $t = $this->children[$key]->sub_makeXHTML($parser);
+                       $e->children[] = trim($t).$s;
+                       $this->children[$key]->children[] = $e;
+               }
+       }
+
+       /**
+       * This function generates the XHTML for the entire subtree
+       */
+       function sub_makeXHTML(& $parser, $tag = '', $attr = '') {
+               $ret = '';
+
+               $attr2 = $this->getSourceAttrs();
+               if ($attr != '' AND $attr2 != '')
+                       $attr .= ' ';
+               $attr .= $attr2;
+
+               if ($tag != '') {
+                       $ret .= '<'.$tag;
+                       if ($attr != '')
+                               $ret .= ' '.$attr;
+                       $ret .= '>';
+               }
+
+               # THIS SHOULD BE DONE IN THE WIKI2XML-PARSER INSTEAD
+               #       foreach ( array_keys ( $this->children ) AS $x )
+               #          $this->fixLinkTails ( $parser , $x ) ;
+
+               foreach ($this->children as $key => $child) {
+                       if (is_string($child)) {
+                               $ret .= $child;
+                       } elseif ($child->name != 'ATTRS') {
+                               $ret .= $child->makeXHTML($parser);
+                       }
+               }
+               if ($tag != '')
+                       $ret .= '</'.$tag.">\n";
+               return $ret;
+       }
+
+       /**
+       * Link functions
+       */
+       function createInternalLink(& $parser, $target, $display_title, $options) {
+               global $wgUser;
+               $skin = $wgUser->getSkin();
+               $tp = explode(':', $target); # tp = target parts
+               $title = ''; # The plain title
+               $language = ''; # The language/meta/etc. part
+               $namespace = ''; # The namespace, if any
+               $subtarget = ''; # The '#' thingy
+
+               $nt = Title :: newFromText($target);
+               $fl = strtoupper($this->attrs['FORCEDLINK']) == 'YES';
+
+               if ($fl || count($tp) == 1) {
+                       # Plain and simple case
+                       $title = $target;
+               } else {
+                       # There's stuff missing here...
+                       if ($nt->getNamespace() == NS_IMAGE) {
+                               $options[] = $display_title;
+                               return $skin->makeImageLinkObj($nt, implode('|', $options));
+                       } else {
+                               # Default
+                               $title = $target;
+                       }
+               }
+
+               if ($language != '') {
+                       # External link within the WikiMedia project
+                       return "{language link}";
+               } else {
+                       if ($namespace != '') {
+                               # Link to another namespace, check for image/media stuff
+                               return "{namespace link}";
+                       } else {
+                               return $skin->makeLink($target, $display_title);
+                       }
+               }
+       }
+
+       /** @todo document */
+       function makeInternalLink(& $parser) {
+               $target = '';
+               $option = array ();
+               foreach ($this->children as $child) {
+                       if (is_string($child)) {
+                               # This shouldn't be the case!
+                       } else {
+                               if ($child->name == 'LINKTARGET') {
+                                       $target = trim($child->makeXHTML($parser));
+                               } else {
+                                       $option[] = trim($child->makeXHTML($parser));
+                               }
+                       }
+               }
+
+               if (count($option) == 0)
+                       $option[] = $target; # Create dummy display title
+               $display_title = array_pop($option);
+               return $this->createInternalLink($parser, $target, $display_title, $option);
+       }
+
+       /** @todo document */
+       function getTemplateXHTML($title, $parts, & $parser) {
+               global $wgLang, $wgUser;
+               $skin = $wgUser->getSkin();
+               $ot = $title; # Original title
+               if (count(explode(':', $title)) == 1)
+                       $title = $wgLang->getNsText(NS_TEMPLATE).":".$title;
+               $nt = Title :: newFromText($title);
+               $id = $nt->getArticleID();
+               if ($id == 0) {
+                       # No/non-existing page
+                       return $skin->makeBrokenLink($title, $ot);
+               }
+
+               $a = 0;
+               $tv = array (); # Template variables
+               foreach ($parts AS $part) {
+                       $a ++;
+                       $x = explode('=', $part, 2);
+                       if (count($x) == 1)
+                               $key = "{$a}";
+                       else
+                               $key = $x[0];
+                       $value = array_pop($x);
+                       $tv[$key] = $value;
+               }
+               $art = new Article($nt);
+               $text = $art->getContent(false);
+               $parser->plain_parse($text, true, $tv);
+
+               return $text;
+       }
+
+       /**
+        * This function actually converts wikiXML into XHTML tags
+        * @todo use switch() !
+        */
+       function makeXHTML(& $parser) {
+               $ret = '';
+               $n = $this->name; # Shortcut
+
+               if ($n == 'EXTENSION') {
+                       # Fix allowed HTML
+                       $old_n = $n;
+                       $ext = strtoupper($this->attrs['NAME']);
+                       
+                       switch($ext) {
+                               case 'B':
+                               case 'STRONG':
+                                       $n = 'BOLD';
+                                       break;
+                               case 'I':
+                               case 'EM':
+                                       $n = 'ITALICS';
+                                       break;
+                               case 'U':
+                                       $n = 'UNDERLINED'; # Hey, virtual wiki tag! ;-)
+                                       break;
+                               case 'S':
+                                       $n = 'STRIKE';
+                                       break;
+                               case 'P':
+                                       $n = 'PARAGRAPH';
+                                       break;
+                               case 'TABLE':
+                                       $n = 'TABLE';
+                                       break;
+                               case 'TR':
+                                       $n = 'TABLEROW';
+                                       break;
+                               case 'TD':
+                                       $n = 'TABLECELL';
+                                       break;
+                               case 'TH':
+                                       $n = 'TABLEHEAD';
+                                       break;
+                               case 'CAPTION':
+                                       $n = 'CAPTION';
+                                       break;
+                               case 'NOWIKI':
+                                       $n = 'NOWIKI';
+                                       break;
+                               }
+                       if ($n != $old_n) {
+                               unset ($this->attrs['NAME']); # Cleanup
+                       } elseif ($parser->nowiki > 0) {
+                               # No 'real' wiki tags allowed in nowiki section
+                               $n = '';
+                       }
+               } // $n = 'EXTENSION'
+
+               switch($n) {
+                       case 'ARTICLE':
+                               $ret .= $this->sub_makeXHTML($parser);
+                               break;
+                       case 'HEADING':
+                               $ret .= $this->sub_makeXHTML($parser, 'h'.$this->attrs['LEVEL']);
+                               break;
+                       case 'PARAGRAPH':
+                               $ret .= $this->sub_makeXHTML($parser, 'p');
+                               break;
+                       case 'BOLD':
+                               $ret .= $this->sub_makeXHTML($parser, 'strong');
+                               break;
+                       case 'ITALICS':
+                               $ret .= $this->sub_makeXHTML($parser, 'em');
+                               break;
+
+                       # These don't exist as wiki markup
+                       case 'UNDERLINED':
+                               $ret .= $this->sub_makeXHTML($parser, 'u');
+                               break;
+                       case 'STRIKE':
+                               $ret .= $this->sub_makeXHTML($parser, 'strike');
+                               break;
+
+                       # HTML comment
+                       case 'COMMENT':
+                               # Comments are parsed out
+                               $ret .= '';
+                               break;
+                               
+
+                       # Links
+                       case 'LINK':
+                               $ret .= $this->makeInternalLink($parser);
+                               break;
+                       case 'LINKTARGET':
+                       case 'LINKOPTION':
+                               $ret .= $this->sub_makeXHTML($parser);
+                               break;
+               
+                       case 'TEMPLATE':
+                               $parts = $this->sub_makeXHTML($parser);
+                               $parts = explode('|', $parts);
+                               $title = array_shift($parts);
+                               $ret .= $this->getTemplateXHTML($title, $parts, & $parser);
+                               break;
+
+                       case 'TEMPLATEVAR':
+                               $x = $this->sub_makeXHTML($parser);
+                               if (isset ($parser->mCurrentTemplateOptions["{$x}"]))
+                                       $ret .= $parser->mCurrentTemplateOptions["{$x}"];
+                               break;
+
+                       # Internal use, not generated by wiki2xml parser                                
+                       case 'IGNORE':
+                               $ret .= $this->sub_makeXHTML($parser);
+
+                       case 'NOWIKI':
+                               $parser->nowiki++;
+                               $ret .= $this->sub_makeXHTML($parser, '');
+                               $parser->nowiki--;
+
+
+                       # Unknown HTML extension
+                       case 'EXTENSION': # This is currently a dummy!!!
+                               $ext = $this->attrs['NAME'];
+
+                               $ret .= '&lt;'.$ext.'&gt;';
+                               $ret .= $this->sub_makeXHTML($parser);
+                               $ret .= '&lt;/'.$ext.'&gt; ';
+                               break;
+
+
+                       # Table stuff
+
+                       case 'TABLE':
+                               $ret .= $this->sub_makeXHTML($parser, 'table');
+                               break;
+                       case 'TABLEROW':
+                               $ret .= $this->sub_makeXHTML($parser, 'tr');
+                               break;
+                       case 'TABLECELL':
+                               $ret .= $this->sub_makeXHTML($parser, 'td');
+                               break;
+                       case 'TABLEHEAD':
+                               $ret .= $this->sub_makeXHTML($parser, 'th');
+                               break;
+                       case 'CAPTION':
+                               $ret .= $this->sub_makeXHTML($parser, 'caption');
+                               break;
+                       case 'ATTRS': # SPECIAL CASE : returning attributes
+                               return $this->getTheseAttrs();
+
+
+                       # Lists stuff
+                       case 'LISTITEM':
+                               if ($parser->mListType == 'dl')
+                                       $ret .= $this->sub_makeXHTML($parser, 'dd');
+                               else
+                                       $ret .= $this->sub_makeXHTML($parser, 'li');
+                               break;
+                       case 'LIST':
+                                       $type = 'ol'; # Default
+                                       if ($this->attrs['TYPE'] == 'bullet')
+                                               $type = 'ul';
+                                       else
+                                               if ($this->attrs['TYPE'] == 'indent')
+                                                       $type = 'dl';
+                                       $oldtype = $parser->mListType;
+                                       $parser->mListType = $type;
+                                       $ret .= $this->sub_makeXHTML($parser, $type);
+                                       $parser->mListType = $oldtype;
+                               break;
+                       
+                       # Something else entirely
+                       default:
+                               $ret .= '&lt;'.$n.'&gt;';
+                               $ret .= $this->sub_makeXHTML($parser);
+                               $ret .= '&lt;/'.$n.'&gt; ';
+                       } // switch($n)
+
+               $ret = "\n{$ret}\n";
+               $ret = str_replace("\n\n", "\n", $ret);
+               return $ret;
+       }
+
+       /**
+        * A function for additional debugging output
+        */
+       function myPrint() {
+               $ret = "<ul>\n";
+               $ret .= "<li> <b> Name: </b> $this->name </li>\n";
+               // print attributes
+               $ret .= '<li> <b> Attributes: </b>';
+               foreach ($this->attrs as $name => $value) {
+                       $ret .= "$name => $value; ";
+               }
+               $ret .= " </li>\n";
+               // print children
+               foreach ($this->children as $child) {
+                       if (is_string($child)) {
+                               $ret .= "<li> $child </li>\n";
+                       } else {
+                               $ret .= $child->myPrint();
+                       }
+               }
+               $ret .= "</ul>\n";
+               return $ret;
+       }
 }
 
-$ancStack = array();    // the stack with ancestral elements
+$ancStack = array (); // the stack with ancestral elements
 
-// Three global functions needed for parsing, sorry guys
+// START Three global functions needed for parsing, sorry guys
+/** @todo document */
 function wgXMLstartElement($parser, $name, $attrs) {
-    global $ancStack;
+       global $ancStack;
 
-    $newElem = new element;
-    $newElem->name = $name;
-    $newElem->attrs = $attrs;
+       $newElem = new element;
+       $newElem->name = $name;
+       $newElem->attrs = $attrs;
 
-    array_push($ancStack, $newElem);
+       array_push($ancStack, $newElem);
 }
 
+/** @todo document */
 function wgXMLendElement($parser, $name) {
-    global $ancStack, $rootElem;
-    // pop element off stack
-    $elem = array_pop ($ancStack);
-    if (count ($ancStack) == 0)
-        $rootElem = $elem;
-    else
-        // add it to its parent
-        array_push ($ancStack[count($ancStack)-1]->children, $elem);
+       global $ancStack, $rootElem;
+       // pop element off stack
+       $elem = array_pop($ancStack);
+       if (count($ancStack) == 0)
+               $rootElem = $elem;
+       else
+               // add it to its parent
+               array_push($ancStack[count($ancStack) - 1]->children, $elem);
 }
 
+/** @todo document */
 function wgXMLcharacterData($parser, $data) {
-    global $ancStack;
-    $data = trim ($data); // Don't add blank lines, they're no use...
-    // add to parent if parent exists
-    if ( $ancStack && $data != "" ) {
-        array_push ($ancStack[count($ancStack)-1]->children, $data);
-    }
+       global $ancStack;
+       $data = trim($data); // Don't add blank lines, they're no use...
+       // add to parent if parent exists
+       if ($ancStack && $data != "") {
+               array_push($ancStack[count($ancStack) - 1]->children, $data);
+       }
 }
-
+// END Three global functions needed for parsing, sorry guys
 
 /**
  * Here's the class that generates a nice tree
+ * @package MediaWiki
+ * @subpackage Experimental
  */
 class xml2php {
 
-    function &scanFile( $filename ) {
-        global $ancStack, $rootElem;
-        $ancStack = array();
-
-        $xml_parser = xml_parser_create();
-        xml_set_element_handler ($xml_parser, 'wgXMLstartElement', 'wgXMLendElement');
-        xml_set_character_data_handler ($xml_parser, 'wgXMLcharacterData');
-        if (!($fp = fopen($filename, 'r'))) {
-            die('could not open XML input');
-        }
-        while ($data = fread($fp, 4096)) {
-            if (!xml_parse($xml_parser, $data, feof($fp))) {
-                die(sprintf("XML error: %s at line %d",
-                    xml_error_string(xml_get_error_code($xml_parser)),
-                    xml_get_current_line_number($xml_parser)));
-            }
-        }
-        xml_parser_free($xml_parser);
-
-        // return the remaining root element we copied in the beginning
-        return $rootElem;
-    }
-
-    function scanString ( $input ) {
-        global $ancStack, $rootElem;
-        $ancStack = array();
-
-        $xml_parser = xml_parser_create();
-        xml_set_element_handler ($xml_parser, 'wgXMLstartElement', 'wgXMLendElement');
-        xml_set_character_data_handler ($xml_parser, 'wgXMLcharacterData');
-
-        if (!xml_parse ($xml_parser, $input, true)) {
-            die (sprintf ("XML error: %s at line %d",
-                xml_error_string(xml_get_error_code($xml_parser)),
-                xml_get_current_line_number($xml_parser)));
-        }
-        xml_parser_free ($xml_parser);
-
-        // return the remaining root element we copied in the beginning
-        return $rootElem;
-    }
+       /** @todo document */
+       function & scanFile($filename) {
+               global $ancStack, $rootElem;
+               $ancStack = array ();
+
+               $xml_parser = xml_parser_create();
+               xml_set_element_handler($xml_parser, 'wgXMLstartElement', 'wgXMLendElement');
+               xml_set_character_data_handler($xml_parser, 'wgXMLcharacterData');
+               if (!($fp = fopen($filename, 'r'))) {
+                       die('could not open XML input');
+               }
+               while ($data = fread($fp, 4096)) {
+                       if (!xml_parse($xml_parser, $data, feof($fp))) {
+                               die(sprintf("XML error: %s at line %d", xml_error_string(xml_get_error_code($xml_parser)), xml_get_current_line_number($xml_parser)));
+                       }
+               }
+               xml_parser_free($xml_parser);
+
+               // return the remaining root element we copied in the beginning
+               return $rootElem;
+       }
 
-}
+       /** @todo document */
+       function scanString($input) {
+               global $ancStack, $rootElem;
+               $ancStack = array ();
 
-/* Example code:
+               $xml_parser = xml_parser_create();
+               xml_set_element_handler($xml_parser, 'wgXMLstartElement', 'wgXMLendElement');
+               xml_set_character_data_handler($xml_parser, 'wgXMLcharacterData');
 
-    $w = new xml2php;
-    $filename = 'sample.xml';
-    $result = $w->scanFile( $filename );
-    print $result->myPrint();
-*/
+               if (!xml_parse($xml_parser, $input, true)) {
+                       die(sprintf("XML error: %s at line %d", xml_error_string(xml_get_error_code($xml_parser)), xml_get_current_line_number($xml_parser)));
+               }
+               xml_parser_free($xml_parser);
 
-$dummytext = "<article><heading level='2'> R-type </heading><paragraph><link><linktarget>image:a.jpg</linktarget><linkoption>1</linkoption><linkoption>2</linkoption><linkoption>3</linkoption><linkoption>text</linkoption></link></paragraph><paragraph>The <link><linktarget>video game</linktarget><linkoption>computer game</linkoption></link> <bold>R-type</bold> is <extension name='nowiki'>cool &amp; stuff</extension> because:</paragraph><list type='bullet'><listitem>it's nice</listitem><listitem>it's fast</listitem><listitem>it has:<list type='bullet'><listitem>graphics</listitem><listitem>sound</listitem></list></listitem></list><table><tablerow><tablecell>Version 1     </tablecell><tablecell>not bad</tablecell></tablerow><tablerow><tablecell>Version 2     </tablecell><tablecell>much better </tablecell></tablerow></table><paragraph>This is a || token in the middle of text.</paragraph></article>" ;
+               // return the remaining root element we copied in the beginning
+               return $rootElem;
+       }
+
+}
 
-class ParserXML EXTENDS Parser
-       {
+/**
+ * @todo document
+ * @package MediaWiki
+ * @subpackage Experimental
+ */
+class ParserXML extends Parser {
        /**#@+
         * @access private
         */
        # Persistent:
-       var $mTagHooks;
+       var $mTagHooks, $mListType;
 
        # Cleared with clearState():
-       var $mOutput, $mAutonumber, $mDTopen, $mStripState = array();
+       var $mOutput, $mAutonumber, $mDTopen, $mStripState = array ();
        var $mVariables, $mIncludeCount, $mArgStack, $mLastSection, $mInPre;
 
        # Temporary:
-       var $mOptions, $mTitle, $mOutputType,
-           $mTemplates,        // cache of already loaded templates, avoids
-                               // multiple SQL queries for the same string
-           $mTemplatePath;     // stores an unsorted hash of all the templates already loaded
-                               // in this path. Used for loop detection.
+       var $mOptions, $mTitle, $mOutputType, $mTemplates, // cache of already loaded templates, avoids
+       // multiple SQL queries for the same string
+       $mTemplatePath; // stores an unsorted hash of all the templates already loaded
+       // in this path. Used for loop detection.
+
+       var $nowikicount, $mCurrentTemplateOptions;
 
        /**#@-*/
 
@@ -291,9 +547,9 @@ class ParserXML EXTENDS Parser
         * @access public
         */
        function ParserXML() {
-               $this->mTemplates = array();
-               $this->mTemplatePath = array();
-               $this->mTagHooks = array();
+               $this->mTemplates = array ();
+               $this->mTemplatePath = array ();
+               $this->mTagHooks = array ();
                $this->clearState();
        }
 
@@ -308,25 +564,79 @@ class ParserXML EXTENDS Parser
                $this->mLastSection = "";
                $this->mDTopen = false;
                $this->mVariables = false;
-               $this->mIncludeCount = array();
-               $this->mStripState = array();
-               $this->mArgStack = array();
+               $this->mIncludeCount = array ();
+               $this->mStripState = array ();
+               $this->mArgStack = array ();
                $this->mInPre = false;
        }
-       
-       function parse( $text, &$title, $options, $linestart = true, $clearState = true ) {
-               global $dummytext ;
-               $text = $dummytext ;
-               
+
+       /**
+       * Turns the wikitext into XML by calling the external parser
+       *
+       */
+       function html2xml(& $text) {
+               global $wgWiki2xml;
+
+               # generating html2xml command path
+               $a = $wgWiki2xml;
+               $a = explode('/', $a);
+               array_pop($a);
+               $a[] = 'html2xml';
+               $html2xml = implode('/', $a);
+               $a = array ();
+
+               $tmpfname = tempnam('/tmp', 'FOO');
+               $handle = fopen($tmpfname, 'w');
+               fwrite($handle, utf8_encode($text));
+               fclose($handle);
+               exec($html2xml.' < '.$tmpfname, $a);
+               $text = utf8_decode(implode("\n", $a));
+               unlink($tmpfname);
+       }
+
+       /** @todo document */
+       function runXMLparser(& $text) {
+               global $wgWiki2xml;
+
+               $this->html2xml($text);
+
+               $tmpfname = tempnam('/tmp', 'FOO');
+               $handle = fopen($tmpfname, 'w');
+               fwrite($handle, $text);
+               fclose($handle);
+               exec($wgWiki2xml.' < '.$tmpfname, $a);
+               $text = utf8_decode(implode("\n", $a));
+               unlink($tmpfname);
+       }
+
+       /** @todo document */
+       function plain_parse(& $text, $inline = false, $templateOptions = array ()) {
+               $this->runXMLparser($text);
+               $nowikicount = 0;
                $w = new xml2php;
-               $result = $w->scanString( $text );
-               $text .= "<hr>" . $result->makeXHTML ( $this );
-               $text .= "<hr>" . $result->myPrint();
-               
-               $this->mOutput->setText ( $text ) ;
+               $result = $w->scanString($text);
+
+               $oldTemplateOptions = $this->mCurrentTemplateOptions;
+               $this->mCurrentTemplateOptions = $templateOptions;
+
+               if ($inline) { # Inline rendering off for templates
+                       if (count($result->children) == 1)
+                               $result->children[0]->name = 'IGNORE';
+               }
+
+               if (1)
+                       $text = $result->makeXHTML($this); # No debugging info
+               else
+                       $text = $result->makeXHTML($this).'<hr>'.$text.'<hr>'.$result->myPrint();
+               $this->mCurrentTemplateOptions = $oldTemplateOptions;
+       }
+
+       /** @todo document */
+       function parse($text, & $title, $options, $linestart = true, $clearState = true) {
+               $this->plain_parse($text);
+               $this->mOutput->setText($text);
                return $this->mOutput;
-       }       
-       
        }
 
-?>
+}
+?>
\ No newline at end of file