<?php
/*************************************************************************************************
**
**	XHTML validator
**	Author: Jan Dolezel, http://www.eithel.net
**	Last modified: 2005-10-26
**
**	License Information:
**	-------------------------------------------------------------------------
**	Copyright (C) 2005 Jan Dolezel
**
**	This program is free software; you can redistribute it 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 more details.
**
*************************************************************************************************/

$temp = "";

class Parser {
	var $xmlParser;
	var $depth;
	var $deep;
	var $pEnvelope = true;
	var $parsedText = "";
	var $mappedTags = array(
		"b"	=> "strong",
		"i"	=> "em");
	var $allowedOpenTags = array("div", "p", "address", "strong", "em", "h4", "h5", "h6", "ul", "ol", "li");
	var $allowedSingleTags = array("br", "img");
	var $allowedOpenAttribTags = array("a", "img", "h3");
	var $attributes = array(
		"a"	=> array("href", "title"),
		"img"	=> array("src", "alt"),
		"h3" => array("id"));
	var $allowedCloseTags = array("a", "div", "p", "address", "strong", "em", "h3", "h4", "h5", "h6", "ul", "ol", "li");
	var $newLineTags = array("br", "div", "p", "address", "h3", "h4", "h5", "h6", "ul", "ol", "li");

	function Parser() {
		$this->xmlParser = xml_parser_create();
		xml_set_element_handler($this->xmlParser, array(&$this, "startElement"), array(&$this, "endElement"));
		xml_set_character_data_handler($this->xmlParser, array(&$this, "characterData"));
		$this->depth = array();
		$this->deep = 0;
		foreach(array_merge($this->allowedOpenTags, $this->allowedOpenAttribTags) as $tag) {
			$this->depth[$tag] = 0;
		}
	}

	function startElement($parser, $name, $attrs) {
		global $temp;
		$tag = strtolower($name);
		if(array_key_exists($tag, $this->mappedTags)) $tag = $this->mappedTags[$tag];
		
		if(in_array($tag, $this->allowedOpenTags)) {
			if($this->depth[$tag] == 0) $temp .= "<" . $tag . ">";
			$this->depth[$tag]++;
		} elseif(in_array($tag, $this->allowedOpenAttribTags)) {
			if($this->depth[$tag] == 0) {
				$temp .= "<" . $tag;
				foreach($this->attributes[$tag] as $attrib) {
					if(isset($attrs[$attrib])) $temp .= " " . strtolower($attrib) . "=\"" . $attrs[$attrib] . "\"";
					if(isset($attrs[strtoupper($attrib)])) $temp .= " " . strtolower($attrib) . "=\"" . $attrs[strtoupper($attrib)] . "\"";
				}
				if($tag == "img") {
					$src = $attrs["SRC"];
					if(strpos($src, "http") === false) $src = "http://" . $_SERVER['SERVER_NAME'] . $src;
					if(($gis = @getimagesize($src)) !== false) {
						list($width, $height, $type, $attr) = $gis;
						$temp .= " " . $attr;
					}
				}
				if(in_array($tag, $this->allowedSingleTags)) $temp .= " />";
				else $temp .= ">";
			}
			$this->depth[$tag]++;
		} elseif(in_array($tag, $this->allowedSingleTags)) {
			$temp .= "<" . $tag . " />";
		} elseif($this->deep > 0) {
			$temp .= "&lt;" . $tag;
			foreach($attrs as $key => $attrib) {
				@$temp .= " " . strtolower($key) . "=&quot;" . $attrib . "&quot;";
			}
			$temp .= "&gt;";
		}
		$this->deep++;
	}

	function endElement($parser, $name) {
		global $temp;
		$this->deep--;
		$tag = strtolower($name);
		if(array_key_exists($tag, $this->mappedTags)) $tag = $this->mappedTags[$tag];

		if(in_array($tag, $this->allowedSingleTags)) {
			/* do nothing */
		} elseif(in_array($tag, $this->allowedCloseTags)) {
			$this->depth[$tag]--;
			if($this->depth[$tag] == 0)	$temp .= "</" . $tag . ">";
		} elseif($this->deep > 0) {
			$temp .= "&lt;/" . $tag . "&gt;";
		}
		if(in_array($tag, $this->newLineTags)) {
			$temp .= "\n";
		}
	}

	function characterData($parser, $data) {
		global $temp;
		if(($this->deep == 1) && (trim($data) != "") && $this->pEnvelope) $temp .= "<p>" . $this->applySlashes(htmlspecialchars($data)) . "</p>";
	    else $temp .= $this->applySlashes(htmlspecialchars($data));
	}

	var $trans = array("'" => "\'", "\r" => "", "\n" => " ", "\t" => " ", "   " => " ", "  " => " ");
	function applySlashes($str) {
		$newlength = strlen($str);
		do {
			$oldlength = $newlength;
			$str = stripslashes($str);
			$newlength = strlen($str);
		} while($oldlength != $newlength);
		return strtr($str, $this->trans);
	}

	function parse($string, $env = true) {
		global $temp;
		$tmp = $temp;
		$temp = "";
		$this->pEnvelope = $env;
	    if (!xml_parse($this->xmlParser, "<x>" . $string . "</x>", true)) {
			$temp = $string;
			$this->parsedText = $temp;
			return false;
	    }
		$this->parsedText = $temp;
		$temp = $tmp;
		return true;
	}
	
	function free() {
		xml_parser_free($this->xmlParser);
	}
	
	function getText() {
		return $this->parsedText;
	}
}
?>