Your IP : 216.73.216.224
HTMLSax3/Decorators.php 0000666 00000007147 15252667057 0010721 0 ustar 00 <?php
class XML_HTMLSax3_Trim {
var $orig_obj;
var $orig_method;
function XML_HTMLSax3_Trim(&$orig_obj, $orig_method) {
$this->orig_obj =& $orig_obj;
$this->orig_method = $orig_method;
}
function trimData(&$parser, $data) {
$data = trim($data);
if ($data != '') {
$this->orig_obj->{$this->orig_method}($parser, $data);
}
}
}
class XML_HTMLSax3_CaseFolding {
var $orig_obj;
var $orig_open_method;
var $orig_close_method;
function XML_HTMLSax3_CaseFolding(&$orig_obj, $orig_open_method, $orig_close_method) {
$this->orig_obj =& $orig_obj;
$this->orig_open_method = $orig_open_method;
$this->orig_close_method = $orig_close_method;
}
function foldOpen(&$parser, $tag, $attrs=array(), $empty = FALSE) {
$this->orig_obj->{$this->orig_open_method}($parser, strtoupper($tag), $attrs, $empty);
}
function foldClose(&$parser, $tag, $empty = FALSE) {
$this->orig_obj->{$this->orig_close_method}($parser, strtoupper($tag), $empty);
}
}
class XML_HTMLSax3_Linefeed {
var $orig_obj;
var $orig_method;
function XML_HTMLSax3_LineFeed(&$orig_obj, $orig_method) {
$this->orig_obj =& $orig_obj;
$this->orig_method = $orig_method;
}
function breakData(&$parser, $data) {
$data = explode("\n",$data);
foreach ( $data as $chunk ) {
$this->orig_obj->{$this->orig_method}($parser, $chunk);
}
}
}
class XML_HTMLSax3_Tab {
var $orig_obj;
var $orig_method;
function XML_HTMLSax3_Tab(&$orig_obj, $orig_method) {
$this->orig_obj =& $orig_obj;
$this->orig_method = $orig_method;
}
function breakData(&$parser, $data) {
$data = explode("\t",$data);
foreach ( $data as $chunk ) {
$this->orig_obj->{$this->orig_method}($this, $chunk);
}
}
}
class XML_HTMLSax3_Entities_Parsed {
var $orig_obj;
var $orig_method;
function XML_HTMLSax3_Entities_Parsed(&$orig_obj, $orig_method) {
$this->orig_obj =& $orig_obj;
$this->orig_method = $orig_method;
}
function breakData(&$parser, $data) {
$data = preg_split('/(&.+?;)/',$data,-1,PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY);
foreach ( $data as $chunk ) {
$chunk = html_entity_decode($chunk,ENT_NOQUOTES);
$this->orig_obj->{$this->orig_method}($this, $chunk);
}
}
}
if (version_compare(phpversion(), '4.3', '<') && !function_exists('html_entity_decode') ) {
function html_entity_decode($str, $style=ENT_NOQUOTES) {
return strtr($str,
array_flip(get_html_translation_table(HTML_ENTITIES,$style)));
}
}
class XML_HTMLSax3_Entities_Unparsed {
var $orig_obj;
var $orig_method;
function XML_HTMLSax3_Entities_Unparsed(&$orig_obj, $orig_method) {
$this->orig_obj =& $orig_obj;
$this->orig_method = $orig_method;
}
function breakData(&$parser, $data) {
$data = preg_split('/(&.+?;)/',$data,-1,PREG_SPLIT_DELIM_CAPTURE | PREG_SPLIT_NO_EMPTY);
foreach ( $data as $chunk ) {
$this->orig_obj->{$this->orig_method}($this, $chunk);
}
}
}
class XML_HTMLSax3_Escape_Stripper {
var $orig_obj;
var $orig_method;
function XML_HTMLSax3_Escape_Stripper(&$orig_obj, $orig_method) {
$this->orig_obj =& $orig_obj;
$this->orig_method = $orig_method;
}
function strip(&$parser, $data) {
if ( substr($data,0,2) == '--' ) {
$patterns = array(
'/^\-\-/', // Opening comment: --
'/\-\-$/', // Closing comment: --
);
$data = preg_replace($patterns,'',$data);
} else if ( substr($data,0,1) == '[' ) {
$patterns = array(
'/^\[.*CDATA.*\[/s', // Opening CDATA
'/\].*\]$/s', // Closing CDATA
);
$data = preg_replace($patterns,'',$data);
}
$this->orig_obj->{$this->orig_method}($this, $data);
}
}
?> HTMLSax3/States.php 0000666 00000012027 15252667057 0010050 0 ustar 00 <?php
define('XML_HTMLSAX3_STATE_STOP', 0);
define('XML_HTMLSAX3_STATE_START', 1);
define('XML_HTMLSAX3_STATE_TAG', 2);
define('XML_HTMLSAX3_STATE_OPENING_TAG', 3);
define('XML_HTMLSAX3_STATE_CLOSING_TAG', 4);
define('XML_HTMLSAX3_STATE_ESCAPE', 6);
define('XML_HTMLSAX3_STATE_JASP', 7);
define('XML_HTMLSAX3_STATE_PI', 8);
class XML_HTMLSax3_StartingState {
function parse(&$context) {
$data = $context->scanUntilString('<');
if ($data != '') {
$context->handler_object_data->
{$context->handler_method_data}($context->htmlsax, $data);
}
$context->IgnoreCharacter();
return XML_HTMLSAX3_STATE_TAG;
}
}
class XML_HTMLSax3_TagState {
function parse(&$context) {
switch($context->ScanCharacter()) {
case '/':
return XML_HTMLSAX3_STATE_CLOSING_TAG;
break;
case '?':
return XML_HTMLSAX3_STATE_PI;
break;
case '%':
return XML_HTMLSAX3_STATE_JASP;
break;
case '!':
return XML_HTMLSAX3_STATE_ESCAPE;
break;
default:
$context->unscanCharacter();
return XML_HTMLSAX3_STATE_OPENING_TAG;
}
}
}
class XML_HTMLSax3_ClosingTagState {
function parse(&$context) {
$tag = $context->scanUntilCharacters('/>');
if ($tag != '') {
$char = $context->scanCharacter();
if ($char == '/') {
$char = $context->scanCharacter();
if ($char != '>') {
$context->unscanCharacter();
}
}
$context->handler_object_element->
{$context->handler_method_closing}($context->htmlsax, $tag, FALSE);
}
return XML_HTMLSAX3_STATE_START;
}
}
class XML_HTMLSax3_OpeningTagState {
function parseAttributes(&$context) {
$Attributes = array();
$context->ignoreWhitespace();
$attributename = $context->scanUntilCharacters("=/> \n\r\t");
while ($attributename != '') {
$attributevalue = NULL;
$context->ignoreWhitespace();
$char = $context->scanCharacter();
if ($char == '=') {
$context->ignoreWhitespace();
$char = $context->ScanCharacter();
if ($char == '"') {
$attributevalue= $context->scanUntilString('"');
$context->IgnoreCharacter();
} else if ($char == "'") {
$attributevalue = $context->scanUntilString("'");
$context->IgnoreCharacter();
} else {
$context->unscanCharacter();
$attributevalue =
$context->scanUntilCharacters("> \n\r\t");
}
} else if ($char !== NULL) {
$attributevalue = NULL;
$context->unscanCharacter();
}
$Attributes[$attributename] = $attributevalue;
$context->ignoreWhitespace();
$attributename = $context->scanUntilCharacters("=/> \n\r\t");
}
return $Attributes;
}
function parse(&$context) {
$tag = $context->scanUntilCharacters("/> \n\r\t");
if ($tag != '') {
$this->attrs = array();
$Attributes = $this->parseAttributes($context);
$char = $context->scanCharacter();
if ($char == '/') {
$char = $context->scanCharacter();
if ($char != '>') {
$context->unscanCharacter();
}
$context->handler_object_element->
{$context->handler_method_opening}($context->htmlsax, $tag,
$Attributes, TRUE);
$context->handler_object_element->
{$context->handler_method_closing}($context->htmlsax, $tag,
TRUE);
} else {
$context->handler_object_element->
{$context->handler_method_opening}($context->htmlsax, $tag,
$Attributes, FALSE);
}
}
return XML_HTMLSAX3_STATE_START;
}
}
class XML_HTMLSax3_EscapeState {
function parse(&$context) {
$char = $context->ScanCharacter();
if ($char == '-') {
$char = $context->ScanCharacter();
if ($char == '-') {
$context->unscanCharacter();
$context->unscanCharacter();
$text = $context->scanUntilString('-->');
$text .= $context->scanCharacter();
$text .= $context->scanCharacter();
} else {
$context->unscanCharacter();
$text = $context->scanUntilString('>');
}
} else if ( $char == '[') {
$context->unscanCharacter();
$text = $context->scanUntilString(']>');
$text.= $context->scanCharacter();
} else {
$context->unscanCharacter();
$text = $context->scanUntilString('>');
}
$context->IgnoreCharacter();
if ($text != '') {
$context->handler_object_escape->
{$context->handler_method_escape}($context->htmlsax, $text);
}
return XML_HTMLSAX3_STATE_START;
}
}
class XML_HTMLSax3_JaspState {
function parse(&$context) {
$text = $context->scanUntilString('%>');
if ($text != '') {
$context->handler_object_jasp->
{$context->handler_method_jasp}($context->htmlsax, $text);
}
$context->IgnoreCharacter();
$context->IgnoreCharacter();
return XML_HTMLSAX3_STATE_START;
}
}
class XML_HTMLSax3_PiState {
function parse(&$context) {
$target = $context->scanUntilCharacters(" \n\r\t");
$data = $context->scanUntilString('?>');
if ($data != '') {
$context->handler_object_pi->
{$context->handler_method_pi}($context->htmlsax, $target, $data);
}
$context->IgnoreCharacter();
$context->IgnoreCharacter();
return XML_HTMLSAX3_STATE_START;
}
}
?> HTMLSax3/index.php 0000666 00000000002 15252667057 0007702 0 ustar 00 X
index.php 0000666 00000000002 15252667057 0006377 0 ustar 00 X
safehtml.php 0000666 00000022246 15252667057 0007111 0 ustar 00 <?php
/**
* SafeHTML Parser
*
* @package SafeHTML
* @author Roman Ivanov <thingol@mail.ru>
* @copyright 2004-2005 Roman Ivanov
* @license http://www.debian.org/misc/bsd.license BSD License (3 Clause)
* @version 1.3.7
* @link http://pixel-apes.com/safehtml/
*/
if (!defined('_ECRIRE_INC_VERSION')) return;
require_once(XML_HTMLSAX3 . 'HTMLSax3.php');
class SafeHTML
{
var $_xhtml = '';
var $_counter = array();
var $_stack = array();
var $_dcCounter = array();
var $_dcStack = array();
var $_listScope = 0;
var $_liStack = array();
var $_protoRegexps = array();
var $_cssRegexps = array();
var $singleTags = array('area', 'br', 'img', 'input', 'hr', 'wbr', );
var $deleteTags = array(
'applet', 'base', 'basefont', 'bgsound', 'blink', 'body',
'embed', 'frame', 'frameset', 'head', 'html', 'ilayer',
'iframe', 'layer', 'link', 'meta', 'object', 'style',
'title', 'script',
);
var $deleteTagsContent = array('script', 'style', 'title', 'xml', );
var $protocolFiltering = 'white';
var $blackProtocols = array(
'about', 'chrome', 'data', 'disk', 'hcp',
'help', 'javascript', 'livescript', 'lynxcgi', 'lynxexec',
'ms-help', 'ms-its', 'mhtml', 'mocha', 'opera',
'res', 'resource', 'shell', 'vbscript', 'view-source',
'vnd.ms.radio', 'wysiwyg',
);
var $whiteProtocols = array(
'ed2k', 'file', 'ftp', 'gopher', 'http', 'https',
'irc', 'mailto', 'news', 'nntp', 'telnet', 'webcal',
'xmpp', 'callto',
);
var $protocolAttributes = array(
'action', 'background', 'codebase', 'dynsrc', 'href', 'lowsrc', 'src',
);
var $cssKeywords = array(
'absolute', 'behavior', 'behaviour', 'content', 'expression',
'fixed', 'include-source', 'moz-binding',
);
var $noClose = array();
var $closeParagraph = array(
'address', 'blockquote', 'center', 'dd', 'dir', 'div',
'dl', 'dt', 'h1', 'h2', 'h3', 'h4',
'h5', 'h6', 'hr', 'isindex', 'listing', 'marquee',
'menu', 'multicol', 'ol', 'p', 'plaintext', 'pre',
'table', 'ul', 'xmp',
);
var $tableTags = array(
'caption', 'col', 'colgroup', 'tbody', 'td', 'tfoot', 'th',
'thead', 'tr',
);
var $listTags = array('dir', 'menu', 'ol', 'ul', 'dl', );
var $attributes = array('dynsrc', 'id', 'name', );
var $attributesNS = array('xml:lang', );
function SafeHTML()
{
//making regular expressions based on Proto & CSS arrays
foreach ($this->blackProtocols as $proto) {
$preg = "/[\s\x01-\x1F]*";
for ($i=0; $i<strlen($proto); $i++) {
$preg .= $proto{$i} . "[\s\x01-\x1F]*";
}
$preg .= ":/i";
$this->_protoRegexps[] = $preg;
}
foreach ($this->cssKeywords as $css) {
$this->_cssRegexps[] = '/' . $css . '/i';
}
return true;
}
function _writeAttrs ($attrs,$tag=null)
{
if (is_array($attrs)) {
foreach ($attrs as $name => $value) {
$name = strtolower($name);
if (strpos($name, 'on') === 0) {
continue;
}
if (strpos($name, 'data') === 0) {
continue;
}
if ($tag !='a' AND in_array($name, $this->attributes)) {
continue;
}
if (!preg_match("/^[a-z0-9-]+$/i", $name)) {
if (!in_array($name, $this->attributesNS))
{
continue;
}
}
if (($value === TRUE) || (is_null($value))) {
$value = $name;
}
if ($name == 'style') {
// removes insignificant backslahes
$value = str_replace("\\", '', $value);
// removes CSS comments
while (1)
{
$_value = preg_replace("!/\*.*?\*/!s", '', $value);
if ($_value == $value) break;
$value = $_value;
}
// replace all & to &
$value = str_replace('&', '&', $value);
$value = str_replace('&', '&', $value);
foreach ($this->_cssRegexps as $css) {
if (preg_match($css, $value)) {
continue 2;
}
}
foreach ($this->_protoRegexps as $proto) {
if (preg_match($proto, $value)) {
continue 2;
}
}
}
$tempval = preg_replace_callback('/&#(\d+);?/m', create_function('$m', 'return chr($m[1]);'), $value);
$tempval = preg_replace_callback('/&#x([0-9a-f]+);?/mi', create_function('$m', 'return chr(hexdec($m[1]));'), $tempval);
if ((in_array($name, $this->protocolAttributes)) &&
(strpos($tempval, ':') !== false))
{
if ($this->protocolFiltering == 'black') {
foreach ($this->_protoRegexps as $proto) {
if (preg_match($proto, $tempval)) continue 2;
}
} else {
$_tempval = explode(':', $tempval);
$proto = $_tempval[0];
if (!in_array($proto, $this->whiteProtocols)) {
continue;
}
}
}
$value = str_replace("\"", """, $value);
$this->_xhtml .= ' ' . $name . '="' . $value . '"';
}
}
return true;
}
function _openHandler(&$parser, $name, $attrs)
{
$name = strtolower($name);
if (in_array($name, $this->deleteTagsContent)) {
array_push($this->_dcStack, $name);
$this->_dcCounter[$name] = isset($this->_dcCounter[$name]) ? $this->_dcCounter[$name]+1 : 1;
}
if (count($this->_dcStack) != 0) {
return true;
}
if (in_array($name, $this->deleteTags)) {
return true;
}
if (!preg_match("/^[a-z0-9]+$/i", $name)) {
if (preg_match("!(?:\@|://)!i", $name)) {
$this->_xhtml .= '<' . $name . '>';
}
return true;
}
if (in_array($name, $this->singleTags)) {
$this->_xhtml .= '<' . $name;
$this->_writeAttrs($attrs);
$this->_xhtml .= ' />';
return true;
}
// TABLES: cannot open table elements when we are not inside table
if ((isset($this->_counter['table'])) && ($this->_counter['table'] <= 0)
&& (in_array($name, $this->tableTags)))
{
return true;
}
// PARAGRAPHS: close paragraph when closeParagraph tags opening
if ((in_array($name, $this->closeParagraph)) && (in_array('p', $this->_stack))) {
$this->_closeHandler($parser, 'p');
}
// LISTS: we should close <li> if <li> of the same level opening
if ($name == 'li' && count($this->_liStack) &&
$this->_listScope == $this->_liStack[count($this->_liStack)-1])
{
$this->_closeHandler($parser, 'li');
}
// LISTS: we want to know on what nesting level of lists we are
if (in_array($name, $this->listTags)) {
$this->_listScope++;
}
if ($name == 'li') {
array_push($this->_liStack, $this->_listScope);
}
$this->_xhtml .= '<' . $name;
$this->_writeAttrs($attrs,$name);
$this->_xhtml .= '>';
array_push($this->_stack,$name);
$this->_counter[$name] = isset($this->_counter[$name]) ? $this->_counter[$name]+1 : 1;
return true;
}
function _closeHandler(&$parser, $name)
{
$name = strtolower($name);
if (isset($this->_dcCounter[$name]) && ($this->_dcCounter[$name] > 0) &&
(in_array($name, $this->deleteTagsContent)))
{
while ($name != ($tag = array_pop($this->_dcStack))) {
$this->_dcCounter[$tag]--;
}
$this->_dcCounter[$name]--;
}
if (count($this->_dcStack) != 0) {
return true;
}
if ((isset($this->_counter[$name])) && ($this->_counter[$name] > 0)) {
while ($name != ($tag = array_pop($this->_stack))) {
$this->_closeTag($tag);
}
$this->_closeTag($name);
}
return true;
}
function _closeTag($tag)
{
if (!in_array($tag, $this->noClose)) {
$this->_xhtml .= '</' . $tag . '>';
}
$this->_counter[$tag]--;
if (in_array($tag, $this->listTags)) {
$this->_listScope--;
}
if ($tag == 'li') {
array_pop($this->_liStack);
}
return true;
}
function _dataHandler(&$parser, $data)
{
if (count($this->_dcStack) == 0) {
$this->_xhtml .= $data;
}
return true;
}
function _escapeHandler(&$parser, $data)
{
return true;
}
function getXHTML ()
{
while ($tag = array_pop($this->_stack)) {
$this->_closeTag($tag);
}
return $this->_xhtml;
}
function clear()
{
$this->_xhtml = '';
return true;
}
function parse($doc)
{
// Save all '<' symbols
$doc = preg_replace("/<(?=[^a-zA-Z\/\!\?\%])/", '<', $doc);
// Web documents shouldn't contains \x00 symbol
$doc = str_replace("\x00", '', $doc);
// Opera6 bug workaround
$doc = str_replace("\xC0\xBC", '<', $doc);
// UTF-7 encoding ASCII decode
$doc = $this->repackUTF7($doc);
// Instantiate the parser
$parser= new XML_HTMLSax3();
// Set up the parser
$parser->set_object($this);
$parser->set_element_handler('_openHandler','_closeHandler');
$parser->set_data_handler('_dataHandler');
$parser->set_escape_handler('_escapeHandler');
$parser->parse($doc);
return $this->getXHTML();
}
function repackUTF7($str)
{
return preg_replace_callback('!\+([0-9a-zA-Z/]+)\-!', array($this, 'repackUTF7Callback'), $str);
}
function repackUTF7Callback($str)
{
$str = base64_decode($str[1]);
$str = preg_replace_callback('/^((?:\x00.)*)((?:[^\x00].)+)/', array($this, 'repackUTF7Back'), $str);
return preg_replace('/\x00(.)/', '$1', $str);
}
function repackUTF7Back($str)
{
return $str[1].'+'.rtrim(base64_encode($str[2]), '=').'-';
}
}
?>
HTMLSax3.php 0000666 00000023731 15252667057 0006611 0 ustar 00 <?php
// +----------------------------------------------------------------------+
// | Copyright (c) 1997-2002 The PHP Group |
// +----------------------------------------------------------------------+
// | This source file is subject to version 2.02 of the PHP license, |
// | that is bundled with this package in the file LICENSE, and is |
// | available at through the world-wide-web at |
// | http://www.php.net/license/3_0.txt. |
// | If you did not receive a copy of the PHP license and are unable to |
// | obtain it through the world-wide-web, please send a note to |
// | license@php.net so we can mail you a copy immediately. |
// +----------------------------------------------------------------------+
// | Authors: Alexander Zhukov <alex@veresk.ru> Original port from Python |
// | Authors: Harry Fuecks <hfuecks@phppatterns.com> Port to PEAR + more |
// | Authors: Many @ Sitepointforums Advanced PHP Forums |
// +----------------------------------------------------------------------+
//
if (!defined('_ECRIRE_INC_VERSION')) return;
if (!defined('XML_HTMLSAX3')) {
define('XML_HTMLSAX3', 'XML/');
}
require_once(XML_HTMLSAX3 . 'HTMLSax3/States.php');
require_once(XML_HTMLSAX3 . 'HTMLSax3/Decorators.php');
class XML_HTMLSax3_StateParser {
var $htmlsax;
var $handler_object_element;
var $handler_method_opening;
var $handler_method_closing;
var $handler_object_data;
var $handler_method_data;
var $handler_object_pi;
var $handler_method_pi;
var $handler_object_jasp;
var $handler_method_jasp;
var $handler_object_escape;
var $handler_method_escape;
var $handler_default;
var $parser_options = array();
var $rawtext;
var $position;
var $length;
var $State = array();
function XML_HTMLSax3_StateParser (& $htmlsax) {
$this->htmlsax = & $htmlsax;
$this->State[XML_HTMLSAX3_STATE_START] = new XML_HTMLSax3_StartingState();
$this->State[XML_HTMLSAX3_STATE_CLOSING_TAG] = new XML_HTMLSax3_ClosingTagState();
$this->State[XML_HTMLSAX3_STATE_TAG] = new XML_HTMLSax3_TagState();
$this->State[XML_HTMLSAX3_STATE_OPENING_TAG] = new XML_HTMLSax3_OpeningTagState();
$this->State[XML_HTMLSAX3_STATE_PI] = new XML_HTMLSax3_PiState();
$this->State[XML_HTMLSAX3_STATE_JASP] = new XML_HTMLSax3_JaspState();
$this->State[XML_HTMLSAX3_STATE_ESCAPE] = new XML_HTMLSax3_EscapeState();
}
function unscanCharacter() {
$this->position -= 1;
}
function ignoreCharacter() {
$this->position += 1;
}
function scanCharacter() {
if ($this->position < $this->length) {
return $this->rawtext{$this->position++};
}
}
function scanUntilString($string) {
$start = $this->position;
$this->position = strpos($this->rawtext, $string, $start);
if ($this->position === FALSE) {
$this->position = $this->length;
}
return substr($this->rawtext, $start, $this->position - $start);
}
function scanUntilCharacters($string) {}
function ignoreWhitespace() {}
function parse($data) {
if ($this->parser_options['XML_OPTION_TRIM_DATA_NODES']==1) {
$decorator = new XML_HTMLSax3_Trim(
$this->handler_object_data,
$this->handler_method_data);
$this->handler_object_data =& $decorator;
$this->handler_method_data = 'trimData';
}
if ($this->parser_options['XML_OPTION_CASE_FOLDING']==1) {
$open_decor = new XML_HTMLSax3_CaseFolding(
$this->handler_object_element,
$this->handler_method_opening,
$this->handler_method_closing);
$this->handler_object_element =& $open_decor;
$this->handler_method_opening ='foldOpen';
$this->handler_method_closing ='foldClose';
}
if ($this->parser_options['XML_OPTION_LINEFEED_BREAK']==1) {
$decorator = new XML_HTMLSax3_Linefeed(
$this->handler_object_data,
$this->handler_method_data);
$this->handler_object_data =& $decorator;
$this->handler_method_data = 'breakData';
}
if ($this->parser_options['XML_OPTION_TAB_BREAK']==1) {
$decorator = new XML_HTMLSax3_Tab(
$this->handler_object_data,
$this->handler_method_data);
$this->handler_object_data =& $decorator;
$this->handler_method_data = 'breakData';
}
if ($this->parser_options['XML_OPTION_ENTITIES_UNPARSED']==1) {
$decorator = new XML_HTMLSax3_Entities_Unparsed(
$this->handler_object_data,
$this->handler_method_data);
$this->handler_object_data =& $decorator;
$this->handler_method_data = 'breakData';
}
if ($this->parser_options['XML_OPTION_ENTITIES_PARSED']==1) {
$decorator = new XML_HTMLSax3_Entities_Parsed(
$this->handler_object_data,
$this->handler_method_data);
$this->handler_object_data =& $decorator;
$this->handler_method_data = 'breakData';
}
// Note switched on by default
if ($this->parser_options['XML_OPTION_STRIP_ESCAPES']==1) {
$decorator = new XML_HTMLSax3_Escape_Stripper(
$this->handler_object_escape,
$this->handler_method_escape);
$this->handler_object_escape =& $decorator;
$this->handler_method_escape = 'strip';
}
$this->rawtext = $data;
$this->length = strlen($data);
$this->position = 0;
$this->_parse();
}
function _parse($state = XML_HTMLSAX3_STATE_START) {
do {
$state = $this->State[$state]->parse($this);
} while ($state != XML_HTMLSAX3_STATE_STOP &&
$this->position < $this->length);
}
}
class XML_HTMLSax3_StateParser_Lt430 extends XML_HTMLSax3_StateParser {
function XML_HTMLSax3_StateParser_Lt430(& $htmlsax) {
parent::XML_HTMLSax3_StateParser($htmlsax);
$this->parser_options['XML_OPTION_TRIM_DATA_NODES'] = 0;
$this->parser_options['XML_OPTION_CASE_FOLDING'] = 0;
$this->parser_options['XML_OPTION_LINEFEED_BREAK'] = 0;
$this->parser_options['XML_OPTION_TAB_BREAK'] = 0;
$this->parser_options['XML_OPTION_ENTITIES_PARSED'] = 0;
$this->parser_options['XML_OPTION_ENTITIES_UNPARSED'] = 0;
$this->parser_options['XML_OPTION_STRIP_ESCAPES'] = 0;
}
function scanUntilCharacters($string) {
$startpos = $this->position;
while ($this->position < $this->length && strpos($string, $this->rawtext{$this->position}) === FALSE) {
$this->position++;
}
return substr($this->rawtext, $startpos, $this->position - $startpos);
}
function ignoreWhitespace() {
while ($this->position < $this->length &&
strpos(" \n\r\t", $this->rawtext{$this->position}) !== FALSE) {
$this->position++;
}
}
function parse($data) {
parent::parse($data);
}
}
class XML_HTMLSax3_StateParser_Gtet430 extends XML_HTMLSax3_StateParser {
function XML_HTMLSax3_StateParser_Gtet430(& $htmlsax) {
parent::XML_HTMLSax3_StateParser($htmlsax);
$this->parser_options['XML_OPTION_TRIM_DATA_NODES'] = 0;
$this->parser_options['XML_OPTION_CASE_FOLDING'] = 0;
$this->parser_options['XML_OPTION_LINEFEED_BREAK'] = 0;
$this->parser_options['XML_OPTION_TAB_BREAK'] = 0;
$this->parser_options['XML_OPTION_ENTITIES_PARSED'] = 0;
$this->parser_options['XML_OPTION_ENTITIES_UNPARSED'] = 0;
$this->parser_options['XML_OPTION_STRIP_ESCAPES'] = 0;
}
function scanUntilCharacters($string) {
$startpos = $this->position;
$length = strcspn($this->rawtext, $string, $startpos);
$this->position += $length;
return substr($this->rawtext, $startpos, $length);
}
function ignoreWhitespace() {
$this->position += strspn($this->rawtext, " \n\r\t", $this->position);
}
function parse($data) {
parent::parse($data);
}
}
class XML_HTMLSax3_NullHandler {
function DoNothing() {
}
}
class XML_HTMLSax3 {
var $state_parser;
function XML_HTMLSax3() {
if (version_compare(phpversion(), '4.3', 'ge')) {
$this->state_parser = new XML_HTMLSax3_StateParser_Gtet430($this);
} else {
$this->state_parser = new XML_HTMLSax3_StateParser_Lt430($this);
}
$nullhandler = new XML_HTMLSax3_NullHandler();
$this->set_object($nullhandler);
$this->set_element_handler('DoNothing', 'DoNothing');
$this->set_data_handler('DoNothing');
$this->set_pi_handler('DoNothing');
$this->set_jasp_handler('DoNothing');
$this->set_escape_handler('DoNothing');
}
function set_object(&$object) {
if ( is_object($object) ) {
$this->state_parser->handler_default =& $object;
return true;
} else {
require_once('PEAR.php');
PEAR::raiseError('XML_HTMLSax3::set_object requires '.
'an object instance');
}
}
function set_option($name, $value=1) {
if ( array_key_exists($name,$this->state_parser->parser_options) ) {
$this->state_parser->parser_options[$name] = $value;
return true;
} else {
require_once('PEAR.php');
PEAR::raiseError('XML_HTMLSax3::set_option('.$name.') illegal');
}
}
function set_data_handler($data_method) {
$this->state_parser->handler_object_data =& $this->state_parser->handler_default;
$this->state_parser->handler_method_data = $data_method;
}
function set_element_handler($opening_method, $closing_method) {
$this->state_parser->handler_object_element =& $this->state_parser->handler_default;
$this->state_parser->handler_method_opening = $opening_method;
$this->state_parser->handler_method_closing = $closing_method;
}
function set_pi_handler($pi_method) {
$this->state_parser->handler_object_pi =& $this->state_parser->handler_default;
$this->state_parser->handler_method_pi = $pi_method;
}
function set_escape_handler($escape_method) {
$this->state_parser->handler_object_escape =& $this->state_parser->handler_default;
$this->state_parser->handler_method_escape = $escape_method;
}
function set_jasp_handler ($jasp_method) {
$this->state_parser->handler_object_jasp =& $this->state_parser->handler_default;
$this->state_parser->handler_method_jasp = $jasp_method;
}
function get_current_position() {
return $this->state_parser->position;
}
function get_length() {
return $this->state_parser->length;
}
function parse($data) {
$this->state_parser->parse($data);
}
}
?>