/libraries/joomla/filter/input.php

<?php
/**
 * @package     Joomla.Platform
 * @subpackage  Filter
 *
 * @copyright   Copyright (C) 2005 - 2012 Open Source Matters, Inc. All rights reserved.
 * @license     GNU General Public License version 2 or later; see LICENSE
 */

defined('JPATH_PLATFORM') or die;

/**
 * JFilterInput is a class for filtering input from any data source
 *
 * Forked from the php input filter library by: Daniel Morris <dan@rootcube.com>
 * Original Contributors: Gianpaolo Racca, Ghislain Picard, Marco Wandschneider, Chris Tobin and Andrew Eddie.
 *
 * @package     Joomla.Platform
 * @subpackage  Filter
 * @since       11.1
 */
class JFilterInput extends JObject
{
	/**
	 * @var    array  A container for JFilterInput instances.
	 * @since  11.3
	 */
	protected static $instances = array();

	/**
	 * @var    array  An array of permitted tags.
	 * @since  11.1
	 */
	public $tagsArray;

	/**
	 * @var    array  An array of permitted tag attributes.
	 * @since  11.1
	 */
	public $attrArray;

	/**
	 * @var    integer  Method for tags: WhiteList method = 0 (default), BlackList method = 1
	 * @since  11.1
	 */
	public $tagsMethod;

	/**
	 * @var    integer  Method for attributes: WhiteList method = 0 (default), BlackList method = 1
	 * @since  11.1
	 */
	public $attrMethod;

	/**
	 * @var    integer  Only auto clean essentials = 0, Allow clean blacklisted tags/attr = 1
	 * @since  11.1
	 */
	public $xssAuto;

	/**
	 * @var    array  A list of the default blacklisted tags.
	 * @since  11.1
	 */
	public $tagBlacklist = array(
		'applet',
		'body',
		'bgsound',
		'base',
		'basefont',
		'embed',
		'frame',
		'frameset',
		'head',
		'html',
		'id',
		'iframe',
		'ilayer',
		'layer',
		'link',
		'meta',
		'name',
		'object',
		'script',
		'style',
		'title',
		'xml'
	);

	/**
	 * @var    array     A list of the default blacklisted tag attributes.  All event handlers implicit.
	 * @since   11.1
	 */
	public $attrBlacklist = array(
		'action',
		'background',
		'codebase',
		'dynsrc',
		'lowsrc'
	);

	/**
	 * Constructor for inputFilter class. Only first parameter is required.
	 *
	 * @param   array    $tagsArray   List of user-defined tags
	 * @param   array    $attrArray   List of user-defined attributes
	 * @param   integer  $tagsMethod  WhiteList method = 0, BlackList method = 1
	 * @param   integer  $attrMethod  WhiteList method = 0, BlackList method = 1
	 * @param   integer  $xssAuto     Only auto clean essentials = 0, Allow clean blacklisted tags/attr = 1
	 *
	 * @since   11.1
	 */
	public function __construct($tagsArray = array(), $attrArray = array(), $tagsMethod = 0, $attrMethod = 0, $xssAuto = 1)
	{
		// Make sure user defined arrays are in lowercase
		$tagsArray = array_map('strtolower', (array) $tagsArray);
		$attrArray = array_map('strtolower', (array) $attrArray);

		// Assign member variables
		$this->tagsArray = $tagsArray;
		$this->attrArray = $attrArray;
		$this->tagsMethod = $tagsMethod;
		$this->attrMethod = $attrMethod;
		$this->xssAuto = $xssAuto;
	}

	/**
	 * Returns an input filter object, only creating it if it doesn't already exist.
	 *
	 * @param   array    $tagsArray   List of user-defined tags
	 * @param   array    $attrArray   List of user-defined attributes
	 * @param   integer  $tagsMethod  WhiteList method = 0, BlackList method = 1
	 * @param   integer  $attrMethod  WhiteList method = 0, BlackList method = 1
	 * @param   integer  $xssAuto     Only auto clean essentials = 0, Allow clean blacklisted tags/attr = 1
	 *
	 * @return  JFilterInput  The JFilterInput object.
	 *
	 * @since   11.1
	 */
	public static function &getInstance($tagsArray = array(), $attrArray = array(), $tagsMethod = 0, $attrMethod = 0, $xssAuto = 1)
	{
		$sig = md5(serialize(array($tagsArray, $attrArray, $tagsMethod, $attrMethod, $xssAuto)));

		if (empty(self::$instances[$sig]))
		{
			self::$instances[$sig] = new JFilterInput($tagsArray, $attrArray, $tagsMethod, $attrMethod, $xssAuto);
		}

		return self::$instances[$sig];
	}

	/**
	 * Method to be called by another php script. Processes for XSS and
	 * specified bad code.
	 *
	 * @param   mixed   $source  Input string/array-of-string to be 'cleaned'
	 * @param   string  $type    Return type for the variable (INT, UINT, FLOAT, BOOLEAN, WORD, ALNUM, CMD, BASE64, STRING, ARRAY, PATH, NONE)
	 *
	 * @return  mixed  'Cleaned' version of input parameter
	 *
	 * @since   11.1
	 */
	public function clean($source, $type = 'string')
	{
		// Handle the type constraint
		switch (strtoupper($type))
		{
			case 'INT':
			case 'INTEGER':
				// Only use the first integer value
				preg_match('/-?[0-9]+/', (string) $source, $matches);
				$result = @ (int) $matches[0];
				break;

			case 'UINT':
				// Only use the first integer value
				preg_match('/-?[0-9]+/', (string) $source, $matches);
				$result = @ abs((int) $matches[0]);
				break;

			case 'FLOAT':
			case 'DOUBLE':
				// Only use the first floating point value
				preg_match('/-?[0-9]+(\.[0-9]+)?/', (string) $source, $matches);
				$result = @ (float) $matches[0];
				break;

			case 'BOOL':
			case 'BOOLEAN':
				$result = (bool) $source;
				break;

			case 'WORD':
				$result = (string) preg_replace('/[^A-Z_]/i', '', $source);
				break;

			case 'ALNUM':
				$result = (string) preg_replace('/[^A-Z0-9]/i', '', $source);
				break;

			case 'CMD':
				$result = (string) preg_replace('/[^A-Z0-9_\.-]/i', '', $source);
				$result = ltrim($result, '.');
				break;

			case 'BASE64':
				$result = (string) preg_replace('/[^A-Z0-9\/+=]/i', '', $source);
				break;

			case 'STRING':
				$result = (string) $this->_remove($this->_decode((string) $source));
				break;

			case 'HTML':
				$result = (string) $this->_remove((string) $source);
				break;

			case 'ARRAY':
				$result = (array) $source;
				break;

			case 'PATH':
				$pattern = '/^[A-Za-z0-9_-]+[A-Za-z0-9_\.-]*([\\\\\/][A-Za-z0-9_-]+[A-Za-z0-9_\.-]*)*$/';
				preg_match($pattern, (string) $source, $matches);
				$result = @ (string) $matches[0];
				break;

			case 'USERNAME':
				$result = (string) preg_replace('/[\x00-\x1F\x7F<>"\'%&]/', '', $source);
				break;

			default:
				// Are we dealing with an array?
				if (is_array($source))
				{
					foreach ($source as $key => $value)
					{
						// filter element for XSS and other 'bad' code etc.
						if (is_string($value))
						{
							$source[$key] = $this->_remove($this->_decode($value));
						}
					}
					$result = $source;
				}
				else
				{
					// Or a string?
					if (is_string($source) && !empty($source))
					{
						// filter source for XSS and other 'bad' code etc.
						$result = $this->_remove($this->_decode($source));
					}
					else
					{
						// Not an array or string.. return the passed parameter
						$result = $source;
					}
				}
				break;
		}

		return $result;
	}

	/**
	 * Function to determine if contents of an attribute are safe
	 *
	 * @param   array  $attrSubSet  A 2 element array for attribute's name, value
	 *
	 * @return  boolean  True if bad code is detected
	 *
	 * @since   11.1
	 */
	public static function checkAttribute($attrSubSet)
	{
		$attrSubSet[0] = strtolower($attrSubSet[0]);
		$attrSubSet[1] = strtolower($attrSubSet[1]);

		return (((strpos($attrSubSet[1], 'expression') !== false) && ($attrSubSet[0]) == 'style') || (strpos($attrSubSet[1], 'javascript:') !== false) ||
			(strpos($attrSubSet[1], 'behaviour:') !== false) || (strpos($attrSubSet[1], 'vbscript:') !== false) ||
			(strpos($attrSubSet[1], 'mocha:') !== false) || (strpos($attrSubSet[1], 'livescript:') !== false));
	}

	/**
	 * Internal method to iteratively remove all unwanted tags and attributes
	 *
	 * @param   string  $source  Input string to be 'cleaned'
	 *
	 * @return  string  'Cleaned' version of input parameter
	 *
	 * @since   11.1
	 */
	protected function _remove($source)
	{
		$loopCounter = 0;

		// Iteration provides nested tag protection
		while ($source != $this->_cleanTags($source))
		{
			$source = $this->_cleanTags($source);
			$loopCounter++;
		}

		return $source;
	}

	/**
	 * Internal method to strip a string of certain tags
	 *
	 * @param   string  $source  Input string to be 'cleaned'
	 *
	 * @return  string  'Cleaned' version of input parameter
	 *
	 * @since   11.1
	 */
	protected function _cleanTags($source)
	{
		// First, pre-process this for illegal characters inside attribute values
		$source = $this->_escapeAttributeValues($source);
		// In the beginning we don't really have a tag, so everything is postTag
		$preTag = null;
		$postTag = $source;
		$currentSpace = false;
		// Setting to null to deal with undefined variables
		$attr = '';

		// Is there a tag? If so it will certainly start with a '<'.
		$tagOpen_start = strpos($source, '<');

		while ($tagOpen_start !== false)
		{
			// Get some information about the tag we are processing
			$preTag .= substr($postTag, 0, $tagOpen_start);
			$postTag = substr($postTag, $tagOpen_start);
			$fromTagOpen = substr($postTag, 1);
			$tagOpen_end = strpos($fromTagOpen, '>');

			// Check for mal-formed tag where we have a second '<' before the first '>'
			$nextOpenTag = (strlen($postTag) > $tagOpen_start) ? strpos($postTag, '<', $tagOpen_start + 1) : false;
			if (($nextOpenTag !== false) && ($nextOpenTag < $tagOpen_end))
			{
				// At this point we have a mal-formed tag -- remove the offending open
				$postTag = substr($postTag, 0, $tagOpen_start) . substr($postTag, $tagOpen_start + 1);
				$tagOpen_start = strpos($postTag, '<');
				continue;
			}

			// Let's catch any non-terminated tags and skip over them
			if ($tagOpen_end === false)
			{
				$postTag = substr($postTag, $tagOpen_start + 1);
				$tagOpen_start = strpos($postTag, '<');
				continue;
			}

			// Do we have a nested tag?
			$tagOpen_nested = strpos($fromTagOpen, '<');
			$tagOpen_nested_end = strpos(substr($postTag, $tagOpen_end), '>');
			if (($tagOpen_nested !== false) && ($tagOpen_nested < $tagOpen_end))
			{
				$preTag .= substr($postTag, 0, ($tagOpen_nested + 1));
				$postTag = substr($postTag, ($tagOpen_nested + 1));
				$tagOpen_start = strpos($postTag, '<');
				continue;
			}

			// Let's get some information about our tag and setup attribute pairs
			$tagOpen_nested = (strpos($fromTagOpen, '<') + $tagOpen_start + 1);
			$currentTag = substr($fromTagOpen, 0, $tagOpen_end);
			$tagLength = strlen($currentTag);
			$tagLeft = $currentTag;
			$attrSet = array();
			$currentSpace = strpos($tagLeft, ' ');

			// Are we an open tag or a close tag?
			if (substr($currentTag, 0, 1) == '/')
			{
				// Close Tag
				$isCloseTag = true;
				list ($tagName) = explode(' ', $currentTag);
				$tagName = substr($tagName, 1);
			}
			else
			{
				// Open Tag
				$isCloseTag = false;
				list ($tagName) = explode(' ', $currentTag);
			}

			/*
			 * Exclude all "non-regular" tagnames
			 * OR no tagname
			 * OR remove if xssauto is on and tag is blacklisted
			 */
			if ((!preg_match("/^[a-z][a-z0-9]*$/i", $tagName)) || (!$tagName) || ((in_array(strtolower($tagName), $this->tagBlacklist)) && ($this->xssAuto)))
			{
				$postTag = substr($postTag, ($tagLength + 2));
				$tagOpen_start = strpos($postTag, '<');
				// Strip tag
				continue;
			}

			/*
			 * Time to grab any attributes from the tag... need this section in
			 * case attributes have spaces in the values.
			 */
			while ($currentSpace !== false)
			{
				$attr = '';
				$fromSpace = substr($tagLeft, ($currentSpace + 1));
				$nextEqual = strpos($fromSpace, '=');
				$nextSpace = strpos($fromSpace, ' ');
				$openQuotes = strpos($fromSpace, '"');
				$closeQuotes = strpos(substr($fromSpace, ($openQuotes + 1)), '"') + $openQuotes + 1;

				$startAtt = '';
				$startAttPosition = 0;

				// Find position of equal and open quotes ignoring
				if (preg_match('#\s*=\s*\"#', $fromSpace, $matches, PREG_OFFSET_CAPTURE))
				{
					$startAtt = $matches[0][0];
					$startAttPosition = $matches[0][1];
					$closeQuotes = strpos(substr($fromSpace, ($startAttPosition + strlen($startAtt))), '"') + $startAttPosition + strlen($startAtt);
					$nextEqual = $startAttPosition + strpos($startAtt, '=');
					$openQuotes = $startAttPosition + strpos($startAtt, '"');
					$nextSpace = strpos(substr($fromSpace, $closeQuotes), ' ') + $closeQuotes;
				}

				// Do we have an attribute to process? [check for equal sign]
				if ($fromSpace != '/' && (($nextEqual && $nextSpace && $nextSpace < $nextEqual) || !$nextEqual))
				{
					if (!$nextEqual)
					{
						$attribEnd = strpos($fromSpace, '/') - 1;
					}
					else
					{
						$attribEnd = $nextSpace - 1;
					}
					// If there is an ending, use this, if not, do not worry.
					if ($attribEnd > 0)
					{
						$fromSpace = substr($fromSpace, $attribEnd + 1);
					}
				}
				if (strpos($fromSpace, '=') !== false)
				{
					// If the attribute value is wrapped in quotes we need to grab the substring from
					// the closing quote, otherwise grab until the next space.
					if (($openQuotes !== false) && (strpos(substr($fromSpace, ($openQuotes + 1)), '"') !== false))
					{
						$attr = substr($fromSpace, 0, ($closeQuotes + 1));
					}
					else
					{
						$attr = substr($fromSpace, 0, $nextSpace);
					}
				}
				// No more equal signs so add any extra text in the tag into the attribute array [eg. checked]
				else
				{
					if ($fromSpace != '/')
					{
						$attr = substr($fromSpace, 0, $nextSpace);
					}
				}

				// Last Attribute Pair
				if (!$attr && $fromSpace != '/')
				{
					$attr = $fromSpace;
				}

				// Add attribute pair to the attribute array
				$attrSet[] = $attr;

				// Move search point and continue iteration
				$tagLeft = substr($fromSpace, strlen($attr));
				$currentSpace = strpos($tagLeft, ' ');
			}

			// Is our tag in the user input array?
			$tagFound = in_array(strtolower($tagName), $this->tagsArray);

			// If the tag is allowed let's append it to the output string.
			if ((!$tagFound && $this->tagsMethod) || ($tagFound && !$this->tagsMethod))
			{
				// Reconstruct tag with allowed attributes
				if (!$isCloseTag)
				{
					// Open or single tag
					$attrSet = $this->_cleanAttributes($attrSet);
					$preTag .= '<' . $tagName;
					for ($i = 0, $count = count($attrSet); $i < $count; $i++)
					{
						$preTag .= ' ' . $attrSet[$i];
					}

					// Reformat single tags to XHTML
					if (strpos($fromTagOpen, '</' . $tagName))
					{
						$preTag .= '>';
					}
					else
					{
						$preTag .= ' />';
					}
				}
				// Closing tag
				else
				{
					$preTag .= '</' . $tagName . '>';
				}
			}

			// Find next tag's start and continue iteration
			$postTag = substr($postTag, ($tagLength + 2));
			$tagOpen_start = strpos($postTag, '<');
		}

		// Append any code after the end of tags and return
		if ($postTag != '<')
		{
			$preTag .= $postTag;
		}

		return $preTag;
	}

	/**
	 * Internal method to strip a tag of certain attributes
	 *
	 * @param   array  $attrSet  Array of attribute pairs to filter
	 *
	 * @return  array  Filtered array of attribute pairs
	 *
	 * @since   11.1
	 */
	protected function _cleanAttributes($attrSet)
	{
		// Initialise variables.
		$newSet = array();

		$count = count($attrSet);
		// Iterate through attribute pairs
		for ($i = 0; $i < $count; $i++)
		{
			// Skip blank spaces
			if (!$attrSet[$i])
			{
				continue;
			}

			// Split into name/value pairs
			$attrSubSet = explode('=', trim($attrSet[$i]), 2);
			// Take the last attribute in case there is an attribute with no value
			$attrSubSet[0] = array_pop(explode(' ', trim($attrSubSet[0])));

			// Remove all "non-regular" attribute names
			// AND blacklisted attributes

			if ((!preg_match('/[a-z]*$/i', $attrSubSet[0]))
				|| (($this->xssAuto) && ((in_array(strtolower($attrSubSet[0]), $this->attrBlacklist))
				|| (substr($attrSubSet[0], 0, 2) == 'on'))))
			{
				continue;
			}

			// XSS attribute value filtering
			if (isset($attrSubSet[1]))
			{
				// trim leading and trailing spaces
				$attrSubSet[1] = trim($attrSubSet[1]);
				// strips unicode, hex, etc
				$attrSubSet[1] = str_replace('&#', '', $attrSubSet[1]);
				// Strip normal newline within attr value
				$attrSubSet[1] = preg_replace('/[\n\r]/', '', $attrSubSet[1]);
				// Strip double quotes
				$attrSubSet[1] = str_replace('"', '', $attrSubSet[1]);
				// Convert single quotes from either side to doubles (Single quotes shouldn't be used to pad attr values)
				if ((substr($attrSubSet[1], 0, 1) == "'") && (substr($attrSubSet[1], (strlen($attrSubSet[1]) - 1), 1) == "'"))
				{
					$attrSubSet[1] = substr($attrSubSet[1], 1, (strlen($attrSubSet[1]) - 2));
				}
				// Strip slashes
				$attrSubSet[1] = stripslashes($attrSubSet[1]);
			}
			else
			{
				continue;
			}

			// Autostrip script tags
			if (self::checkAttribute($attrSubSet))
			{
				continue;
			}

			// Is our attribute in the user input array?
			$attrFound = in_array(strtolower($attrSubSet[0]), $this->attrArray);

			// If the tag is allowed lets keep it
			if ((!$attrFound && $this->attrMethod) || ($attrFound && !$this->attrMethod))
			{
				// Does the attribute have a value?
				if (empty($attrSubSet[1]) === false)
				{
					$newSet[] = $attrSubSet[0] . '="' . $attrSubSet[1] . '"';
				}
				elseif ($attrSubSet[1] === "0")
				{
					// Special Case
					// Is the value 0?
					$newSet[] = $attrSubSet[0] . '="0"';
				}
				else
				{
					// Leave empty attributes alone
					$newSet[] = $attrSubSet[0] . '=""';
				}
			}
		}

		return $newSet;
	}

	/**
	 * Try to convert to plaintext
	 *
	 * @param   string  $source  The source string.
	 *
	 * @return  string  Plaintext string
	 *
	 * @since   11.1
	 */
	protected function _decode($source)
	{
		static $ttr;

		if (!is_array($ttr))
		{
			// Entity decode
			$trans_tbl = get_html_translation_table(HTML_ENTITIES);
			foreach ($trans_tbl as $k => $v)
			{
				$ttr[$v] = utf8_encode($k);
			}
		}
		$source = strtr($source, $ttr);
		// Convert decimal
		$source = preg_replace('/&#(\d+);/me', "utf8_encode(chr(\\1))", $source); // decimal notation
		// Convert hex
		$source = preg_replace('/&#x([a-f0-9]+);/mei', "utf8_encode(chr(0x\\1))", $source); // hex notation
		return $source;
	}

	/**
	 * Escape < > and " inside attribute values
	 *
	 * @param   string  $source  The source string.
	 *
	 * @return  string  Filtered string
	 *
	 * @since    11.1
	 */
	protected function _escapeAttributeValues($source)
	{
		$alreadyFiltered = '';
		$remainder = $source;
		$badChars = array('<', '"', '>');
		$escapedChars = array('&lt;', '&quot;', '&gt;');
		// Process each portion based on presence of =" and "<space>, "/>, or ">
		// See if there are any more attributes to process
		while (preg_match('#<[^>]*?=\s*?(\"|\')#s', $remainder, $matches, PREG_OFFSET_CAPTURE))
		{
			// get the portion before the attribute value
			$quotePosition = $matches[0][1];
			$nextBefore = $quotePosition + strlen($matches[0][0]);

			// Figure out if we have a single or double quote and look for the matching closing quote
			// Closing quote should be "/>, ">, "<space>, or " at the end of the string
			$quote = substr($matches[0][0], -1);
			$pregMatch = ($quote == '"') ? '#(\"\s*/\s*>|\"\s*>|\"\s+|\"$)#' : "#(\'\s*/\s*>|\'\s*>|\'\s+|\'$)#";

			// get the portion after attribute value
			if (preg_match($pregMatch, substr($remainder, $nextBefore), $matches, PREG_OFFSET_CAPTURE))
			{
				// We have a closing quote
				$nextAfter = $nextBefore + $matches[0][1];
			}
			else
			{
				// No closing quote
				$nextAfter = strlen($remainder);
			}
			// Get the actual attribute value
			$attributeValue = substr($remainder, $nextBefore, $nextAfter - $nextBefore);
			// Escape bad chars
			$attributeValue = str_replace($badChars, $escapedChars, $attributeValue);
			$attributeValue = $this->_stripCSSExpressions($attributeValue);
			$alreadyFiltered .= substr($remainder, 0, $nextBefore) . $attributeValue . $quote;
			$remainder = substr($remainder, $nextAfter + 1);
		}

		// At this point, we just have to return the $alreadyFiltered and the $remainder
		return $alreadyFiltered . $remainder;
	}

	/**
	 * Remove CSS Expressions in the form of <property>:expression(...)
	 *
	 * @param   string  $source  The source string.
	 *
	 * @return  string  Filtered string
	 *
	 * @since   11.1
	 */
	protected function _stripCSSExpressions($source)
	{
		// Strip any comments out (in the form of /*...*/)
		$test = preg_replace('#\/\*.*\*\/#U', '', $source);
		// Test for :expression
		if (!stripos($test, ':expression'))
		{
			// Not found, so we are done
			$return = $source;
		}
		else
		{
			// At this point, we have stripped out the comments and have found :expression
			// Test stripped string for :expression followed by a '('
			if (preg_match_all('#:expression\s*\(#', $test, $matches))
			{
				// If found, remove :expression
				$test = str_ireplace(':expression', '', $test);
				$return = $test;
			}
		}
		return $return;
	}
}