WB 2.08.x

14-May-2007 Matthias Gallas

+	Added support for ISO-8859-6.(Thanks to thorn)

		$pos = strpos($foo, '<');

		if ($pos === false) { // "===" means identicaly

			$foo = preg_replace('/('.$string.')/i', '<span class="highlight">$1</span>',$foo);

		}

// converts $charset_in to $charset_out or 

// UTF-8 to HTML-ENTITIES or HTML-ENTITIES to UTF-8

	// try mb_convert_encoding(). This can handle to or from HTML-ENTITIES, too

		// there's no GB2312 or ISO-8859-11 encoding in php's mb_* functions

		if ($charset_in=='ISO-8859-11' || $charset_in=='GB2312') {

			if (function_exists('iconv')) {

				$string = iconv($charset_in, 'UTF-8', $string);

			else {

				if ($charset_in == 'GB2312') {

					$string=my_mysql_iconv($string, 'gb2312', 'utf8');

				} else {

					$string=my_mysql_iconv($string, 'tis620', 'utf8');

				}

			}

			$charset_in='UTF-8';

			if ($charset_out == 'UTF-8') {

				return $string;

			}

		if ($charset_out=='ISO-8859-11' || $charset_out=='GB2312') {

			$string=mb_convert_encoding($string, 'UTF-8', $charset_in);

			if (function_exists('iconv')) {

				$string = iconv('UTF-8', $charset_out, $string);

			else {

				if ($charset_out == 'GB2312') {

					$string=my_mysql_iconv($string, 'utf8', 'gb2312');

				} else {

					$string=my_mysql_iconv($string, 'utf8', 'tis620');

		} else {

			$string=mb_convert_encoding($string, $charset_out, $charset_in);

		return $string;

	// try iconv(). This can't handle to or from HTML-ENTITIES.

	if (function_exists('iconv') && $charset_out!='HTML-ENTITIES' && $charset_in!='HTML-ENTITIES' ) {

		$string = iconv($charset_in, $charset_out, $string);

		return $string;

	}

	// do the UTF-8->HTML-ENTITIES or HTML-ENTITIES->UTF-8 translation

	if (($charset_in=='HTML-ENTITIES' && $charset_out=='UTF-8') || ($charset_in=='UTF-8' && $charset_out=='HTML-ENTITIES')) {

		$named_to_numbered_entities=array(

			' '=>' ','¡'=>'¡','¢'=>'¢','£'=>'£','¤'=>'¤',

			'¥'=>'¥','¦'=>'¦','§'=>'§','¨'=>'¨','ª'=>'ª',

			'«'=>'«','¬'=>'¬','­'=>'­','®'=>'®','¯'=>'¯',

			'°'=>'°','±'=>'±','²'=>'²','³'=>'³','´'=>'´',

			'µ'=>'µ','¶'=>'¶','·'=>'·','¸'=>'¸','¹'=>'¹',

			'º'=>'º','»'=>'»','¼'=>'¼','½'=>'½','¾'=>'¾',

			'¿'=>'¿','÷'=>'÷','∅'=>'∅','€'=>'€',

			'Á'=>'Á','á'=>'á','Â'=>'Â',

			'â'=>'â','Æ'=>'Æ','æ'=>'æ','À'=>'À','à'=>'à',

			'Å'=>'Å','å'=>'å','Ã'=>'Ã','ã'=>'ã','Ä'=>'Ä',

			'ä'=>'ä','Ç'=>'Ç','ç'=>'ç','É'=>'É','é'=>'é',

			'Ê'=>'Ê','ê'=>'ê','È'=>'È','è'=>'è','Ë'=>'Ë',

			'ë'=>'ë','Í'=>'Í','í'=>'í','Î'=>'Î','î'=>'î',

			'Ì'=>'Ì','ì'=>'ì','Ï'=>'Ï','ï'=>'ï','Ñ'=>'Ñ',

			'ñ'=>'ñ','Ó'=>'Ó','ó'=>'ó','Ô'=>'Ô','ô'=>'ô',

			'Œ'=>'Œ','œ'=>'œ','Ò'=>'Ò','ò'=>'ò','Õ'=>'Õ',

			'õ'=>'õ','Ö'=>'Ö','ö'=>'ö','Š'=>'Š','š'=>'š',

			'ß'=>'ß','Ú'=>'Ú','ú'=>'ú','Û'=>'Û','û'=>'û',

			'Ù'=>'Ù','ù'=>'ù','Ü'=>'Ü','ü'=>'ü','Ý'=>'Ý',

			'ý'=>'ý','Ÿ'=>'Ÿ','ÿ'=>'ÿ','©'=>'©','®'=>'®',

			'Ð'=>'Ð','×'=>'×','Ø'=>'Ø','Þ'=>'Þ','ð'=>'ð',

			'ø'=>'ø','þ'=>'þ');

		$numbered_to_named_entities=array('Á'=>'Á','á'=>'á','Â'=>'Â',

			' '=>' ','¡'=>'¡','¢'=>'¢','£'=>'£','¤'=>'¤',

			'¥'=>'¥','¦'=>'¦','§'=>'§','¨'=>'¨','ª'=>'ª',

			'«'=>'«','¬'=>'¬','­'=>'­','®'=>'®','¯'=>'¯',

			'°'=>'°','±'=>'±','²'=>'²','³'=>'³','´'=>'´',

			'µ'=>'µ','¶'=>'¶','·'=>'·','¸'=>'¸','¹'=>'¹',

			'º'=>'º','»'=>'»','¼'=>'¼','½'=>'½','¾'=>'¾',

			'¿'=>'¿','÷'=>'÷','∅'=>'∅','€'=>'€',

			'â'=>'â','Æ'=>'Æ','æ'=>'æ','À'=>'À','à'=>'à',

			'Å'=>'Å','å'=>'å','Ã'=>'Ã','ã'=>'ã','Ä'=>'Ä',

			'ä'=>'ä','Ç'=>'Ç','ç'=>'ç','É'=>'É','é'=>'é',

			'Ê'=>'Ê','ê'=>'ê','È'=>'È','è'=>'è','Ë'=>'Ë',

			'ë'=>'ë','Í'=>'Í','í'=>'í','Î'=>'Î','î'=>'î',

			'Ì'=>'Ì','ì'=>'ì','Ï'=>'Ï','ï'=>'ï','Ñ'=>'Ñ',

			'ñ'=>'ñ','Ó'=>'Ó','ó'=>'ó','Ô'=>'Ô','ô'=>'ô',

			'Œ'=>'Œ','œ'=>'œ','Ò'=>'Ò','ò'=>'ò','Õ'=>'Õ',

			'õ'=>'õ','Ö'=>'Ö','ö'=>'ö','Š'=>'Š','š'=>'š',

			'ß'=>'ß','Ú'=>'Ú','ú'=>'ú','Û'=>'Û','û'=>'û',

			'Ù'=>'Ù','ù'=>'ù','Ü'=>'Ü','ü'=>'ü','Ý'=>'Ý',

			'ý'=>'ý','Ÿ'=>'Ÿ','ÿ'=>'ÿ','©'=>'©','®'=>'®',

			'Ð'=>'Ð','×'=>'×','Ø'=>'Ø','Þ'=>'Þ','ð'=>'ð',

			'ø'=>'ø','þ'=>'þ');

		if ($charset_in == 'HTML-ENTITIES') {

			$string = strtr($string, $named_to_numbered_entities);

			$string = preg_replace("/&#([0-9]+);/e", "code_to_utf8($1)", $string);

		}

		elseif ($charset_out == 'HTML-ENTITIES') {

			$string = preg_replace("/&#([0-9]+);/e", "code_to_utf8($1)", $string);

			$char = "";

			while (strlen($string) > 0) {

				preg_match("/^(.)(.*)$/su", $string, $match);

				if (strlen($match[1]) > 1) {

					$char .= "&#".uniord($match[1]).";";

				} else $char .= $match[1];

				$string = $match[2];

			}

			$string = $char;

			$string_htmlspecialchars_decode=array("&lt;"=>"<", "&gt;"=>">", "&amp;"=>"&", "&quot;"=>"\"", "&#039;"=>"\'");

			$string = strtr($string, $string_htmlspecialchars_decode);

			$string = strtr($string, $numbered_to_named_entities);

		}

		return $string;

	}

	// mb_convert_encoding() and iconv() aren't available, so use my_mysql_iconv()

	if ($charset_in == 'ISO-8859-1') { $mysqlcharset_from = 'latin1'; }

	elseif ($charset_in == 'ISO-8859-2') { $mysqlcharset_from = 'latin2'; }

	elseif ($charset_in == 'ISO-8859-3') { $mysqlcharset_from = 'latin1'; }

	elseif ($charset_in == 'ISO-8859-4') { $mysqlcharset_from = 'latin7'; }

	elseif ($charset_in == 'ISO-8859-5') { $string = convert_cyr_string ($string, "iso8859-5", "windows-1251" ); $mysqlcharset_from = 'cp1251'; }

	elseif ($charset_in == 'ISO-8859-6') { $mysqlcharset_from = ''; } //?

	elseif ($charset_in == 'ISO-8859-7') { $mysqlcharset_from = 'greek'; }

	elseif ($charset_in == 'ISO-8859-8') { $mysqlcharset_from = 'hebrew'; }

	elseif ($charset_in == 'ISO-8859-9') { $mysqlcharset_from = 'latin5'; }

	elseif ($charset_in == 'ISO-8859-10') { $mysqlcharset_from = 'latin1'; }

	elseif ($charset_in == 'BIG5') { $mysqlcharset_from = 'big5'; }

	elseif ($charset_in == 'ISO-2022-JP') { $mysqlcharset_from = ''; } //?

	elseif ($charset_in == 'ISO-2022-KR') { $mysqlcharset_from = ''; } //?

	elseif ($charset_in == 'GB2312') { $mysqlcharset_from = 'gb2312'; }

	elseif ($charset_in == 'ISO-8859-11') { $mysqlcharset_from = 'tis620'; }

	elseif ($charset_in == 'UTF-8') { $mysqlcharset_from = 'utf8'; }

	else { $mysqlcharset_from = 'latin1'; }

	if ($charset_out == 'ISO-8859-1') { $mysqlcharset_to = 'latin1'; }

	elseif ($charset_out == 'ISO-8859-2') { $mysqlcharset_to = 'latin2'; }

	elseif ($charset_out == 'ISO-8859-3') { $mysqlcharset_to = 'latin1'; }

	elseif ($charset_out == 'ISO-8859-4') { $mysqlcharset_to = 'latin7'; }

	elseif ($charset_out == 'ISO-8859-5') { $mysqlcharset_to = 'cp1251'; } // use convert_cyr_string afterwards

	elseif ($charset_out == 'ISO-8859-6') { $mysqlcharset_to = ''; } //?

	elseif ($charset_out == 'ISO-8859-7') { $mysqlcharset_to = 'greek'; }

	elseif ($charset_out == 'ISO-8859-8') { $mysqlcharset_to = 'hebrew'; }

	elseif ($charset_out == 'ISO-8859-9') { $mysqlcharset_to = 'latin5'; }

	elseif ($charset_out == 'ISO-8859-10') { $mysqlcharset_to = 'latin1'; }

	elseif ($charset_out == 'BIG5') { $mysqlcharset_to = 'big5'; }

	elseif ($charset_out == 'ISO-2022-JP') { $mysqlcharset_to = ''; } //?

	elseif ($charset_out == 'ISO-2022-KR') { $mysqlcharset_to = ''; } //?

	elseif ($charset_out == 'GB2312') { $mysqlcharset_to = 'gb2312'; }

	elseif ($charset_out == 'ISO-8859-11') { $mysqlcharset_to = 'tis620'; }

	elseif ($charset_out == 'UTF-8') { $mysqlcharset_to = 'utf8'; }

	else { $mysqlcharset_to = 'latin1'; }

	if ($mysqlcharset_from!="" && $mysqlcharset_to!="" && $mysqlcharset_from!=$mysqlcharset_to) {

		$string=my_mysql_iconv($string, $mysqlcharset_from, $mysqlcharset_to);

		if ($mysqlcharset_to == 'cp1251') { 

			$string = convert_cyr_string ($string, "windows-1251", "iso-8859-5" );

		}

		return($string);

	}

	// $string is unchanged. This will happen if we have to deal with ISO-8859-6 or ISO-2022-JP or -KR

	// and mbstring _and_ iconv aren't available.

	return $string;

	if (!is_UTF8($string)) {

	} else {

	// check if we really get UTF-8. We don't get UTF-8 if charset is ISO-8859-11 or GB2312 and mb_string AND iconv aren't available.

	if (is_UTF8($string)) {

		$string=mb_convert_encoding_wrapper($string, 'HTML-ENTITIES', 'UTF-8');

		$string=mb_convert_encoding_wrapper($string, 'UTF-8', 'HTML-ENTITIES');

	} else {

	}

// function to check if a string is UTF-8

function is_UTF8 ($string) {

	return preg_match('%^(?:[\x09\x0A\x0D\x20-\x7E]|[\xC2-\xDF][\x80-\xBF]|\xE0[\xA0-\xBF][\x80-\xBF]|[\xE1-\xEC\xEE\xEF][\x80-\xBF]{2}|\xED[\x80-\x9F][\x80-\xBF]|\xF0[\x90-\xBF][\x80-\xBF]{2}|[\xF1-\xF3][\x80-\xBF]{3}|\xF4[\x80-\x8F][\x80-\xBF]{2})*$%xs', $string);

}

	if ($charset_out == '') { $charset_out = 'ISO-8859-1'; }

	if($charset_out!='UTF-8' && is_UTF8($string)) {

		$string=mb_convert_encoding_wrapper($string, $charset_out, 'UTF-8');

	}

	if ($charset_in == "") { $charset_in = 'ISO-8859-1'; }

	if (is_UTF8($string)) {

		$string=mb_convert_encoding_wrapper($string,'HTML-ENTITIES','UTF-8');

	}

// and numbered-entities into hex

	$string = preg_replace('/&#([0-9]+);/e', "dechex('$1')",  $string);