Index: branches/2.6.x/wb/framework/functions.php =================================================================== --- branches/2.6.x/wb/framework/functions.php (revision 453) +++ branches/2.6.x/wb/framework/functions.php (revision 454) @@ -341,7 +341,7 @@ // Function as replecement for php's htmlspecialchars() function my_htmlspecialchars($string) { $string = umlauts_to_entities($string); - $string = entities_to_umlauts($string); + $string = entities_to_umlauts($string, DEFAULT_CHARSET, 1); return($string); } @@ -363,6 +363,138 @@ return $converted_string; } +// Function as wrapper for mb_convert_encoding +function mb_convert_encoding_wrapper($string, $charset_out, $charset_in) { + if (function_exists('mb_convert_encoding')) { + $string=mb_convert_encoding($string, $charset_out, $charset_in); + } else { + if ($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') { $mysqlcharset_from = 'latin1'; } //? BROKEN + elseif ($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') { $mysqlcharset_from = 'latin1'; } //? BROKEN + elseif ($charset_in == 'ISO-2022-KR') { $mysqlcharset_from = 'latin1'; } //? BROKEN + elseif ($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') { $mysqlcharset_to = 'latin1'; } //? BROKEN + elseif ($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') { $mysqlcharset_to = 'latin1'; } //? BROKEN + elseif ($charset_out == 'ISO-2022-KR') { $mysqlcharset_to = 'latin1'; } //? BROKEN + elseif ($charset_out == 'GB2312') { $mysqlcharset_to = 'gb2312'; } + elseif ($charset_out == 'ISO-8859-11') { $mysqlcharset_to = 'tis620'; } + elseif ($charset_out == 'UTF-8') { $mysqlcharset_to = 'utf8'; } + else { $mysqlcharset_to = 'latin1'; } + + if ($charset_in == 'HTML-ENTITIES') { $mysqlcharset_from = 'html'; } // special-case + if ($charset_out == 'HTML-ENTITIES') { $mysqlcharset_to = 'html'; } // special-case + + // use mysql to convert the string + if ($mysqlcharset_from!="html" && $mysqlcharset_to!="html" && $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" ); + } + } + // do the utf8->htmlentities or htmlentities->utf8 translation + if (($mysqlcharset_from=='html' && $mysqlcharset_to=='utf8') || ($mysqlcharset_from=='utf8' && $mysqlcharset_to=='html')) { + if ($mysqlcharset_from == 'html') { + $named_to_numbered_entities=array('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'); + $string = strtr($string, $named_to_numbered_entities); + $string = preg_replace("/&#([0-9]+);/e", "code_to_utf8($1)", $string); + } + elseif ($mysqlcharset_to == 'html') { + $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("<"=>"<", ">"=>">", "&"=>"&", """=>"\"", "''"=>"\'"); + $string = strtr($string, $string_htmlspecialchars_decode); + $numbered_to_named_entities=array('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'); + $string = strtr($string, $numbered_to_named_entities); + } + } + } + return($string); +} +// support-function for mb_convert_encoding_wrapper() +function uniord($c) { + $ud = 0; + if (ord($c{0}) >= 0 && ord($c{0}) <= 127) $ud = ord($c{0}); + if (ord($c{0}) >= 192 && ord($c{0}) <= 223) $ud = (ord($c{0})-192)*64 + (ord($c{1})-128); + if (ord($c{0}) >= 224 && ord($c{0}) <= 239) $ud = (ord($c{0})-224)*4096 + (ord($c{1})-128)*64 + (ord($c{2})-128); + if (ord($c{0}) >= 240 && ord($c{0}) <= 247) $ud = (ord($c{0})-240)*262144 + (ord($c{1})-128)*4096 + (ord($c{2})-128)*64 + (ord($c{3})-128); + if (ord($c{0}) >= 248 && ord($c{0}) <= 251) $ud = (ord($c{0})-248)*16777216 + (ord($c{1})-128)*262144 + (ord($c{2})-128)*4096 + (ord($c{3})-128)*64 + (ord($c{4})-128); + if (ord($c{0}) >= 252 && ord($c{0}) <= 253) $ud = (ord($c{0})-252)*1073741824 + (ord($c{1})-128)*16777216 + (ord($c{2})-128)*262144 + (ord($c{3})-128)*4096 + (ord($c{4})-128)*64 + (ord($c{5})-128); + if (ord($c{0}) >= 254 && ord($c{0}) <= 255) $ud = false; // error + return $ud; +} +// support-function for mb_convert_encoding_wrapper() +function code_to_utf8($num) { + if ($num <= 0x7F) { + return chr($num); + } elseif ($num <= 0x7FF) { + return chr(($num >> 6) + 192) . chr(($num & 63) + 128); + } elseif ($num <= 0xFFFF) { + return chr(($num >> 12) + 224) . chr((($num >> 6) & 63) + 128) . chr(($num & 63) + 128); + } elseif ($num <= 0x1FFFFF) { + return chr(($num >> 18) + 240) . chr((($num >> 12) & 63) + 128) . chr((($num >> 6) & 63) + 128) . chr(($num & 63) + 128); + } + return " "; +} + // Function to convert a string from mixed html-entities/umlauts to pure utf-8-umlauts function string_to_utf8($string, $charset=DEFAULT_CHARSET) { $charset = strtoupper($charset); @@ -373,11 +505,11 @@ $string=my_mysql_iconv($string, 'gb2312', 'utf8'); } elseif ($charset == "ISO-8859-11") { $string=my_mysql_iconv($string, 'tis620', 'utf8'); - } else { - $string=mb_convert_encoding($string, 'UTF-8', $charset); + } elseif ($charset != "UTF-8") { + $string=mb_convert_encoding_wrapper($string, 'UTF-8', $charset); } - $string=mb_convert_encoding($string, 'HTML-ENTITIES', 'UTF-8'); - $string=mb_convert_encoding($string, 'UTF-8', 'HTML-ENTITIES'); + $string=mb_convert_encoding_wrapper($string, 'HTML-ENTITIES', 'UTF-8'); + $string=mb_convert_encoding_wrapper($string, 'UTF-8', 'HTML-ENTITIES'); return($string); } @@ -394,7 +526,7 @@ } elseif ($charset_out == "ISO-8859-11") { $string=my_mysql_iconv($string, 'utf8', 'tis620'); } else { - $string=mb_convert_encoding($string, $charset_out, 'UTF-8'); + $string=mb_convert_encoding_wrapper($string, $charset_out, 'UTF-8'); } } if($convert_htmlspecialchars == 1) { @@ -413,7 +545,7 @@ if($convert_htmlspecialchars == 1) { $string=htmlspecialchars($string,ENT_QUOTES); } - $string=mb_convert_encoding($string,'HTML-ENTITIES','UTF-8'); + $string=mb_convert_encoding_wrapper($string,'HTML-ENTITIES','UTF-8'); return($string); }