Source for file php5.php
Documentation is available at php5.php
/* vim: set expandtab tabstop=4 shiftwidth=4 softtabstop=4 foldmethod=marker: */ // +----------------------------------------------------------------------+ // | This library is free software; you can redistribute it and/or modify | // | it under the terms of the GNU Lesser General Public License as | // | published by the Free Software Foundation; either version 2.1 of the | // | License, or (at your option) any later version. | // | This library 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 | // | Lesser General Public License for more details. | // | You should have received a copy of the GNU Lesser General Public | // | License along with this library; if not, write to the Free Software | // | Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 | // +----------------------------------------------------------------------+ * Encode/decode Internationalized Domain Names. * The class allows to convert internationalized domain names * (see RFC 3490 for details) as they can be used with various registries worldwide * to be translated between their original (localized) form and their encoded form * as it will be used in the DNS (Domain Name System). * The class provides two public methods, encode() and decode(), which do exactly * what you would expect them to do. You are allowed to use complete domain names, * simple strings and complete email addresses as well. That means, that you might * use any of the following notations: * - xn--brse-5qa.xn--knrz-1ra.info * Unicode input might be given as either UTF-8 string, UCS-4 string or UCS-4 * array. Unicode output is available in the same formats. * You can select your preferred format via {@link set_paramter()}. * ACE input and output is always expected to be ASCII. * @author Markus Nix <mnix@docuverse.de> * @author Matthias Sommerfeld <mso@phlylabs.de> * @version $Id: IDNA.php,v 0.4.0 2005/07/30 06:06 phlylabs_de Exp $ * These Unicode codepoints are * mapped to nothing, See RFC3454 for details private static $_np_map_nothing = array ( private static $_general_prohibited = array ( * Codepints prohibited by Nameprep private static $_np_prohibit = array ( * Codepoint ranges prohibited by nameprep private static $_np_prohibit_ranges = array ( array (0x1D173 , 0x1D17A ), array (0xF0000 , 0xFFFFD ), array (0x100000 , 0x10FFFD ), * Replacement mappings (casemapping, replacement sequences, ...) private static $_np_replacemaps = array ( 0xDF => array (0x73 , 0x73 ), 0x130 => array (0x69 , 0x307 ), 0x149 => array (0x2BC , 0x6E ), 0x1F0 => array (0x6A , 0x30C ), 0x37A => array (0x20 , 0x3B9 ), 0x390 => array (0x3B9 , 0x308 , 0x301 ), 0x3B0 => array (0x3C5 , 0x308 , 0x301 ), 0x587 => array (0x565 , 0x582 ), 0x1E96 => array (0x68 , 0x331 ), 0x1E97 => array (0x74 , 0x308 ), 0x1E98 => array (0x77 , 0x30A ), 0x1E99 => array (0x79 , 0x30A ), 0x1E9A => array (0x61 , 0x2BE ), 0x1F50 => array (0x3C5 , 0x313 ), 0x1F52 => array (0x3C5 , 0x313 , 0x300 ), 0x1F54 => array (0x3C5 , 0x313 , 0x301 ), 0x1F56 => array (0x3C5 , 0x313 , 0x342 ), 0x1F80 => array (0x1F00 , 0x3B9 ), 0x1F81 => array (0x1F01 , 0x3B9 ), 0x1F82 => array (0x1F02 , 0x3B9 ), 0x1F83 => array (0x1F03 , 0x3B9 ), 0x1F84 => array (0x1F04 , 0x3B9 ), 0x1F85 => array (0x1F05 , 0x3B9 ), 0x1F86 => array (0x1F06 , 0x3B9 ), 0x1F87 => array (0x1F07 , 0x3B9 ), 0x1F88 => array (0x1F00 , 0x3B9 ), 0x1F89 => array (0x1F01 , 0x3B9 ), 0x1F8A => array (0x1F02 , 0x3B9 ), 0x1F8B => array (0x1F03 , 0x3B9 ), 0x1F8C => array (0x1F04 , 0x3B9 ), 0x1F8D => array (0x1F05 , 0x3B9 ), 0x1F8E => array (0x1F06 , 0x3B9 ), 0x1F8F => array (0x1F07 , 0x3B9 ), 0x1F90 => array (0x1F20 , 0x3B9 ), 0x1F91 => array (0x1F21 , 0x3B9 ), 0x1F92 => array (0x1F22 , 0x3B9 ), 0x1F93 => array (0x1F23 , 0x3B9 ), 0x1F94 => array (0x1F24 , 0x3B9 ), 0x1F95 => array (0x1F25 , 0x3B9 ), 0x1F96 => array (0x1F26 , 0x3B9 ), 0x1F97 => array (0x1F27 , 0x3B9 ), 0x1F98 => array (0x1F20 , 0x3B9 ), 0x1F99 => array (0x1F21 , 0x3B9 ), 0x1F9A => array (0x1F22 , 0x3B9 ), 0x1F9B => array (0x1F23 , 0x3B9 ), 0x1F9C => array (0x1F24 , 0x3B9 ), 0x1F9D => array (0x1F25 , 0x3B9 ), 0x1F9E => array (0x1F26 , 0x3B9 ), 0x1F9F => array (0x1F27 , 0x3B9 ), 0x1FA0 => array (0x1F60 , 0x3B9 ), 0x1FA1 => array (0x1F61 , 0x3B9 ), 0x1FA2 => array (0x1F62 , 0x3B9 ), 0x1FA3 => array (0x1F63 , 0x3B9 ), 0x1FA4 => array (0x1F64 , 0x3B9 ), 0x1FA5 => array (0x1F65 , 0x3B9 ), 0x1FA6 => array (0x1F66 , 0x3B9 ), 0x1FA7 => array (0x1F67 , 0x3B9 ), 0x1FA8 => array (0x1F60 , 0x3B9 ), 0x1FA9 => array (0x1F61 , 0x3B9 ), 0x1FAA => array (0x1F62 , 0x3B9 ), 0x1FAB => array (0x1F63 , 0x3B9 ), 0x1FAC => array (0x1F64 , 0x3B9 ), 0x1FAD => array (0x1F65 , 0x3B9 ), 0x1FAE => array (0x1F66 , 0x3B9 ), 0x1FAF => array (0x1F67 , 0x3B9 ), 0x1FB2 => array (0x1F70 , 0x3B9 ), 0x1FB3 => array (0x3B1 , 0x3B9 ), 0x1FB4 => array (0x3AC , 0x3B9 ), 0x1FB6 => array (0x3B1 , 0x342 ), 0x1FB7 => array (0x3B1 , 0x342 , 0x3B9 ), 0x1FBC => array (0x3B1 , 0x3B9 ), 0x1FC2 => array (0x1F74 , 0x3B9 ), 0x1FC3 => array (0x3B7 , 0x3B9 ), 0x1FC4 => array (0x3AE , 0x3B9 ), 0x1FC6 => array (0x3B7 , 0x342 ), 0x1FC7 => array (0x3B7 , 0x342 , 0x3B9 ), 0x1FCC => array (0x3B7 , 0x3B9 ), 0x1FD2 => array (0x3B9 , 0x308 , 0x300 ), 0x1FD3 => array (0x3B9 , 0x308 , 0x301 ), 0x1FD6 => array (0x3B9 , 0x342 ), 0x1FD7 => array (0x3B9 , 0x308 , 0x342 ), 0x1FE2 => array (0x3C5 , 0x308 , 0x300 ), 0x1FE3 => array (0x3C5 , 0x308 , 0x301 ), 0x1FE4 => array (0x3C1 , 0x313 ), 0x1FE6 => array (0x3C5 , 0x342 ), 0x1FE7 => array (0x3C5 , 0x308 , 0x342 ), 0x1FF2 => array (0x1F7C , 0x3B9 ), 0x1FF3 => array (0x3C9 , 0x3B9 ), 0x1FF4 => array (0x3CE , 0x3B9 ), 0x1FF6 => array (0x3C9 , 0x342 ), 0x1FF7 => array (0x3C9 , 0x342 , 0x3B9 ), 0x1FFC => array (0x3C9 , 0x3B9 ), 0x20A8 => array (0x72 , 0x73 ), 0x2103 => array (0xB0 , 0x63 ), 0x2109 => array (0xB0 , 0x66 ), 0x2116 => array (0x6E , 0x6F ), 0x2120 => array (0x73 , 0x6D ), 0x2121 => array (0x74 , 0x65 , 0x6C ), 0x2122 => array (0x74 , 0x6D ), 0x3371 => array (0x68 , 0x70 , 0x61 ), 0x3373 => array (0x61 , 0x75 ), 0x3375 => array (0x6F , 0x76 ), 0x3380 => array (0x70 , 0x61 ), 0x3381 => array (0x6E , 0x61 ), 0x3382 => array (0x3BC , 0x61 ), 0x3383 => array (0x6D , 0x61 ), 0x3384 => array (0x6B , 0x61 ), 0x3385 => array (0x6B , 0x62 ), 0x3386 => array (0x6D , 0x62 ), 0x3387 => array (0x67 , 0x62 ), 0x338A => array (0x70 , 0x66 ), 0x338B => array (0x6E , 0x66 ), 0x338C => array (0x3BC , 0x66 ), 0x3390 => array (0x68 , 0x7A ), 0x3391 => array (0x6B , 0x68 , 0x7A ), 0x3392 => array (0x6D , 0x68 , 0x7A ), 0x3393 => array (0x67 , 0x68 , 0x7A ), 0x3394 => array (0x74 , 0x68 , 0x7A ), 0x33A9 => array (0x70 , 0x61 ), 0x33AA => array (0x6B , 0x70 , 0x61 ), 0x33AB => array (0x6D , 0x70 , 0x61 ), 0x33AC => array (0x67 , 0x70 , 0x61 ), 0x33B4 => array (0x70 , 0x76 ), 0x33B5 => array (0x6E , 0x76 ), 0x33B6 => array (0x3BC , 0x76 ), 0x33B7 => array (0x6D , 0x76 ), 0x33B8 => array (0x6B , 0x76 ), 0x33B9 => array (0x6D , 0x76 ), 0x33BA => array (0x70 , 0x77 ), 0x33BB => array (0x6E , 0x77 ), 0x33BC => array (0x3BC , 0x77 ), 0x33BD => array (0x6D , 0x77 ), 0x33BE => array (0x6B , 0x77 ), 0x33BF => array (0x6D , 0x77 ), 0x33C0 => array (0x6B , 0x3C9 ), 0x33C1 => array (0x6D , 0x3C9 ), /* 0x33C2 => array(0x61, 0x2E, 0x6D, 0x2E), */ 0x33C3 => array (0x62 , 0x71 ), 0x33C6 => array (0x63 , 0x2215 , 0x6B , 0x67 ), 0x33C7 => array (0x63 , 0x6F , 0x2E ), 0x33C8 => array (0x64 , 0x62 ), 0x33C9 => array (0x67 , 0x79 ), 0x33CB => array (0x68 , 0x70 ), 0x33CD => array (0x6B , 0x6B ), 0x33CE => array (0x6B , 0x6D ), 0x33D7 => array (0x70 , 0x68 ), 0x33D9 => array (0x70 , 0x70 , 0x6D ), 0x33DA => array (0x70 , 0x72 ), 0x33DC => array (0x73 , 0x76 ), 0x33DD => array (0x77 , 0x62 ), 0xFB00 => array (0x66 , 0x66 ), 0xFB01 => array (0x66 , 0x69 ), 0xFB02 => array (0x66 , 0x6C ), 0xFB03 => array (0x66 , 0x66 , 0x69 ), 0xFB04 => array (0x66 , 0x66 , 0x6C ), 0xFB05 => array (0x73 , 0x74 ), 0xFB06 => array (0x73 , 0x74 ), 0xFB13 => array (0x574 , 0x576 ), 0xFB14 => array (0x574 , 0x565 ), 0xFB15 => array (0x574 , 0x56B ), 0xFB16 => array (0x57E , 0x576 ), 0xFB17 => array (0x574 , 0x56D ), 0x10400 => array (0x10428 ), 0x10401 => array (0x10429 ), 0x10402 => array (0x1042A ), 0x10403 => array (0x1042B ), 0x10404 => array (0x1042C ), 0x10405 => array (0x1042D ), 0x10406 => array (0x1042E ), 0x10407 => array (0x1042F ), 0x10408 => array (0x10430 ), 0x10409 => array (0x10431 ), 0x1040A => array (0x10432 ), 0x1040B => array (0x10433 ), 0x1040C => array (0x10434 ), 0x1040D => array (0x10435 ), 0x1040E => array (0x10436 ), 0x1040F => array (0x10437 ), 0x10410 => array (0x10438 ), 0x10411 => array (0x10439 ), 0x10412 => array (0x1043A ), 0x10413 => array (0x1043B ), 0x10414 => array (0x1043C ), 0x10415 => array (0x1043D ), 0x10416 => array (0x1043E ), 0x10417 => array (0x1043F ), 0x10418 => array (0x10440 ), 0x10419 => array (0x10441 ), 0x1041A => array (0x10442 ), 0x1041B => array (0x10443 ), 0x1041C => array (0x10444 ), 0x1041D => array (0x10445 ), 0x1041E => array (0x10446 ), 0x1041F => array (0x10447 ), 0x10420 => array (0x10448 ), 0x10421 => array (0x10449 ), 0x10422 => array (0x1044A ), 0x10423 => array (0x1044B ), 0x10424 => array (0x1044C ), 0x10425 => array (0x1044D ), 0x213B => array (0x66 , 0x61 , 0x78 ), 0x3250 => array (0x70 , 0x74 , 0x65 ), 0x32CC => array (0x68 , 0x67 ), 0x32CE => array (0x65 , 0x76 ), 0x32CF => array (0x6C , 0x74 , 0x64 ), 0x337A => array (0x69 , 0x75 ), 0x33DE => array (0x76 , 0x2215 , 0x6D ), 0x33DF => array (0x61 , 0x2215 , 0x6D ) * Normalization Combining Classes; Code Points not listed private static $_np_norm_combcls = array ( private $_punycode_prefix = 'xn--'; private $_invalid_ucs = 0x80000000; private $_max_ucs = 0x10FFFF; private $_initial_bias = 72; private $_initial_n = 0x80; private $_sbase = 0xAC00; private $_lbase = 0x1100; private $_vbase = 0x1161; private $_tbase = 0x11a7; * lcount * tcount * vcount private $_scount = 11172; * Default encoding for encode()'s input and decode()'s output is UTF-8; * Other possible encodings are ucs4_string and ucs4_array * See {@link setParams()} for how to select these private $_api_encoding = 'utf8'; * Overlong UTF-8 encodings are forbidden private $_allow_overlong = false; private $_strict_mode = false; $this->_slast = $this->_sbase + $this->_lcount * $this->_vcount * $this->_tcount; * Sets a new option value. Available options and values: * [utf8 - Use either UTF-8 or ISO-8859-1 as input (true for UTF-8, false * otherwise); The output is always UTF-8] * [overlong - Unicode does not allow unnecessarily long encodings of chars, * to allow this, set this parameter to true, else to false; * [strict - true: strict mode, good for registration purposes - Causes errors * on failures; false: loose mode, ideal for "wildlife" applications * by silently ignoring errors and returning the original input instead] * @param mixed $option Parameter to set (string: single parameter; array of Parameter => Value pairs) * @param string $value Value to use (if parameter 1 is a string) * @return boolean true on success, false otherwise public function setParams($option, $value = false ) $option = array ($option => $value); foreach ($option as $k => $v) { $this->_api_encoding = $v; throw new Exception ('Set Parameter: Unknown parameter '. $v. ' for option '. $k); $this->_allow_overlong = ($v) ? true : false; $this->_strict_mode = ($v) ? true : false; * Encode a given UTF-8 domain name. * @param string $decoded Domain name (UTF-8 or UCS-4) * [@param string $encoding Desired input encoding, see {@link set_parameter}] * @return string Encoded Domain name (ACE string) * @return mixed processed string public function encode($decoded, $one_time_encoding = false ) // Forcing conversion of input to UCS4 array // If one time encoding is given, use this, else the objects property switch (($one_time_encoding) ? $one_time_encoding : $this->_api_encoding) { $decoded = $this->_utf8_to_ucs4 ($decoded); $decoded = $this->_ucs4_string_to_ucs4 ($decoded); case 'ucs4_array': // No break; before this line. Catch case, but do nothing throw new Exception ('Unsupported input format'); // No input, no output, what else did you expect? if (empty ($decoded)) return ''; foreach ($decoded as $k => $v) { // Make sure to use just the plain dot // It's right, no break here // The codepoints above have to be converted to dots anyway // Stumbling across an anchoring character // Neither email addresses nor URLs allowed in strict mode if ($this->_strict_mode) { throw new Exception ('Neither email addresses nor URLs are allowed in strict mode.'); $encoded = $this->_encode (array_slice($decoded, $last_begin, (($k)- $last_begin))); $output .= $this->_ucs4_to_utf8 (array_slice($decoded, $last_begin, (($k)- $last_begin))); $output .= chr($decoded[$k]); // Catch the rest of the string $encoded = $this->_encode (array_slice($decoded, $last_begin, (($inp_len)- $last_begin))); $output .= $this->_ucs4_to_utf8 (array_slice($decoded, $last_begin, (($inp_len)- $last_begin))); if ($output = $this->_encode ($decoded)) { return $this->_ucs4_to_utf8 ($decoded); * Decode a given ACE domain name. * @param string $encoded Domain name (ACE string) * @param string $encoding Desired output encoding, see {@link set_parameter} * @return string Decoded Domain name (UTF-8 or UCS-4) public function decode($input, $one_time_encoding = false ) if ($one_time_encoding) { switch ($one_time_encoding) { throw new Exception ('Unknown encoding '. $one_time_encoding); // Make sure to drop any newline characters around // Negotiate input and try to determine, wether it is a plain string, // an email address or something like a complete URL if (strpos($input, '@')) { // Maybe it is an email address if ($this->_strict_mode) { throw new Exception ('Only simple domain name parts can be handled in strict mode'); list ($email_pref, $input) = explode('@', $input, 2 ); foreach ($arr as $k => $v) { $conv = $this->_decode ($v); if ($conv) $arr[$k] = $conv; $return = $email_pref . '@' . join('.', $arr); } elseif (preg_match('![:\./]!', $input)) { // Or a complete domain name (with or without paths / parameters) if ($this->_strict_mode) { throw new Exception ('Only simple domain name parts can be handled in strict mode'); if (isset ($parsed['host'])) { $arr = explode('.', $parsed['host']); foreach ($arr as $k => $v) { $conv = $this->_decode ($v); if ($conv) $arr[$k] = $conv; $parsed['host'] = join('.', $arr); if (isset ($parsed['scheme'])) { $parsed['scheme'] .= (strtolower($parsed['scheme']) == 'mailto') ? ':' : '://'; $return = join('', $parsed); } else { // parse_url seems to have failed, try without it foreach ($arr as $k => $v) { $conv = $this->_decode ($v); if ($conv) $arr[$k] = $conv; $return = join('.', $arr); } else { // Otherwise we consider it being a pure domain name string $return = $this->_decode ($input); // The output is UTF-8 by default, other output formats need conversion here // If one time encoding is given, use this, else the objects property switch (($one_time_encoding) ? $one_time_encoding : $this->_api_encoding) { return $this->_ucs4_to_ucs4_string ($this->_utf8_to_ucs4 ($return)); return $this->_utf8_to_ucs4 ($return); throw new Exception ('Unsupported output format'); * The actual encoding algorithm. private function _encode ($decoded) // We cannot encode a domain name containing the Punycode prefix $extract = strlen($this->_punycode_prefix); $check_pref = $this->_utf8_to_ucs4 ($this->_punycode_prefix); if ($check_pref == $check_deco) { throw new Exception ('This is already a punycode string'); // We will not try to encode strings consisting of basic code points only foreach ($decoded as $k => $v) { if ($this->_strict_mode) { throw new Exception ('The given string does not contain encodable chars'); $decoded = $this->_nameprep ($decoded); // hmm, serious - rethrow $deco_len = count($decoded); // How many chars have been consumed // Start with the prefix; copy it to output $encoded = $this->_punycode_prefix; // Copy all basic code points to output for ($i = 0; $i < $deco_len; ++ $i) { $encoded .= chr($decoded[$i]); // All codepoints were basic ones if ($codecount == $deco_len) { // Start with the prefix; copy it to output $encoded = $this->_punycode_prefix . $encoded; // If we have basic code points in output, add an hyphen to the end // Now find and encode all non-basic code points $cur_code = $this->_initial_n; $bias = $this->_initial_bias; while ($codecount < $deco_len) { // Find the smallest code point >= the current code point and // remember the last ouccrence of it in the input for ($i = 0 , $next_code = $this->_max_ucs; $i < $deco_len; $i++ ) { if ($decoded[$i] >= $cur_code && $decoded[$i] <= $next_code) { $next_code = $decoded[$i]; $delta += ($next_code - $cur_code) * ($codecount + 1 ); // Scan input again and encode all characters whose code point is $cur_code for ($i = 0; $i < $deco_len; $i++ ) { if ($decoded[$i] < $cur_code) { } else if ($decoded[$i] == $cur_code) { for ($q = $delta, $k = $this->_base; 1; $k += $this->_base) { (($k >= $bias + $this->_tmax)? $this->_tmax : $k - $bias); $encoded .= $this->_encodeDigit (ceil($t + (($q - $t) % ($this->_base - $t)))); $q = ($q - $t) / ($this->_base - $t); $encoded .= $this->_encodeDigit ($q); $bias = $this->_adapt ($delta, $codecount + 1 , $is_first); * The actual decoding algorithm. private function _decode ($encoded) // We do need to find the Punycode prefix // If nothing left after removing the prefix, it is hopeless // Find last occurence of the delimiter $delim_pos = strrpos($encoded, '-'); if ($delim_pos > strlen($this->_punycode_prefix)) { for ($k = strlen($this->_punycode_prefix); $k < $delim_pos; ++ $k) { $decoded[] = ord($encoded{$k}); $deco_len = count($decoded); // Wandering through the strings; init $bias = $this->_initial_bias; $char = $this->_initial_n; for ($enco_idx = ($delim_pos)? ($delim_pos + 1 ) : 0; $enco_idx < $enco_len; ++ $deco_len) { for ($old_idx = $idx, $w = 1 , $k = $this->_base; 1 ; $k += $this->_base) { $digit = $this->_decodeDigit ($encoded{$enco_idx++ }); (($k >= $bias + $this->_tmax)? $this->_tmax : ($k - $bias)); $w = (int) ($w * ($this->_base - $t)); $bias = $this->_adapt ($idx - $old_idx, $deco_len + 1 , $is_first); $char += (int) ($idx / ($deco_len + 1 )); // Make room for the decoded char for ($i = $deco_len; $i > $idx; $i-- ) { $decoded[$i] = $decoded[($i - 1 )]; $decoded[$idx++ ] = $char; return $this->_ucs4_to_utf8 ($decoded); * Adapt the bias according to the current code point and position. private function _adapt ($delta, $npoints, $is_first) $delta = (int) ($is_first ? ($delta / $this->_damp) : ($delta / 2 )); $delta += (int) ($delta / $npoints); for ($k = 0; $delta > (($this->_base - $this->_tmin) * $this->_tmax) / 2; $k += $this->_base) { $delta = (int) ($delta / ($this->_base - $this->_tmin)); return (int) ($k + ($this->_base - $this->_tmin + 1 ) * $delta / ($delta + $this->_skew)); * Encoding a certain digit. private function _encodeDigit ($d) return chr($d + 22 + 75 * ($d < 26 )); * Decode a certain digit. private function _decodeDigit ($cp) return ($cp - 48 < 10 )? $cp - 22 : (($cp - 65 < 26 )? $cp - 65 : (($cp - 97 < 26 )? $cp - 97 : $this->_base)); * Do Nameprep according to RFC3491 and RFC3454. * @param array $input Unicode Characters * @return string Unicode Characters, Nameprep'd private function _nameprep ($input) // Walking through the input array, performing the required steps on each of // the input chars and putting the result into the output array // While mapping required chars we apply the cannonical ordering // Map to nothing == skip that code point if (in_array($v, self ::$_np_map_nothing)) { // Try to find prohibited input if (in_array ($v, self ::$_np_prohibit) || in_array ($v, self ::$_general_prohibited)) { throw new Exception ('NAMEPREP: Prohibited input U+' . sprintf('%08X', $v)); foreach (self ::$_np_prohibit_ranges as $range) { if ($range[0 ] <= $v && $v <= $range[1 ]) { throw new Exception ('NAMEPREP: Prohibited input U+' . sprintf('%08X', $v)); // Hangul syllable decomposition if (0xAC00 <= $v && $v <= 0xD7AF ) { foreach ($this->_hangulDecompose ($v) as $out) { } else if (isset (self ::$_np_replacemaps[$v])) { // There's a decomposition mapping for that code point foreach ($this->_applyCannonicalOrdering (self ::$_np_replacemaps[$v]) as $out) { $out_len = count ($output); for ($i = 0; $i < $out_len; ++ $i) { $class = $this->_getCombiningClass ($output[$i]); if ((!$last_class || $last_class != $class) && $class) { $seq_len = $i - $last_starter; $out = $this->_combine (array_slice($output, $last_starter, $seq_len)); // On match: Replace the last starter with the composed character and remove // the now redundant non-starter(s) $output[$last_starter] = $out; if (count($out) != $seq_len) { for ($j = $i + 1; $j < $out_len; ++ $j) { $output[$j - 1 ] = $output[$j]; unset ($output[$out_len]); // Rewind the for loop by one, since there can be more possible compositions $last_class = ($i == $last_starter)? 0 : $this->_getCombiningClass ($output[$i - 1 ]); // The current class is 0 * Decomposes a Hangul syllable * (see http://www.unicode.org/unicode/reports/tr15/#Hangul). * @param integer $char 32bit UCS4 code point * @return array Either Hangul Syllable decomposed or original 32bit * value as one value array private function _hangulDecompose ($char) $sindex = $char - $this->_sbase; if ($sindex < 0 || $sindex >= $this->_scount) { $T = $this->_tbase + $sindex % $this->_tcount; $result[] = (int) ($this->_lbase + $sindex / $this->_ncount); $result[] = (int) ($this->_vbase + ($sindex % $this->_ncount) / $this->_tcount); if ($T != $this->_tbase) { * Ccomposes a Hangul syllable * (see http://www.unicode.org/unicode/reports/tr15/#Hangul). * @param array $input Decomposed UCS4 sequence * @return array UCS4 sequence with syllables composed private function _hangulCompose ($input) $inp_len = count($input); $result[] = $last; // copy first char from input to output for ($i = 1; $i < $inp_len; ++ $i) { // Find out, wether two current characters from L and V $lindex = $last - $this->_lbase; if (0 <= $lindex && $lindex < $this->_lcount) { $vindex = $char - $this->_vbase; if (0 <= $vindex && $vindex < $this->_vcount) { // create syllable of form LV $last = ($this->_sbase + ($lindex * $this->_vcount + $vindex) * $this->_tcount); $out_off = count($result) - 1; $result[$out_off] = $last; // reset last // Find out, wether two current characters are LV and T $sindex = $last - $this->_sbase; if (0 <= $sindex && $sindex < $this->_scount && ($sindex % $this->_tcount) == 0 ) { $tindex = $char - $this->_tbase; if (0 <= $tindex && $tindex <= $this->_tcount) { // create syllable of form LVT $out_off = count($result) - 1; $result[$out_off] = $last; // reset last // if neither case was true, just add the character * Returns the combining class of a certain wide char. * @param integer $char Wide char to check (32bit integer) * @return integer Combining class if found, else 0 private function _getCombiningClass ($char) return isset (self ::$_np_norm_combcls[$char])? self ::$_np_norm_combcls[$char] : 0; * Apllies the cannonical ordering of a decomposed UCS4 sequence. * @param array $input Decomposed UCS4 sequence * @return array Ordered USC4 sequence private function _applyCannonicalOrdering ($input) $last = $this->_getCombiningClass ($input[0 ]); for ($i = 0; $i < $size - 1; ++ $i) { $next = $this->_getCombiningClass ($input[$i + 1 ]); if ($next != 0 && $last > $next) { // Move item leftward until it fits for ($j = $i + 1; $j > 0; -- $j) { if ($this->_getCombiningClass ($input[$j - 1 ]) <= $next) { $input[$j] = $input[$j - 1 ]; // Reentering the loop looking at the old character again * Do composition of a sequence of starter and non-starter. * @param array $input UCS4 Decomposed sequence * @return array Ordered USC4 sequence private function _combine ($input) $inp_len = count($input); // Is it a Hangul syllable? $hangul = $this->_hangulCompose ($input); // This place is probably wrong if (count($hangul) != $inp_len) { foreach (self ::$_np_replacemaps as $np_src => $np_target) { if ($np_target[0 ] != $input[0 ]) { if (count ($np_target) != $inp_len) { foreach ($input as $k2 => $v2) { if ($v2 == $np_target[$k2]) { * This converts an UTF-8 encoded string to its UCS-4 (array) representation * By talking about UCS-4 we mean arrays of 32bit integers representing * each of the "chars". This is due to PHP not being able to handle strings with * bit depth different from 8. This applies to the reverse method _ucs4_to_utf8(), too. * The following UTF-8 encodings are supported: * bytes bits representation * 3 16 1110xxxx 10xxxxxx 10xxxxxx * 4 21 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx * 5 26 111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx * 6 31 1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx * Each x represents a bit that can be used to store character data. private function _utf8_to_ucs4 ($input) for ($k = 0; $k < $inp_len; ++ $k) { $v = ord($input{$k}); // Extract byte from input string if ($v < 128 ) { // We found an ASCII char - put into stirng as is throw new Exception ('Conversion from UTF-8 to UCS-4 failed: malformed input at byte '. $k); if ('next' == $mode) { // Try to find the next start byte; determine the width of the Unicode char if ($v >> 5 == 6 ) { // &110xxxxx 10xxxxx $next_byte = 0; // Tells, how many times subsequent bitmasks must rotate 6bits to the left } elseif ($v >> 4 == 14 ) { // &1110xxxx 10xxxxxx 10xxxxxx } elseif ($v >> 3 == 30 ) { // &11110xxx 10xxxxxx 10xxxxxx 10xxxxxx } elseif ($v >> 2 == 62 ) { // &111110xx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx } elseif ($v >> 1 == 126 ) { // &1111110x 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx 10xxxxxx throw new Exception ('This might be UTF-8, but I don\'t understand it at byte '. $k); $output[$out_len] = (int) $v; if (!$this->_allow_overlong && $test == 'range') { if (($v < 0xA0 && $start_byte == 0xE0 ) || ($v < 0x90 && $start_byte == 0xF0 ) || ($v > 0x8F && $start_byte == 0xF4 )) { throw new Exception ('Bogus UTF-8 character detected (out of legal range) at byte '. $k); if ($v >> 6 == 2 ) { // Bit mask must be 10xxxxxx $v = ($v - 128 ) << ($next_byte * 6 ); $output[($out_len - 1 )] += $v; throw new Exception ('Conversion from UTF-8 to UCS-4 failed: malformed input at byte '. $k); * Convert UCS-4 array into UTF-8 string. private function _ucs4_to_utf8 ($input) // 7bit are transferred literally } else if ($v < 1 << 11 ) { $output .= chr(192 + ($v >> 6 )) } else if ($v < 1 << 16 ) { $output .= chr(224 + ($v >> 12 )) . chr(128 + (($v >> 6 ) & 63 )) } else if ($v < 1 << 21 ) { $output .= chr(240 + ($v >> 18 )) . chr(128 + (($v >> 12 ) & 63 )) . chr(128 + (($v >> 6 ) & 63 )) } else if ($v < 1 << 26 ) { $output .= chr(248 + ($v >> 24 )) . chr(128 + (($v >> 18 ) & 63 )) . chr(128 + (($v >> 12 ) & 63 )) . chr(128 + (($v >> 6 ) & 63 )) } else if ($v < 1 << 31 ) { $output .= chr(252 + ($v >> 30 )) . chr(128 + (($v >> 24 ) & 63 )) . chr(128 + (($v >> 18 ) & 63 )) . chr(128 + (($v >> 12 ) & 63 )) . chr(128 + (($v >> 6 ) & 63 )) throw new Exception ('Conversion from UCS-4 to UTF-8 failed: malformed input at byte ' . $k); * Convert UCS-4 array into UCS-4 string private function _ucs4_to_ucs4_string ($input) // Take array values and split output to 4 bytes per value // The bit mask is 255, which reads &11111111 $output .= ($v & (255 << 24 ) >> 24 ) . ($v & (255 << 16 ) >> 16 ) . ($v & (255 << 8 ) >> 8 ) . ($v & 255 ); * Convert UCS-4 strin into UCS-4 garray private function _ucs4_string_to_ucs4 ($input) // Input length must be dividable by 4 throw new Exception ('Input UCS4 string is broken'); // Empty input - return empty output if (!$inp_len) return $output; for ($i = 0 , $out_len = -1; $i < $inp_len; ++ $i) { // Increment output position every 4 input bytes $output[$out_len] += ord($input{$i}) << (8 * (3 - ($i % 4 ) ) ); * Echo hex representation of UCS4 sequence. * @param array $input UCS4 sequence * @param boolean $include_bit Include bitmask in output private static function _showHex ($input, $include_bit = false ) foreach ($input as $k => $v) { echo '[', $k, '] => ', sprintf('%X', $v); echo ' (', Net_IDNA::_showBitmask ($v), ')'; * Gives you a bit representation of given Byte (8 bits), Word (16 bits) or DWord (32 bits) * Output width is automagically determined private static function _showBitmask ($octet) if ($octet >= (1 << 16 )) { } else if ($octet >= (1 << 8 )) { for ($i = $w; $i > -1; $i-- ) { $return .= ($octet & (1 << $i))? 1 : '0';
Documentation generated on Mon, 11 Mar 2019 14:16:07 -0400 by phpDocumentor 1.4.4. PEAR Logo Copyright © PHP Group 2004.
|
