utf8/transliterate_to_ascii.php

   1 <?php defined('SYSPATH') or die('No direct script access.');
   2 /**
   3  * UTF8::transliterate_to_ascii
   4  *
   5  * @package    Kohana
   6  * @author     Kohana Team
   7  * @copyright  (c) 2007-2011 Kohana Team
   8  * @copyright  (c) 2005 Harry Fuecks
   9  * @license    http://www.gnu.org/licenses/old-licenses/lgpl-2.1.txt
  10  */
  11 function _transliterate_to_ascii($str, $case = 0)
  12 {
  13         static $utf8_lower_accents = NULL;
  14         static $utf8_upper_accents = NULL;
  15
  16         if ($case <= 0)
  17         {
  18                 if ($utf8_lower_accents === NULL)
  19                 {
  20                         $utf8_lower_accents = array(
  21                                 'à' => 'a',  'ô' => 'o',  'ď' => 'd',  'ḟ' => 'f',  'ë' => 'e',  'š' => 's',  'ơ' => 'o',
  22                                 'ß' => 'ss', 'ă' => 'a',  'ř' => 'r',  'ț' => 't',  'ň' => 'n',  'ā' => 'a',  'ķ' => 'k',
  23                                 'ŝ' => 's',  'ỳ' => 'y',  'ņ' => 'n',  'ĺ' => 'l',  'ħ' => 'h',  'ṗ' => 'p',  'ó' => 'o',
  24                                 'ú' => 'u',  'ě' => 'e',  'é' => 'e',  'ç' => 'c',  'ẁ' => 'w',  'ċ' => 'c',  'õ' => 'o',
  25                                 'ṡ' => 's',  'ø' => 'o',  'ģ' => 'g',  'ŧ' => 't',  'ș' => 's',  'ė' => 'e',  'ĉ' => 'c',
  26                                 'ś' => 's',  'î' => 'i',  'ű' => 'u',  'ć' => 'c',  'ę' => 'e',  'ŵ' => 'w',  'ṫ' => 't',
  27                                 'ū' => 'u',  'č' => 'c',  'ö' => 'o',  'è' => 'e',  'ŷ' => 'y',  'ą' => 'a',  'ł' => 'l',
  28                                 'ų' => 'u',  'ů' => 'u',  'ş' => 's',  'ğ' => 'g',  'ļ' => 'l',  'ƒ' => 'f',  'ž' => 'z',
  29                                 'ẃ' => 'w',  'ḃ' => 'b',  'å' => 'a',  'ì' => 'i',  'ï' => 'i',  'ḋ' => 'd',  'ť' => 't',
  30                                 'ŗ' => 'r',  'ä' => 'a',  'í' => 'i',  'ŕ' => 'r',  'ê' => 'e',  'ü' => 'u',  'ò' => 'o',
  31                                 'ē' => 'e',  'ñ' => 'n',  'ń' => 'n',  'ĥ' => 'h',  'ĝ' => 'g',  'đ' => 'd',  'ĵ' => 'j',
  32                                 'ÿ' => 'y',  'ũ' => 'u',  'ŭ' => 'u',  'ư' => 'u',  'ţ' => 't',  'ý' => 'y',  'ő' => 'o',
  33                                 'â' => 'a',  'ľ' => 'l',  'ẅ' => 'w',  'ż' => 'z',  'ī' => 'i',  'ã' => 'a',  'ġ' => 'g',
  34                                 'ṁ' => 'm',  'ō' => 'o',  'ĩ' => 'i',  'ù' => 'u',  'į' => 'i',  'ź' => 'z',  'á' => 'a',
  35                                 'û' => 'u',  'þ' => 'th', 'ð' => 'dh', 'æ' => 'ae', 'µ' => 'u',  'ĕ' => 'e',  'ı' => 'i',
  36                         );
  37                 }
  38
  39                 $str = str_replace(
  40                         array_keys($utf8_lower_accents),
  41                         array_values($utf8_lower_accents),
  42                         $str
  43                 );
  44         }
  45
  46         if ($case >= 0)
  47         {
  48                 if ($utf8_upper_accents === NULL)
  49                 {
  50                         $utf8_upper_accents = array(
  51                                 'À' => 'A',  'Ô' => 'O',  'Ď' => 'D',  'Ḟ' => 'F',  'Ë' => 'E',  'Š' => 'S',  'Ơ' => 'O',
  52                                 'Ă' => 'A',  'Ř' => 'R',  'Ț' => 'T',  'Ň' => 'N',  'Ā' => 'A',  'Ķ' => 'K',  'Ĕ' => 'E',
  53                                 'Ŝ' => 'S',  'Ỳ' => 'Y',  'Ņ' => 'N',  'Ĺ' => 'L',  'Ħ' => 'H',  'Ṗ' => 'P',  'Ó' => 'O',
  54                                 'Ú' => 'U',  'Ě' => 'E',  'É' => 'E',  'Ç' => 'C',  'Ẁ' => 'W',  'Ċ' => 'C',  'Õ' => 'O',
  55                                 'Ṡ' => 'S',  'Ø' => 'O',  'Ģ' => 'G',  'Ŧ' => 'T',  'Ș' => 'S',  'Ė' => 'E',  'Ĉ' => 'C',
  56                                 'Ś' => 'S',  'Î' => 'I',  'Ű' => 'U',  'Ć' => 'C',  'Ę' => 'E',  'Ŵ' => 'W',  'Ṫ' => 'T',
  57                                 'Ū' => 'U',  'Č' => 'C',  'Ö' => 'O',  'È' => 'E',  'Ŷ' => 'Y',  'Ą' => 'A',  'Ł' => 'L',
  58                                 'Ų' => 'U',  'Ů' => 'U',  'Ş' => 'S',  'Ğ' => 'G',  'Ļ' => 'L',  'Ƒ' => 'F',  'Ž' => 'Z',
  59                                 'Ẃ' => 'W',  'Ḃ' => 'B',  'Å' => 'A',  'Ì' => 'I',  'Ï' => 'I',  'Ḋ' => 'D',  'Ť' => 'T',
  60                                 'Ŗ' => 'R',  'Ä' => 'A',  'Í' => 'I',  'Ŕ' => 'R',  'Ê' => 'E',  'Ü' => 'U',  'Ò' => 'O',
  61                                 'Ē' => 'E',  'Ñ' => 'N',  'Ń' => 'N',  'Ĥ' => 'H',  'Ĝ' => 'G',  'Đ' => 'D',  'Ĵ' => 'J',
  62                                 'Ÿ' => 'Y',  'Ũ' => 'U',  'Ŭ' => 'U',  'Ư' => 'U',  'Ţ' => 'T',  'Ý' => 'Y',  'Ő' => 'O',
  63                                 'Â' => 'A',  'Ľ' => 'L',  'Ẅ' => 'W',  'Ż' => 'Z',  'Ī' => 'I',  'Ã' => 'A',  'Ġ' => 'G',
  64                                 'Ṁ' => 'M',  'Ō' => 'O',  'Ĩ' => 'I',  'Ù' => 'U',  'Į' => 'I',  'Ź' => 'Z',  'Á' => 'A',
  65                                 'Û' => 'U',  'Þ' => 'Th', 'Ð' => 'Dh', 'Æ' => 'Ae', 'İ' => 'I',
  66                         );
  67                 }
  68
  69                 $str = str_replace(
  70                         array_keys($utf8_upper_accents),
  71                         array_values($utf8_upper_accents),
  72                         $str
  73                 );
  74         }
  75
  76         return $str;
  77 }