1                 : /*
       2                 :    +----------------------------------------------------------------------+
       3                 :    | PHP Version 6                                                        |
       4                 :    +----------------------------------------------------------------------+
       5                 :    | Copyright (c) 1997-2009 The PHP Group                                |
       6                 :    +----------------------------------------------------------------------+
       7                 :    | This source file is subject to version 3.01 of the PHP license,      |
       8                 :    | that is bundled with this package in the file LICENSE, and is        |
       9                 :    | available through the world-wide-web at the following url:           |
      10                 :    | http://www.php.net/license/3_01.txt                                  |
      11                 :    | If you did not receive a copy of the PHP license and are unable to   |
      12                 :    | obtain it through the world-wide-web, please send a note to          |
      13                 :    | license@php.net so we can mail you a copy immediately.               |
      14                 :    +----------------------------------------------------------------------+
      15                 :    | Author: Bjørn Borud - Guardian Networks AS <borud@guardian.no>       |
      16                 :    +----------------------------------------------------------------------+
      17                 :  */
      18                 : /* $Id: soundex.c 276986 2009-03-10 23:40:06Z helly $ */
      19                 : 
      20                 : #include "php.h"
      21                 : #include <stdlib.h>
      22                 : #include <errno.h>
      23                 : #include <ctype.h>
      24                 : #include "php_string.h"
      25                 : 
      26                 : /* Simple soundex algorithm as described by Knuth in TAOCP, vol 3 */
      27                 : /* {{{ proto string soundex(string str)
      28                 :    Calculate the soundex key of a string */
      29                 : PHP_FUNCTION(soundex)
      30              42 : {
      31                 :         char    *str;
      32                 :         int     i, _small, str_len, code, last;
      33                 :         char    soundex[4 + 1];
      34                 : 
      35                 :         static char soundex_table[26] =
      36                 :         {0,                                                     /* A */
      37                 :          '1',                                           /* B */
      38                 :          '2',                                           /* C */
      39                 :          '3',                                           /* D */
      40                 :          0,                                                     /* E */
      41                 :          '1',                                           /* F */
      42                 :          '2',                                           /* G */
      43                 :          0,                                                     /* H */
      44                 :          0,                                                     /* I */
      45                 :          '2',                                           /* J */
      46                 :          '2',                                           /* K */
      47                 :          '4',                                           /* L */
      48                 :          '5',                                           /* M */
      49                 :          '5',                                           /* N */
      50                 :          0,                                                     /* O */
      51                 :          '1',                                           /* P */
      52                 :          '2',                                           /* Q */
      53                 :          '6',                                           /* R */
      54                 :          '2',                                           /* S */
      55                 :          '3',                                           /* T */
      56                 :          0,                                                     /* U */
      57                 :          '1',                                           /* V */
      58                 :          0,                                                     /* W */
      59                 :          '2',                                           /* X */
      60                 :          0,                                                     /* Y */
      61                 :          '2'};                                          /* Z */
      62                 : 
      63              42 :         if (zend_parse_parameters(ZEND_NUM_ARGS() TSRMLS_CC, "s", &str, &str_len) == FAILURE) {
      64               3 :                 return;
      65                 :         }
      66              39 :         if (str_len == 0) {
      67               1 :                 RETURN_FALSE;
      68                 :         }
      69                 : 
      70                 :         /* build soundex string */
      71              38 :         last = -1;
      72             238 :         for (i = 0, _small = 0; i < str_len && _small < 4; i++) {
      73                 :                 /* convert chars to upper case and strip non-letter chars */
      74                 :                 /* BUG: should also map here accented letters used in non */
      75                 :                 /* English words or names (also found in English text!): */
      76                 :                 /* esstsett, thorn, n-tilde, c-cedilla, s-caron, ... */
      77             200 :                 code = toupper((int)(unsigned char)str[i]);
      78             200 :                 if (code >= 'A' && code <= 'Z') {
      79             195 :                         if (_small == 0) {
      80                 :                                 /* remember first valid char */
      81              37 :                                 soundex[_small++] = code;
      82              37 :                                 last = soundex_table[code - 'A'];
      83                 :                         }
      84                 :                         else {
      85                 :                                 /* ignore sequences of consonants with same soundex */
      86                 :                                 /* code in trail, and vowels unless they separate */
      87                 :                                 /* consonant letters */
      88             158 :                                 code = soundex_table[code - 'A'];
      89             158 :                                 if (code != last) {
      90             116 :                                         if (code != 0) {
      91              66 :                                                 soundex[_small++] = code;
      92                 :                                         }
      93             116 :                                         last = code;
      94                 :                                 }
      95                 :                         }
      96                 :                 }
      97                 :         }
      98                 :         /* pad with '0' and terminate with 0 ;-) */
      99             125 :         while (_small < 4) {
     100              49 :                 soundex[_small++] = '0';
     101                 :         }
     102              38 :         soundex[_small] = '\0';
     103                 : 
     104              38 :         RETURN_STRINGL(soundex, _small, 1);
     105                 : }
     106                 : /* }}} */
     107                 : 
     108                 : /*
     109                 :  * Local variables:
     110                 :  * tab-width: 4
     111                 :  * c-basic-offset: 4
     112                 :  * End:
     113                 :  * vim600: sw=4 ts=4 fdm=marker
     114                 :  * vim<600: sw=4 ts=4
     115                 :  */