1 /*
2 +----------------------------------------------------------------------+
3 | PHP Version 7 |
4 +----------------------------------------------------------------------+
5 | Copyright (c) 1997-2016 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: Hartmut Holzgraefe <hholzgra@php.net> |
16 +----------------------------------------------------------------------+
17 */
18 /* $Id$ */
19
20 #include "php.h"
21 #include <stdlib.h>
22 #include <errno.h>
23 #include <ctype.h>
24 #include "php_string.h"
25
26 #define LEVENSHTEIN_MAX_LENGTH 255
27
28 /* {{{ reference_levdist
29 * reference implementation, only optimized for memory usage, not speed */
30 static zend_long reference_levdist(const char *s1, size_t l1, const char *s2, size_t l2, zend_long cost_ins, zend_long cost_rep, zend_long cost_del )
31 {
32 zend_long *p1, *p2, *tmp;
33 zend_long c0, c1, c2;
34 size_t i1, i2;
35
36 if (l1 == 0) {
37 return l2 * cost_ins;
38 }
39 if (l2 == 0) {
40 return l1 * cost_del;
41 }
42
43 if ((l1 > LEVENSHTEIN_MAX_LENGTH) || (l2 > LEVENSHTEIN_MAX_LENGTH)) {
44 return -1;
45 }
46 p1 = safe_emalloc((l2 + 1), sizeof(zend_long), 0);
47 p2 = safe_emalloc((l2 + 1), sizeof(zend_long), 0);
48
49 for (i2 = 0; i2 <= l2; i2++) {
50 p1[i2] = i2 * cost_ins;
51 }
52 for (i1 = 0; i1 < l1 ; i1++) {
53 p2[0] = p1[0] + cost_del;
54
55 for (i2 = 0; i2 < l2; i2++) {
56 c0 = p1[i2] + ((s1[i1] == s2[i2]) ? 0 : cost_rep);
57 c1 = p1[i2 + 1] + cost_del;
58 if (c1 < c0) {
59 c0 = c1;
60 }
61 c2 = p2[i2] + cost_ins;
62 if (c2 < c0) {
63 c0 = c2;
64 }
65 p2[i2 + 1] = c0;
66 }
67 tmp = p1;
68 p1 = p2;
69 p2 = tmp;
70 }
71 c0 = p1[l2];
72
73 efree(p1);
74 efree(p2);
75
76 return c0;
77 }
78 /* }}} */
79
80 /* {{{ custom_levdist
81 */
82 static int custom_levdist(char *str1, char *str2, char *callback_name)
83 {
84 php_error_docref(NULL, E_WARNING, "The general Levenshtein support is not there yet");
85 /* not there yet */
86
87 return -1;
88 }
89 /* }}} */
90
91 /* {{{ proto int levenshtein(string str1, string str2[, int cost_ins, int cost_rep, int cost_del])
92 Calculate Levenshtein distance between two strings */
93 PHP_FUNCTION(levenshtein)
94 {
95 int argc = ZEND_NUM_ARGS();
96 char *str1, *str2;
97 char *callback_name;
98 size_t str1_len, str2_len, callback_len;
99 zend_long cost_ins, cost_rep, cost_del;
100 zend_long distance = -1;
101
102 switch (argc) {
103 case 2: /* just two strings: use maximum performance version */
104 if (zend_parse_parameters(2, "ss", &str1, &str1_len, &str2, &str2_len) == FAILURE) {
105 return;
106 }
107 distance = reference_levdist(str1, str1_len, str2, str2_len, 1, 1, 1);
108 break;
109
110 case 5: /* more general version: calc cost by ins/rep/del weights */
111 if (zend_parse_parameters(5, "sslll", &str1, &str1_len, &str2, &str2_len, &cost_ins, &cost_rep, &cost_del) == FAILURE) {
112 return;
113 }
114 distance = reference_levdist(str1, str1_len, str2, str2_len, cost_ins, cost_rep, cost_del);
115 break;
116
117 case 3: /* most general version: calc cost by user-supplied function */
118 if (zend_parse_parameters(3, "sss", &str1, &str1_len, &str2, &str2_len, &callback_name, &callback_len) == FAILURE) {
119 return;
120 }
121 distance = custom_levdist(str1, str2, callback_name);
122 break;
123
124 default:
125 WRONG_PARAM_COUNT;
126 }
127
128 if (distance < 0 && /* TODO */ ZEND_NUM_ARGS() != 3) {
129 php_error_docref(NULL, E_WARNING, "Argument string(s) too long");
130 }
131
132 RETURN_LONG(distance);
133 }
134 /* }}} */
135
136 /*
137 * Local variables:
138 * tab-width: 4
139 * c-basic-offset: 4
140 * End:
141 * vim600: sw=4 ts=4 fdm=marker
142 * vim<600: sw=4 ts=4
143 */