-
Notifications
You must be signed in to change notification settings - Fork 40
/
euclidean.c
198 lines (168 loc) · 4.37 KB
/
euclidean.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
/*----------------------------------------------------------------------------
*
* euclidean.c
*
* Euclidean distance
*
* aims to measure the distance between two strings
*
* X is a list of n-grams
* Y is a list of n-grams
* T is a set of n-grams of X and/or Y
*
* For each n-gram in T we count occorrences in X and Y; we sum the
* quadratic difference between nx and ny. At the end, we get the
* square root of the sum.
*
* For example:
*
* x: euler = {eu, ul, le, er}
* y: heuser = {he, eu, us, se, er}
* t: {eu, ul, le, er, he, us, se}
*
* eu ul le er he us se
* s = sqrt((1 - 1)^2 + (1 - 0)^2 + (1 - 0)^2 + (1 - 1)^2 + (0 - 1)^2 + (0 - 1)^2 + (0 - 1)^2) =
* s = sqrt(5) = 2.236067977...
*
* PS> we call n-grams: (i) n-sequence of letters (ii) n-sequence of words
*
*
* Copyright (c) 2008-2020, Euler Taveira de Oliveira
*
*----------------------------------------------------------------------------
*/
#include "similarity.h"
#include "tokenizer.h"
#include <math.h>
/* GUC variables */
int pgs_euclidean_tokenizer = PGS_UNIT_ALNUM;
double pgs_euclidean_threshold = 0.7f;
bool pgs_euclidean_is_normalized = true;
PG_FUNCTION_INFO_V1(euclidean);
Datum
euclidean(PG_FUNCTION_ARGS)
{
char *a, *b;
TokenList *s, *t, *u;
Token *p, *q, *r;
double totdistance;
double totpossible;
float8 res;
a = DatumGetPointer(DirectFunctionCall1(textout,
PointerGetDatum(PG_GETARG_TEXT_P(0))));
b = DatumGetPointer(DirectFunctionCall1(textout,
PointerGetDatum(PG_GETARG_TEXT_P(1))));
if (strlen(a) > PGS_MAX_STR_LEN || strlen(b) > PGS_MAX_STR_LEN)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("argument exceeds the maximum length of %d bytes",
PGS_MAX_STR_LEN)));
/* lists */
s = initTokenList(0);
t = initTokenList(0);
/* set list */
u = initTokenList(1);
switch (pgs_euclidean_tokenizer)
{
case PGS_UNIT_WORD:
tokenizeBySpace(s, a);
tokenizeBySpace(t, b);
/* all tokens in a set */
tokenizeBySpace(u, a);
tokenizeBySpace(u, b);
break;
case PGS_UNIT_GRAM:
tokenizeByGram(s, a);
tokenizeByGram(t, b);
/* all tokens in a set */
tokenizeByGram(u, a);
tokenizeByGram(u, b);
break;
case PGS_UNIT_CAMELCASE:
tokenizeByCamelCase(s, a);
tokenizeByCamelCase(t, b);
/* all tokens in a set */
tokenizeByCamelCase(u, a);
tokenizeByCamelCase(u, b);
break;
case PGS_UNIT_ALNUM: /* default */
default:
tokenizeByNonAlnum(s, a);
tokenizeByNonAlnum(t, b);
/* all tokens in a set */
tokenizeByNonAlnum(u, a);
tokenizeByNonAlnum(u, b);
break;
}
elog(DEBUG3, "Token List A");
printToken(s);
elog(DEBUG3, "Token List B");
printToken(t);
elog(DEBUG3, "All Token List");
printToken(u);
totpossible = sqrt(s->size * s->size + t->size * t->size);
totdistance = 0.0;
p = u->head;
while (p != NULL)
{
int acnt = 0;
int bcnt = 0;
q = s->head;
while (q != NULL)
{
elog(DEBUG4, "p: %s; q: %s", p->data, q->data);
if (strcmp(p->data, q->data) == 0)
{
acnt++;
break;
}
q = q->next;
}
r = t->head;
while (r != NULL)
{
elog(DEBUG4, "p: %s; r: %s", p->data, r->data);
if (strcmp(p->data, r->data) == 0)
{
bcnt++;
break;
}
r = r->next;
}
totdistance += (acnt - bcnt) * (acnt - bcnt);
elog(DEBUG2,
"\"%s\" => acnt(%d); bcnt(%d); totdistance(%.2f)",
p->data, acnt, bcnt, totdistance);
p = p->next;
}
totdistance = sqrt(totdistance);
elog(DEBUG1, "is normalized: %d", pgs_euclidean_is_normalized);
elog(DEBUG1, "total possible: %.2f", totpossible);
elog(DEBUG1, "total distance: %.2f", totdistance);
destroyTokenList(s);
destroyTokenList(t);
destroyTokenList(u);
if (pgs_euclidean_is_normalized)
res = (totpossible - totdistance) / totpossible;
else
res = totdistance;
PG_RETURN_FLOAT8(res);
}
PG_FUNCTION_INFO_V1(euclidean_op);
Datum euclidean_op(PG_FUNCTION_ARGS)
{
float8 res;
/*
* store *_is_normalized value temporarily 'cause
* threshold (we're comparing against) is normalized
*/
bool tmp = pgs_euclidean_is_normalized;
pgs_euclidean_is_normalized = true;
res = DatumGetFloat8(DirectFunctionCall2(
euclidean,
PG_GETARG_DATUM(0),
PG_GETARG_DATUM(1)));
/* we're done; back to the previous value */
pgs_euclidean_is_normalized = tmp;
PG_RETURN_BOOL(res >= pgs_euclidean_threshold);
}