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
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
|
/*
* morphology module
* New dictionary is include in dict.h. For languages which
* use latin charset it may be need to modify mapdict table.
* Teodor Sigaev <teodor@stack.net>
*/
#include "postgres.h"
#include "utils/elog.h"
#include "utils/palloc.h"
#include "utils/builtins.h"
#include "catalog/pg_control.h"
#include "utils/pg_locale.h"
#include "morph.h"
#include "deflex.h"
/*
* Struct for calling dictionaries
* All of this methods are optional, but
* if all methods are NULL, then dictionary does nothing :)
* Return value of lemmatize must be palloced or the same.
* Return value of init must be malloced in other case
* it will be free in end of transaction!
*/
typedef struct
{
char localename[LOCALE_NAME_BUFLEN];
/* init dictionary */
void *(*init) (void);
/* close dictionary */
void (*close) (void *);
/* find in dictionary */
char *(*lemmatize) (void *, char *, int *);
int (*is_stoplemm) (void *, char *, int);
int (*is_stemstoplemm) (void *, char *, int);
} DICT;
/* insert all dictionaries */
#define DICT_BODY
#include "dict.h"
#undef DICT_BODY
/* fill dictionary's structure */
#define DICT_TABLE
DICT dicts[] = {
{
"C", NULL, NULL, NULL, NULL, NULL /* fake dictionary */
}
#include "dict.h"
};
#undef DICT_TABLE
/* array for storing dictinary's objects (if needed) */
void *dictobjs[
lengthof(dicts)];
#define STOPLEXEM -2
#define BYLOCALE -1
#define NODICT 0
#define DEFAULTDICT 1
#define MAXNDICT 2
typedef int2 MAPDICT[MAXNDICT];
#define GETDICT(x,i) *( ((int2*)(x)) + (i) )
/* map dictionaries for lexem type */
static MAPDICT mapdict[] = {
{NODICT, NODICT}, /* not used */
{DEFAULTDICT, NODICT}, /* LATWORD */
{BYLOCALE, NODICT}, /* NONLATINWORD */
{BYLOCALE, DEFAULTDICT}, /* UWORD */
{NODICT, NODICT}, /* EMAIL */
{NODICT, NODICT}, /* FURL */
{NODICT, NODICT}, /* HOST */
{NODICT, NODICT}, /* FLOAT */
{NODICT, NODICT}, /* FINT */
{BYLOCALE, DEFAULTDICT}, /* PARTWORD */
{BYLOCALE, NODICT}, /* NONLATINPARTWORD */
{DEFAULTDICT, NODICT}, /* LATPARTWORD */
{STOPLEXEM, NODICT}, /* SPACE */
{STOPLEXEM, NODICT}, /* SYMTAG */
{STOPLEXEM, NODICT}, /* HTTP */
{BYLOCALE, DEFAULTDICT}, /* DEFISWORD */
{DEFAULTDICT, NODICT}, /* DEFISLATWORD */
{BYLOCALE, NODICT}, /* DEFISNONLATINWORD */
{NODICT, NODICT}, /* URI */
{NODICT, NODICT} /* FILEPATH */
};
static bool inited = false;
void
initmorph(void)
{
int i,
j,
k;
MAPDICT *md;
bool needinit[lengthof(dicts)];
#ifdef USE_LOCALE
PG_LocaleCategories lc;
int bylocaledict = NODICT;
#endif
if (inited)
return;
for (i = 1; i < lengthof(dicts); i++)
needinit[i] = false;
#ifdef USE_LOCALE
PGLC_current(&lc);
for (i = 1; i < lengthof(dicts); i++)
if (strcmp(dicts[i].localename, lc.lang) == 0)
{
bylocaledict = i;
break;
}
PGLC_free_categories(&lc);
#endif
for (i = 1; i < lengthof(mapdict); i++)
{
k = 0;
md = &mapdict[i];
for (j = 0; j < MAXNDICT; j++)
{
GETDICT(md, k) = GETDICT(md, j);
if (GETDICT(md, k) == NODICT)
break;
else if (GETDICT(md, k) == BYLOCALE)
{
#ifdef USE_LOCALE
if (bylocaledict == NODICT)
continue;
GETDICT(md, k) = bylocaledict;
#else
continue;
#endif
}
if (GETDICT(md, k) >= (int2) lengthof(dicts))
continue;
needinit[GETDICT(md, k)] = true;
k++;
}
for (; k < MAXNDICT; k++)
if (GETDICT(md, k) != STOPLEXEM)
GETDICT(md, k) = NODICT;
}
for (i = 1; i < lengthof(dicts); i++)
if (needinit[i] && dicts[i].init)
dictobjs[i] = (*(dicts[i].init)) ();
inited = true;
return;
}
char *
lemmatize(char *word, int *len, int type)
{
int2 nd;
int i;
DICT *dict;
for (i = 0; i < MAXNDICT; i++)
{
nd = GETDICT(&mapdict[type], i);
if (nd == NODICT)
{
/* there is no dictionary */
return word;
}
else if (nd == STOPLEXEM)
{
/* word is stopword */
return NULL;
}
else
{
dict = &dicts[nd];
if (dict->is_stoplemm && (*(dict->is_stoplemm)) (dictobjs[nd], word, *len))
return NULL;
if (dict->lemmatize)
{
int oldlen = *len;
char *newword = (*(dict->lemmatize)) (dictobjs[nd], word, len);
/* word is recognized by distionary */
if (newword != word || *len != oldlen)
{
if (dict->is_stemstoplemm &&
(*(dict->is_stemstoplemm)) (dictobjs[nd], word, *len))
{
if (newword != word && newword)
pfree(newword);
return NULL;
}
return newword;
}
}
}
}
return word;
}
bool
is_stoptype(int type)
{
return (GETDICT(&mapdict[type], 0) == STOPLEXEM) ? true : false;
}
|
