1 |
#include <string.h>
|
2 |
#include <stdio.h>
|
3 |
#include <glib.h>
|
4 |
#include "debug.h"
|
5 |
#include "linguistics.h"
|
6 |
|
7 |
static const char *special[][3]={
|
8 |
/* Capital Diacritics */
|
9 |
/* ¨ Diaresis */
|
10 |
{"Ä","A","AE"},
|
11 |
{"Ë","E"},
|
12 |
{"Ï","I"},
|
13 |
{"Ö","O","OE"},
|
14 |
{"Ü","U","UE"},
|
15 |
{"Ÿ","Y"},
|
16 |
/* ˝ Double Acute Accent */
|
17 |
{"Ő","O","Ö"},
|
18 |
{"Ű","U","Ü"},
|
19 |
/* ´ Acute Accent */
|
20 |
{"Á","A"},
|
21 |
{"Ć","C"},
|
22 |
{"É","E"},
|
23 |
{"Í","I"},
|
24 |
{"Ĺ","L"},
|
25 |
{"Ń","N"},
|
26 |
{"Ó","O"},
|
27 |
{"Ŕ","R"},
|
28 |
{"Ś","S"},
|
29 |
{"Ú","U"},
|
30 |
{"Ý","Y"},
|
31 |
{"Ź","Z"},
|
32 |
/* ˛ Ogonek (nosinė) */
|
33 |
{"Ą","A"},
|
34 |
{"Ę","E"},
|
35 |
{"Į","I"},
|
36 |
{"Ų","U"},
|
37 |
/* ˙ Dot */
|
38 |
{"Ċ","C"},
|
39 |
{"Ė","E"},
|
40 |
{"Ġ","G"},
|
41 |
{"İ","I"},
|
42 |
{"Ŀ","L"},
|
43 |
{"Ż","Z"},
|
44 |
/* – Stroke */
|
45 |
{"Đ","D","DJ"}, /* Croatian Dj, not to be confused with the similar-looking Icelandic Eth */
|
46 |
{"Ħ","H"},
|
47 |
{"Ł","L"},
|
48 |
{"Ŧ","T"},
|
49 |
/* ˚ Ring */
|
50 |
{"Å","A","AA"},
|
51 |
{"Ů","U"},
|
52 |
/* ˇ Caron (haček, paukščiukas) */
|
53 |
{"Č","C"},
|
54 |
{"Ď","D"},
|
55 |
{"Ě","E"},
|
56 |
{"Ľ","L"},
|
57 |
{"Ň","N"},
|
58 |
{"Ř","R"},
|
59 |
{"Š","S"},
|
60 |
{"Ť","T"},
|
61 |
{"Ž","Z"},
|
62 |
/* / Slash */
|
63 |
{"Ø","O","OE"},
|
64 |
/* ¯ Macron */
|
65 |
{"Ā","A","AA"},
|
66 |
{"Ē","E","EE"},
|
67 |
{"Ī","I","II"},
|
68 |
{"Ō","O","OO"},
|
69 |
{"Ū","U","UU"},
|
70 |
/* ˘ Brevis */
|
71 |
{"Ă","A"},
|
72 |
{"Ĕ","E"},
|
73 |
{"Ğ","G"},
|
74 |
{"Ĭ","I"},
|
75 |
{"Ŏ","O"},
|
76 |
{"Ŭ","U"},
|
77 |
/* ^ Circumflex */
|
78 |
{"Â","A"},
|
79 |
{"Ĉ","C"},
|
80 |
{"Ê","E"},
|
81 |
{"Ĝ","G"},
|
82 |
{"Ĥ","H"},
|
83 |
{"Î","I"},
|
84 |
{"Ĵ","J"},
|
85 |
{"Ô","O"},
|
86 |
{"Ŝ","S"},
|
87 |
{"Û","U"},
|
88 |
{"Ŵ","W"},
|
89 |
{"Ŷ","Y"},
|
90 |
/* ¸ Cedilla */
|
91 |
{"Ç","C"},
|
92 |
{"Ģ","G","GJ"},
|
93 |
{"Ķ","K","KJ"},
|
94 |
{"Ļ","L","LJ"},
|
95 |
{"Ņ","N","NJ"},
|
96 |
{"Ŗ","R"},
|
97 |
{"Ş","S"},
|
98 |
{"Ţ","T"},
|
99 |
/* ~ Tilde */
|
100 |
{"Ã","A"},
|
101 |
{"Ĩ","I"},
|
102 |
{"Ñ","N"},
|
103 |
{"Õ","O"},
|
104 |
{"Ũ","U"},
|
105 |
/* ` Grave */
|
106 |
{"À","A"},
|
107 |
{"È","E"},
|
108 |
{"Ì","I"},
|
109 |
{"Ò","O"},
|
110 |
{"Ù","U"},
|
111 |
/* ligatures */
|
112 |
{"Æ","A","AE"},
|
113 |
{"IJ","IJ"},
|
114 |
{"Œ","O","OE"},
|
115 |
/* special letters */
|
116 |
{"Ð","D","DH"}, /* Icelandic Eth, not to be confused with the similar-looking Croatian Dj */
|
117 |
{"Ŋ","N","NG"},
|
118 |
{"Þ","T","TH"},
|
119 |
/* Small Diacritics */
|
120 |
/* ¨ Diaresis */
|
121 |
{"ä","a","ae"},
|
122 |
{"ë","e"},
|
123 |
{"ï","i"},
|
124 |
{"ö","o","oe"},
|
125 |
{"ü","u","ue"},
|
126 |
{"ÿ","y"},
|
127 |
/* ˝ Double Acute Accent */
|
128 |
{"ő","o","ö"},
|
129 |
{"ű","u","ü"},
|
130 |
/* ´ Acute Accent */
|
131 |
{"á","a"},
|
132 |
{"ć","c"},
|
133 |
{"é","e"},
|
134 |
{"í","i"},
|
135 |
{"ĺ","l"},
|
136 |
{"ń","n"},
|
137 |
{"ó","o"},
|
138 |
{"ŕ","r"},
|
139 |
{"ś","s"},
|
140 |
{"ú","u"},
|
141 |
{"ý","y"},
|
142 |
{"ź","z"},
|
143 |
/* ˛ Ogonek (nosinė) */
|
144 |
{"ą","a"},
|
145 |
{"ę","e"},
|
146 |
{"į","i"},
|
147 |
{"ų","u"},
|
148 |
/* ˙ Dot (and dotless i) */
|
149 |
{"ċ","c"},
|
150 |
{"ė","e"},
|
151 |
{"ġ","g"},
|
152 |
{"ı","i"},
|
153 |
{"ŀ","l"},
|
154 |
{"ż","z"},
|
155 |
/* – Stroke */
|
156 |
{"đ","d","dj"},
|
157 |
{"ħ","h"},
|
158 |
{"ł","l"},
|
159 |
{"ŧ","t"},
|
160 |
/* ˚ Ring */
|
161 |
{"å","a", "aa"},
|
162 |
{"ů","u"},
|
163 |
/* ˇ Caron (haček, paukščiukas) */
|
164 |
{"č","c"},
|
165 |
{"ď","d"},
|
166 |
{"ě","e"},
|
167 |
{"ľ","l"},
|
168 |
{"ň","n"},
|
169 |
{"ř","r"},
|
170 |
{"š","s"},
|
171 |
{"ť","t"},
|
172 |
{"ž","z"},
|
173 |
/* / Slash */
|
174 |
{"ø","o", "oe"},
|
175 |
/* Macron */
|
176 |
{"ā","a","aa"},
|
177 |
{"ē","e","ee"},
|
178 |
{"ī","i","ii"},
|
179 |
{"ō","o","oo"},
|
180 |
{"ū","u","uu"},
|
181 |
/* ˘ Brevis */
|
182 |
{"ă","a"},
|
183 |
{"ĕ","e"},
|
184 |
{"ğ","g"},
|
185 |
{"ĭ","i"},
|
186 |
{"ŏ","o"},
|
187 |
{"ŭ","u"},
|
188 |
/* ^ Circumflex */
|
189 |
{"â","a"},
|
190 |
{"ĉ","c"},
|
191 |
{"ê","e"},
|
192 |
{"ĝ","g"},
|
193 |
{"ĥ","h"},
|
194 |
{"î","i"},
|
195 |
{"ĵ","j"},
|
196 |
{"ô","o"},
|
197 |
{"ŝ","s"},
|
198 |
{"û","u"},
|
199 |
{"ŵ","w"},
|
200 |
{"ŷ","y"},
|
201 |
/* ¸ Cedilla */
|
202 |
{"ç","c"},
|
203 |
{"ģ","g","gj"},
|
204 |
{"ķ","k","kj"},
|
205 |
{"ļ","l","lj"},
|
206 |
{"ņ","n","nj"},
|
207 |
{"ŗ","r"},
|
208 |
{"ş","s"},
|
209 |
{"ţ","t"},
|
210 |
/* ~ Tilde */
|
211 |
{"ã","a"},
|
212 |
{"ĩ","i"},
|
213 |
{"õ","o"},
|
214 |
{"ñ","n"},
|
215 |
{"ũ","u"},
|
216 |
/* ` Grave */
|
217 |
{"à","a"},
|
218 |
{"è","e"},
|
219 |
{"ì","i"},
|
220 |
{"ò","o"},
|
221 |
{"ù","u"},
|
222 |
/* ligatures */
|
223 |
{"æ","a","ae"},
|
224 |
{"ij","ij"},
|
225 |
{"œ","o","oe"},
|
226 |
{"ß","s","ss"},
|
227 |
/* special letters */
|
228 |
{"ð","d","dh"},
|
229 |
{"ŋ","n","ng"},
|
230 |
{"þ","t","th"},
|
231 |
|
232 |
/* Cyrillic capital */
|
233 |
|
234 |
{"Ё","Е"},
|
235 |
{"Й","И"},
|
236 |
{"І","I"},
|
237 |
{"Ї","I"},
|
238 |
{"Ў","У"},
|
239 |
{"Є","Е","Э"},
|
240 |
{"Ґ","Г"},
|
241 |
{"Ѓ","Г"},
|
242 |
{"Ђ","Д"},
|
243 |
{"Ќ","К"},
|
244 |
//{"Љ","Л","ЛЬ"},
|
245 |
//{"Њ","Н","НЬ"},
|
246 |
{"Џ","Ц"},
|
247 |
|
248 |
/* Cyrillic small */
|
249 |
|
250 |
{"ё","е"},
|
251 |
{"й","и"},
|
252 |
{"і","i"},
|
253 |
{"ї","i"},
|
254 |
{"ў","у"},
|
255 |
//{"є","е","э"},
|
256 |
{"ґ","г"},
|
257 |
{"ѓ","г"},
|
258 |
{"ђ","д"},
|
259 |
{"ќ","к"},
|
260 |
//{"љ","л","ль"},
|
261 |
//{"њ","н","нь"},
|
262 |
{"џ","ц"},
|
263 |
|
264 |
};
|
265 |
|
266 |
static GHashTable *special_hash;
|
267 |
|
268 |
/* Array of strings for case conversion
|
269 |
* Even elements of array are strings of upper-case letters
|
270 |
* Odd elements of array are strings of lower-case letters, in the order corresponding to directly preceeding even element.
|
271 |
* Last element of array should be NULL.
|
272 |
*/
|
273 |
static const char *upperlower[]={
|
274 |
/*Latin diacritics*/
|
275 |
"ÄËÏÖÜŸŐŰÁĆÉÍĹŃÓŔŚÚÝŹĄĘĮŲĊĖĠİĿŻĐĦŁŦÅŮČĎĚĽŇŘŠŤŽØĀĒĪŌŪĂĔĞĬŎŬÂĈÊĜĤÎĴÔŜÛŴŶÇĢĶĻŅŖŞŢÃĨÑÕŨÀÈÌÒÙÆIJŒÐŊÞ",
|
276 |
"äëïöüÿőűáćéíĺńóŕśúýźąęįųċėġıŀżđħłŧåůčďěľňřšťžøāēīōūăĕğĭŏŭâĉêĝĥîĵôŝûŵŷçģķļņŗşţãĩõñũàèìòùæijœðŋþ",
|
277 |
/*Cyrillic*/
|
278 |
"АБВГҐЃДЂЕЄЁЖЗИЙКЌЛЉМНЊОПРСТУФХЦЏЧШЩЪЫЬЭЮЯІЇЎ",
|
279 |
"абвгґѓдђеєёжзийкќлљмнњопрстуфхцџчшщъыьэюяіїў",
|
280 |
|
281 |
NULL
|
282 |
};
|
283 |
|
284 |
static GHashTable *casefold_hash;
|
285 |
|
286 |
|
287 |
struct special_pos {
|
288 |
char **variants;
|
289 |
int n;
|
290 |
char *s1, *s2;
|
291 |
};
|
292 |
|
293 |
|
294 |
|
295 |
static char**
|
296 |
linguistics_get_special(char *str, char *end)
|
297 |
{
|
298 |
char buf[10];
|
299 |
int len;
|
300 |
if(!end)
|
301 |
end=g_utf8_find_next_char(str,NULL);
|
302 |
len=end-str+1;
|
303 |
g_strlcpy(buf,str,len>10?10:len);
|
304 |
return g_hash_table_lookup(special_hash,buf);
|
305 |
}
|
306 |
|
307 |
|
308 |
/*
|
309 |
* @brief Prepare an utf-8 string for case insensitive comparison.
|
310 |
* @param in String to prepeare.
|
311 |
* @return String prepared for case insensitive search. Result shoud be g_free()d after use.
|
312 |
*/
|
313 |
char*
|
314 |
linguistics_casefold(char *in)
|
315 |
{
|
316 |
int len=strlen(in);
|
317 |
char *src=in;
|
318 |
char *ret=g_new(char,len+1);
|
319 |
char *dest=ret;
|
320 |
char buf[10];
|
321 |
while(*src && dest-ret<len){
|
322 |
if(*src>='A' && *src<='Z') {
|
323 |
*dest++=*src++ - 'A' + 'a';
|
324 |
} else if (!(*src&128)) {
|
325 |
*dest++=*src++;
|
326 |
} else {
|
327 |
int charlen;
|
328 |
char *tmp, *folded;
|
329 |
tmp=g_utf8_find_next_char(src,NULL);
|
330 |
charlen=tmp-src+1;
|
331 |
g_strlcpy(buf,src,charlen>10?10:charlen);
|
332 |
folded=g_hash_table_lookup(casefold_hash,buf);
|
333 |
|
334 |
if(folded)
|
335 |
{
|
336 |
while(*folded && dest-ret<len)
|
337 |
*dest++=*folded++;
|
338 |
src=tmp;
|
339 |
}
|
340 |
else
|
341 |
{
|
342 |
while(src<tmp && dest-ret<len)
|
343 |
*dest++=*src++;
|
344 |
}
|
345 |
}
|
346 |
}
|
347 |
*dest=0;
|
348 |
if(*src)
|
349 |
dbg(0,"Casefolded string for '%s' needs extra space, result is trucated to '%s'.\n",in,ret);
|
350 |
return ret;
|
351 |
}
|
352 |
|
353 |
/**
|
354 |
* @brief Compare two strings using special characters expansion.
|
355 |
*
|
356 |
* @param str first string to compare, special characters are expanded.
|
357 |
* @param match second string to compare, special characters are not expanded.
|
358 |
* @param partial if = 1 then str string may be shorter than match string, in which case the rest from str isn't analysed.
|
359 |
* @return =0 strings matched, =1 not matched. Note this function return value is not fully compatible with strcmp().
|
360 |
*/
|
361 |
|
362 |
int
|
363 |
linguistics_compare(char *str, char *match, int partial)
|
364 |
{
|
365 |
char *s1=str, *s2=match;
|
366 |
char **sp;
|
367 |
int ret=0;
|
368 |
int got_match;
|
369 |
GList *l=NULL;
|
370 |
while (*s1 && *s2) {
|
371 |
int j;
|
372 |
struct special_pos *spp;
|
373 |
char *utf_boundary, *tmp;
|
374 |
/* Skip all matching chars */
|
375 |
for(j=0;s1[j] && s1[j]==s2[j];j++);
|
376 |
if(!s2[j] && (partial || !s1[j])) {
|
377 |
/* MATCH! */
|
378 |
ret=0;
|
379 |
break;
|
380 |
}
|
381 |
/* Find beginning of first mismatching utf-8 encoded char */
|
382 |
utf_boundary=s1;
|
383 |
while(*(tmp=g_utf8_find_next_char(utf_boundary, NULL))) {
|
384 |
if(tmp>s1+j)
|
385 |
break;
|
386 |
utf_boundary=tmp;
|
387 |
}
|
388 |
/* Push first mismatching char to the list if it's a special char */
|
389 |
sp=linguistics_get_special(utf_boundary,tmp);
|
390 |
|
391 |
if(sp)
|
392 |
{
|
393 |
spp=g_new(struct special_pos,1);
|
394 |
spp->variants=sp;
|
395 |
spp->n=1;
|
396 |
spp->s1=utf_boundary;
|
397 |
spp->s2=s2+(utf_boundary-s1);
|
398 |
l=g_list_prepend(l,spp);
|
399 |
}
|
400 |
|
401 |
/* Try to find a match using special char variants from the list */
|
402 |
got_match=0;
|
403 |
while(l && !got_match) {
|
404 |
spp=l->data;
|
405 |
s1=spp->s1;
|
406 |
s2=spp->s2;
|
407 |
while(spp->n<3 && !got_match) {
|
408 |
char *s=spp->variants[(spp->n)++];
|
409 |
int len;
|
410 |
if(!s)
|
411 |
break;
|
412 |
len=strlen(s);
|
413 |
if(!strncmp(s,s2,len)) {
|
414 |
s2+=len;
|
415 |
s1+=strlen(spp->variants[0]);
|
416 |
got_match=1;
|
417 |
break;
|
418 |
}
|
419 |
}
|
420 |
if(spp->n>=3 || !spp->variants[spp->n]) {
|
421 |
/* No matches for current top list element, go to the closest special char towards beginning of the string */
|
422 |
g_free(spp);
|
423 |
l=g_list_delete_link(l,l);
|
424 |
}
|
425 |
}
|
426 |
if(!got_match) {
|
427 |
/* NO MATCH
|
428 |
* FIXME: If we're going to use this function to sort a string list alphabetically we should use
|
429 |
* utf-aware comparison here.
|
430 |
*/
|
431 |
ret=1;
|
432 |
break;
|
433 |
}
|
434 |
}
|
435 |
while(l) {
|
436 |
g_free(l->data);
|
437 |
l=g_list_delete_link(l,l);
|
438 |
}
|
439 |
return ret;
|
440 |
}
|
441 |
|
442 |
|
443 |
char *
|
444 |
linguistics_expand_special(char *str, int mode)
|
445 |
{
|
446 |
char *in=str;
|
447 |
char *out,*ret;
|
448 |
int found=0;
|
449 |
|
450 |
if (!str)
|
451 |
{
|
452 |
return NULL;
|
453 |
}
|
454 |
|
455 |
ret=g_strdup(str);
|
456 |
out=ret;
|
457 |
|
458 |
if (!mode)
|
459 |
{
|
460 |
return ret;
|
461 |
}
|
462 |
|
463 |
while (*in)
|
464 |
{
|
465 |
char *next=g_utf8_find_next_char(in, NULL);
|
466 |
int i,len=next-in;
|
467 |
int match=0;
|
468 |
if (len > 1)
|
469 |
{
|
470 |
for (i = 0 ; i < sizeof(special)/sizeof(special[0]); i++)
|
471 |
{
|
472 |
const char *search=special[i][0];
|
473 |
if (!strncmp(in,search,len))
|
474 |
{
|
475 |
const char *replace=special[i][mode];
|
476 |
if (replace)
|
477 |
{
|
478 |
int replace_len=strlen(replace);
|
479 |
|
480 |
if (replace_len > len)
|
481 |
{
|
482 |
fprintf(stderr,"* ERROR !! ERROR !! found %s %s %d %s %d\n",in,search,len,replace,replace_len);
|
483 |
}
|
484 |
dbg_assert(replace_len <= len);
|
485 |
if (replace_len > len)
|
486 |
{
|
487 |
out+=len;
|
488 |
match=0;
|
489 |
break;
|
490 |
}
|
491 |
else
|
492 |
{
|
493 |
// fprintf(stderr," GOOD !! GOOD !! found %s %s %d %s %d\n",in,search,len,replace,replace_len);
|
494 |
strcpy(out, replace);
|
495 |
out+=replace_len;
|
496 |
match=1;
|
497 |
break;
|
498 |
}
|
499 |
}
|
500 |
}
|
501 |
}
|
502 |
}
|
503 |
|
504 |
if (match)
|
505 |
{
|
506 |
found=1;
|
507 |
in=next;
|
508 |
}
|
509 |
else
|
510 |
{
|
511 |
while (len-- > 0)
|
512 |
{
|
513 |
*out++=*in++;
|
514 |
}
|
515 |
}
|
516 |
}
|
517 |
*out++='\0';
|
518 |
if (!found)
|
519 |
{
|
520 |
if (ret)
|
521 |
{
|
522 |
g_free(ret);
|
523 |
}
|
524 |
ret=NULL;
|
525 |
}
|
526 |
return ret;
|
527 |
}
|
528 |
|
529 |
char *
|
530 |
linguistics_next_word(char *str)
|
531 |
{
|
532 |
char* ret=strtok(str, " -/()\"\',.;_[]{}\\");
|
533 |
return ret;
|
534 |
|
535 |
// int len=strcspn(str, " -/()");
|
536 |
// if (!str[len] || !str[len+1])
|
537 |
// return NULL;
|
538 |
// return str+len+1;
|
539 |
|
540 |
}
|
541 |
|
542 |
int
|
543 |
linguistics_search(char *str)
|
544 |
{
|
545 |
if (!g_strcasecmp(str,"str"))
|
546 |
return 0;
|
547 |
if (!g_strcasecmp(str,"str."))
|
548 |
return 0;
|
549 |
if (!g_strcasecmp(str,"strasse"))
|
550 |
return 0;
|
551 |
if (!g_strcasecmp(str,"weg"))
|
552 |
return 0;
|
553 |
return 1;
|
554 |
}
|
555 |
|
556 |
/**
|
557 |
* @brief Copy one utf8 encoded char to newly allocated buffer.
|
558 |
*
|
559 |
* @param s pointer to the beginning of the char.
|
560 |
* @return newly allocated nul-terminated string containing one utf8 encoded character.
|
561 |
*/
|
562 |
static char
|
563 |
*linguistics_dup_utf8_char(const char *s)
|
564 |
{
|
565 |
char *ret, *next;
|
566 |
next=g_utf8_find_next_char(s,NULL);
|
567 |
ret=g_new(char, next-s+1);
|
568 |
g_strlcpy(ret,s,next-s+1);
|
569 |
return ret;
|
570 |
}
|
571 |
|
572 |
void
|
573 |
linguistics_init(void)
|
574 |
{
|
575 |
int i;
|
576 |
special_hash=g_hash_table_new(g_str_hash, g_str_equal);
|
577 |
casefold_hash=g_hash_table_new(g_str_hash, g_str_equal);
|
578 |
|
579 |
for (i = 0 ; i < sizeof(special)/sizeof(special[0]); i++)
|
580 |
{
|
581 |
g_hash_table_insert(special_hash,(gpointer)special[i][0],special[i]);
|
582 |
}
|
583 |
|
584 |
for (i = 0 ; upperlower[i]; i+=2)
|
585 |
{
|
586 |
int j,k;
|
587 |
for(j=0,k=0;upperlower[i][j] && upperlower[i+1][k];)
|
588 |
{
|
589 |
char *s1=linguistics_dup_utf8_char(upperlower[i]+j);
|
590 |
char *s2=linguistics_dup_utf8_char(upperlower[i+1]+k);
|
591 |
g_hash_table_insert(casefold_hash,s1,s2);
|
592 |
j+=strlen(s1);
|
593 |
k+=strlen(s2);
|
594 |
}
|
595 |
}
|
596 |
}
|
597 |
|