| OLD | NEW |
|---|
| (Empty) |
| 1 #include "license.hunspell" | |
| 2 #include "license.myspell" | |
| 3 | |
| 4 #include <stdlib.h> | |
| 5 #include <string.h> | |
| 6 #include <stdio.h> | |
| 7 | |
| 8 #include "hunspell.hxx" | |
| 9 #include "hunspell.h" | |
| 10 #ifndef HUNSPELL_CHROME_CLIENT | |
| 11 #ifndef MOZILLA_CLIENT | |
| 12 # include "config.h" | |
| 13 #endif | |
| 14 #endif | |
| 15 #include "csutil.hxx" | |
| 16 | |
| 17 #ifdef HUNSPELL_CHROME_CLIENT | |
| 18 Hunspell::Hunspell(const unsigned char* bdict_data, size_t bdict_length) | |
| 19 #else | |
| 20 Hunspell::Hunspell(const char * affpath, const char * dpath, const char * key) | |
| 21 #endif | |
| 22 { | |
| 23 encoding = NULL; | |
| 24 csconv = NULL; | |
| 25 utf8 = 0; | |
| 26 complexprefixes = 0; | |
| 27 #ifndef HUNSPELL_CHROME_CLIENT | |
| 28 affixpath = mystrdup(affpath); | |
| 29 #endif | |
| 30 maxdic = 0; | |
| 31 | |
| 32 #ifdef HUNSPELL_CHROME_CLIENT | |
| 33 bdict_reader = new hunspell::BDictReader; | |
| 34 bdict_reader->Init(bdict_data, bdict_length); | |
| 35 | |
| 36 pHMgr[0] = new HashMgr(bdict_reader); | |
| 37 if (pHMgr[0]) maxdic = 1; | |
| 38 | |
| 39 pAMgr = new AffixMgr(bdict_reader, pHMgr, &maxdic); | |
| 40 #else | |
| 41 /* first set up the hash manager */ | |
| 42 pHMgr[0] = new HashMgr(dpath, affpath, key); | |
| 43 if (pHMgr[0]) maxdic = 1; | |
| 44 | |
| 45 /* next set up the affix manager */ | |
| 46 /* it needs access to the hash manager lookup methods */ | |
| 47 pAMgr = new AffixMgr(affpath, pHMgr, &maxdic, key); | |
| 48 #endif | |
| 49 | |
| 50 /* get the preferred try string and the dictionary */ | |
| 51 /* encoding from the Affix Manager for that dictionary */ | |
| 52 char * try_string = pAMgr->get_try_string(); | |
| 53 encoding = pAMgr->get_encoding(); | |
| 54 langnum = pAMgr->get_langnum(); | |
| 55 utf8 = pAMgr->get_utf8(); | |
| 56 if (!utf8) | |
| 57 csconv = get_current_cs(encoding); | |
| 58 complexprefixes = pAMgr->get_complexprefixes(); | |
| 59 wordbreak = pAMgr->get_breaktable(); | |
| 60 | |
| 61 /* and finally set up the suggestion manager */ | |
| 62 #ifdef HUNSPELL_CHROME_CLIENT | |
| 63 pSMgr = new SuggestMgr(bdict_reader, try_string, MAXSUGGESTION, pAMgr); | |
| 64 #else | |
| 65 pSMgr = new SuggestMgr(try_string, MAXSUGGESTION, pAMgr); | |
| 66 #endif | |
| 67 if (try_string) free(try_string); | |
| 68 } | |
| 69 | |
| 70 Hunspell::~Hunspell() | |
| 71 { | |
| 72 if (pSMgr) delete pSMgr; | |
| 73 if (pAMgr) delete pAMgr; | |
| 74 for (int i = 0; i < maxdic; i++) delete pHMgr[i]; | |
| 75 maxdic = 0; | |
| 76 pSMgr = NULL; | |
| 77 pAMgr = NULL; | |
| 78 #ifdef MOZILLA_CLIENT | |
| 79 delete [] csconv; | |
| 80 #endif | |
| 81 csconv= NULL; | |
| 82 if (encoding) free(encoding); | |
| 83 encoding = NULL; | |
| 84 #ifdef HUNSPELL_CHROME_CLIENT | |
| 85 if (bdict_reader) delete bdict_reader; | |
| 86 bdict_reader = NULL; | |
| 87 #else | |
| 88 if (affixpath) free(affixpath); | |
| 89 affixpath = NULL; | |
| 90 #endif | |
| 91 } | |
| 92 | |
| 93 #ifndef HUNSPELL_CHROME_CLIENT | |
| 94 // load extra dictionaries | |
| 95 int Hunspell::add_dic(const char * dpath, const char * key) { | |
| 96 if (maxdic == MAXDIC || !affixpath) return 1; | |
| 97 pHMgr[maxdic] = new HashMgr(dpath, affixpath, key); | |
| 98 if (pHMgr[maxdic]) maxdic++; else return 1; | |
| 99 return 0; | |
| 100 } | |
| 101 #endif | |
| 102 | |
| 103 // make a copy of src at destination while removing all leading | |
| 104 // blanks and removing any trailing periods after recording | |
| 105 // their presence with the abbreviation flag | |
| 106 // also since already going through character by character, | |
| 107 // set the capitalization type | |
| 108 // return the length of the "cleaned" (and UTF-8 encoded) word | |
| 109 | |
| 110 int Hunspell::cleanword2(char * dest, const char * src, | |
| 111 w_char * dest_utf, int * nc, int * pcaptype, int * pabbrev) | |
| 112 { | |
| 113 unsigned char * p = (unsigned char *) dest; | |
| 114 const unsigned char * q = (const unsigned char * ) src; | |
| 115 | |
| 116 // first skip over any leading blanks | |
| 117 while ((*q != '\0') && (*q == ' ')) q++; | |
| 118 | |
| 119 // now strip off any trailing periods (recording their presence) | |
| 120 *pabbrev = 0; | |
| 121 int nl = strlen((const char *)q); | |
| 122 while ((nl > 0) && (*(q+nl-1)=='.')) { | |
| 123 nl--; | |
| 124 (*pabbrev)++; | |
| 125 } | |
| 126 | |
| 127 // if no characters are left it can't be capitalized | |
| 128 if (nl <= 0) { | |
| 129 *pcaptype = NOCAP; | |
| 130 *p = '\0'; | |
| 131 return 0; | |
| 132 } | |
| 133 | |
| 134 strncpy(dest, (char *) q, nl); | |
| 135 *(dest + nl) = '\0'; | |
| 136 nl = strlen(dest); | |
| 137 if (utf8) { | |
| 138 *nc = u8_u16(dest_utf, MAXWORDLEN, dest); | |
| 139 // don't check too long words | |
| 140 if (*nc >= MAXWORDLEN) return 0; | |
| 141 if (*nc == -1) { // big Unicode character (non BMP area) | |
| 142 *pcaptype = NOCAP; | |
| 143 return nl; | |
| 144 } | |
| 145 *pcaptype = get_captype_utf8(dest_utf, *nc, langnum); | |
| 146 } else { | |
| 147 *pcaptype = get_captype(dest, nl, csconv); | |
| 148 *nc = nl; | |
| 149 } | |
| 150 return nl; | |
| 151 } | |
| 152 | |
| 153 int Hunspell::cleanword(char * dest, const char * src, | |
| 154 int * pcaptype, int * pabbrev) | |
| 155 { | |
| 156 unsigned char * p = (unsigned char *) dest; | |
| 157 const unsigned char * q = (const unsigned char * ) src; | |
| 158 int firstcap = 0; | |
| 159 | |
| 160 // first skip over any leading blanks | |
| 161 while ((*q != '\0') && (*q == ' ')) q++; | |
| 162 | |
| 163 // now strip off any trailing periods (recording their presence) | |
| 164 *pabbrev = 0; | |
| 165 int nl = strlen((const char *)q); | |
| 166 while ((nl > 0) && (*(q+nl-1)=='.')) { | |
| 167 nl--; | |
| 168 (*pabbrev)++; | |
| 169 } | |
| 170 | |
| 171 // if no characters are left it can't be capitalized | |
| 172 if (nl <= 0) { | |
| 173 *pcaptype = NOCAP; | |
| 174 *p = '\0'; | |
| 175 return 0; | |
| 176 } | |
| 177 | |
| 178 // now determine the capitalization type of the first nl letters | |
| 179 int ncap = 0; | |
| 180 int nneutral = 0; | |
| 181 int nc = 0; | |
| 182 | |
| 183 if (!utf8) { | |
| 184 while (nl > 0) { | |
| 185 nc++; | |
| 186 if (csconv[(*q)].ccase) ncap++; | |
| 187 if (csconv[(*q)].cupper == csconv[(*q)].clower) nneutral++; | |
| 188 *p++ = *q++; | |
| 189 nl--; | |
| 190 } | |
| 191 // remember to terminate the destination string | |
| 192 *p = '\0'; | |
| 193 firstcap = csconv[(unsigned char)(*dest)].ccase; | |
| 194 } else { | |
| 195 unsigned short idx; | |
| 196 w_char t[MAXWORDLEN]; | |
| 197 nc = u8_u16(t, MAXWORDLEN, src); | |
| 198 for (int i = 0; i < nc; i++) { | |
| 199 idx = (t[i].h << 8) + t[i].l; | |
| 200 unsigned short low = unicodetolower(idx, langnum); | |
| 201 if (idx != low) ncap++; | |
| 202 if (unicodetoupper(idx, langnum) == low) nneutral++; | |
| 203 } | |
| 204 u16_u8(dest, MAXWORDUTF8LEN, t, nc); | |
| 205 if (ncap) { | |
| 206 idx = (t[0].h << 8) + t[0].l; | |
| 207 firstcap = (idx != unicodetolower(idx, langnum)); | |
| 208 } | |
| 209 } | |
| 210 | |
| 211 // now finally set the captype | |
| 212 if (ncap == 0) { | |
| 213 *pcaptype = NOCAP; | |
| 214 } else if ((ncap == 1) && firstcap) { | |
| 215 *pcaptype = INITCAP; | |
| 216 } else if ((ncap == nc) || ((ncap + nneutral) == nc)){ | |
| 217 *pcaptype = ALLCAP; | |
| 218 } else if ((ncap > 1) && firstcap) { | |
| 219 *pcaptype = HUHINITCAP; | |
| 220 } else { | |
| 221 *pcaptype = HUHCAP; | |
| 222 } | |
| 223 return strlen(dest); | |
| 224 } | |
| 225 | |
| 226 void Hunspell::mkallcap(char * p) | |
| 227 { | |
| 228 if (utf8) { | |
| 229 w_char u[MAXWORDLEN]; | |
| 230 int nc = u8_u16(u, MAXWORDLEN, p); | |
| 231 unsigned short idx; | |
| 232 for (int i = 0; i < nc; i++) { | |
| 233 idx = (u[i].h << 8) + u[i].l; | |
| 234 if (idx != unicodetoupper(idx, langnum)) { | |
| 235 u[i].h = (unsigned char) (unicodetoupper(idx, langnum) >> 8); | |
| 236 u[i].l = (unsigned char) (unicodetoupper(idx, langnum) & 0x00FF); | |
| 237 } | |
| 238 } | |
| 239 u16_u8(p, MAXWORDUTF8LEN, u, nc); | |
| 240 } else { | |
| 241 while (*p != '\0') { | |
| 242 *p = csconv[((unsigned char) *p)].cupper; | |
| 243 p++; | |
| 244 } | |
| 245 } | |
| 246 } | |
| 247 | |
| 248 int Hunspell::mkallcap2(char * p, w_char * u, int nc) | |
| 249 { | |
| 250 if (utf8) { | |
| 251 unsigned short idx; | |
| 252 for (int i = 0; i < nc; i++) { | |
| 253 idx = (u[i].h << 8) + u[i].l; | |
| 254 unsigned short up = unicodetoupper(idx, langnum); | |
| 255 if (idx != up) { | |
| 256 u[i].h = (unsigned char) (up >> 8); | |
| 257 u[i].l = (unsigned char) (up & 0x00FF); | |
| 258 } | |
| 259 } | |
| 260 u16_u8(p, MAXWORDUTF8LEN, u, nc); | |
| 261 return strlen(p); | |
| 262 } else { | |
| 263 while (*p != '\0') { | |
| 264 *p = csconv[((unsigned char) *p)].cupper; | |
| 265 p++; | |
| 266 } | |
| 267 } | |
| 268 return nc; | |
| 269 } | |
| 270 | |
| 271 | |
| 272 void Hunspell::mkallsmall(char * p) | |
| 273 { | |
| 274 while (*p != '\0') { | |
| 275 *p = csconv[((unsigned char) *p)].clower; | |
| 276 p++; | |
| 277 } | |
| 278 } | |
| 279 | |
| 280 int Hunspell::mkallsmall2(char * p, w_char * u, int nc) | |
| 281 { | |
| 282 if (utf8) { | |
| 283 unsigned short idx; | |
| 284 for (int i = 0; i < nc; i++) { | |
| 285 idx = (u[i].h << 8) + u[i].l; | |
| 286 unsigned short low = unicodetolower(idx, langnum); | |
| 287 if (idx != low) { | |
| 288 u[i].h = (unsigned char) (low >> 8); | |
| 289 u[i].l = (unsigned char) (low & 0x00FF); | |
| 290 } | |
| 291 } | |
| 292 u16_u8(p, MAXWORDUTF8LEN, u, nc); | |
| 293 return strlen(p); | |
| 294 } else { | |
| 295 while (*p != '\0') { | |
| 296 *p = csconv[((unsigned char) *p)].clower; | |
| 297 p++; | |
| 298 } | |
| 299 } | |
| 300 return nc; | |
| 301 } | |
| 302 | |
| 303 // convert UTF-8 sharp S codes to latin 1 | |
| 304 char * Hunspell::sharps_u8_l1(char * dest, char * source) { | |
| 305 char * p = dest; | |
| 306 *p = *source; | |
| 307 for (p++, source++; *(source - 1); p++, source++) { | |
| 308 *p = *source; | |
| 309 if (*source == '\x9F') *--p = '\xDF'; | |
| 310 } | |
| 311 return dest; | |
| 312 } | |
| 313 | |
| 314 // recursive search for right ss - sharp s permutations | |
| 315 hentry * Hunspell::spellsharps(char * base, char * pos, int n, | |
| 316 int repnum, char * tmp, int * info, char **root) { | |
| 317 pos = strstr(pos, "ss"); | |
| 318 if (pos && (n < MAXSHARPS)) { | |
| 319 *pos = '\xC3'; | |
| 320 *(pos + 1) = '\x9F'; | |
| 321 hentry * h = spellsharps(base, pos + 2, n + 1, repnum + 1, tmp, info, ro
ot); | |
| 322 if (h) return h; | |
| 323 *pos = 's'; | |
| 324 *(pos + 1) = 's'; | |
| 325 h = spellsharps(base, pos + 2, n + 1, repnum, tmp, info, root); | |
| 326 if (h) return h; | |
| 327 } else if (repnum > 0) { | |
| 328 if (utf8) return checkword(base, info, root); | |
| 329 return checkword(sharps_u8_l1(tmp, base), info, root); | |
| 330 } | |
| 331 return NULL; | |
| 332 } | |
| 333 | |
| 334 int Hunspell::is_keepcase(const hentry * rv) { | |
| 335 return pAMgr && rv->astr && pAMgr->get_keepcase() && | |
| 336 TESTAFF(rv->astr, pAMgr->get_keepcase(), rv->alen); | |
| 337 } | |
| 338 | |
| 339 /* insert a word to the beginning of the suggestion array and return ns */ | |
| 340 int Hunspell::insert_sug(char ***slst, char * word, int ns) { | |
| 341 char * dup = mystrdup(word); | |
| 342 if (!dup) return ns; | |
| 343 if (ns == MAXSUGGESTION) { | |
| 344 ns--; | |
| 345 free((*slst)[ns]); | |
| 346 } | |
| 347 for (int k = ns; k > 0; k--) (*slst)[k] = (*slst)[k - 1]; | |
| 348 (*slst)[0] = dup; | |
| 349 return ns + 1; | |
| 350 } | |
| 351 | |
| 352 int Hunspell::spell(const char * word, int * info, char ** root) | |
| 353 { | |
| 354 #ifdef HUNSPELL_CHROME_CLIENT | |
| 355 if (pHMgr[0]) pHMgr[0]->EmptyHentryCache(); | |
| 356 #endif | |
| 357 struct hentry * rv=NULL; | |
| 358 // need larger vector. For example, Turkish capital letter I converted a | |
| 359 // 2-byte UTF-8 character (dotless i) by mkallsmall. | |
| 360 char cw[MAXWORDUTF8LEN]; | |
| 361 char wspace[MAXWORDUTF8LEN]; | |
| 362 w_char unicw[MAXWORDLEN]; | |
| 363 // Hunspell supports XML input of the simplified API (see manual) | |
| 364 if (strcmp(word, SPELL_XML) == 0) return 1; | |
| 365 int nc = strlen(word); | |
| 366 int wl2 = 0; | |
| 367 if (utf8) { | |
| 368 if (nc >= MAXWORDUTF8LEN) return 0; | |
| 369 } else { | |
| 370 if (nc >= MAXWORDLEN) return 0; | |
| 371 } | |
| 372 int captype = 0; | |
| 373 int abbv = 0; | |
| 374 int wl = 0; | |
| 375 | |
| 376 // input conversion | |
| 377 RepList * rl = (pAMgr) ? pAMgr->get_iconvtable() : NULL; | |
| 378 if (rl && rl->conv(word, wspace)) wl = cleanword2(cw, wspace, unicw, &nc, &cap
type, &abbv); | |
| 379 else wl = cleanword2(cw, word, unicw, &nc, &captype, &abbv); | |
| 380 | |
| 381 int info2 = 0; | |
| 382 if (wl == 0 || maxdic == 0) return 1; | |
| 383 if (root) *root = NULL; | |
| 384 | |
| 385 // allow numbers with dots, dashes and commas (but forbid double separators: "
..", "--" etc.) | |
| 386 enum { NBEGIN, NNUM, NSEP }; | |
| 387 int nstate = NBEGIN; | |
| 388 int i; | |
| 389 | |
| 390 for (i = 0; (i < wl); i++) { | |
| 391 if ((cw[i] <= '9') && (cw[i] >= '0')) { | |
| 392 nstate = NNUM; | |
| 393 } else if ((cw[i] == ',') || (cw[i] == '.') || (cw[i] == '-')) { | |
| 394 if ((nstate == NSEP) || (i == 0)) break; | |
| 395 nstate = NSEP; | |
| 396 } else break; | |
| 397 } | |
| 398 if ((i == wl) && (nstate == NNUM)) return 1; | |
| 399 if (!info) info = &info2; else *info = 0; | |
| 400 | |
| 401 switch(captype) { | |
| 402 case HUHCAP: | |
| 403 case HUHINITCAP: | |
| 404 *info += SPELL_ORIGCAP; | |
| 405 case NOCAP: { | |
| 406 rv = checkword(cw, info, root); | |
| 407 if ((abbv) && !(rv)) { | |
| 408 memcpy(wspace,cw,wl); | |
| 409 *(wspace+wl) = '.'; | |
| 410 *(wspace+wl+1) = '\0'; | |
| 411 rv = checkword(wspace, info, root); | |
| 412 } | |
| 413 break; | |
| 414 } | |
| 415 case ALLCAP: { | |
| 416 *info += SPELL_ORIGCAP; | |
| 417 rv = checkword(cw, info, root); | |
| 418 if (rv) break; | |
| 419 if (abbv) { | |
| 420 memcpy(wspace,cw,wl); | |
| 421 *(wspace+wl) = '.'; | |
| 422 *(wspace+wl+1) = '\0'; | |
| 423 rv = checkword(wspace, info, root); | |
| 424 if (rv) break; | |
| 425 } | |
| 426 // Spec. prefix handling for Catalan, French, Italian: | |
| 427 // prefixes separated by apostrophe (SANT'ELIA -> Sant'+Elia). | |
| 428 if (pAMgr && strchr(cw, '\'')) { | |
| 429 wl = mkallsmall2(cw, unicw, nc); | |
| 430 //There are no really sane circumstances where this could fail, | |
| 431 //but anyway... | |
| 432 if (char * apostrophe = strchr(cw, '\'')) { | |
| 433 if (utf8) { | |
| 434 w_char tmpword[MAXWORDLEN]; | |
| 435 *apostrophe = '\0'; | |
| 436 wl2 = u8_u16(tmpword, MAXWORDLEN, cw); | |
| 437 *apostrophe = '\''; | |
| 438 if (wl2 < nc) { | |
| 439 mkinitcap2(apostrophe + 1, unicw + wl2 + 1, nc - wl2
- 1); | |
| 440 rv = checkword(cw, info, root); | |
| 441 if (rv) break; | |
| 442 } | |
| 443 } else { | |
| 444 mkinitcap2(apostrophe + 1, unicw, nc); | |
| 445 rv = checkword(cw, info, root); | |
| 446 if (rv) break; | |
| 447 } | |
| 448 } | |
| 449 mkinitcap2(cw, unicw, nc); | |
| 450 rv = checkword(cw, info, root); | |
| 451 if (rv) break; | |
| 452 } | |
| 453 if (pAMgr && pAMgr->get_checksharps() && strstr(cw, "SS")) { | |
| 454 char tmpword[MAXWORDUTF8LEN]; | |
| 455 wl = mkallsmall2(cw, unicw, nc); | |
| 456 memcpy(wspace,cw,(wl+1)); | |
| 457 rv = spellsharps(wspace, wspace, 0, 0, tmpword, info, root); | |
| 458 if (!rv) { | |
| 459 wl2 = mkinitcap2(cw, unicw, nc); | |
| 460 rv = spellsharps(cw, cw, 0, 0, tmpword, info, root); | |
| 461 } | |
| 462 if ((abbv) && !(rv)) { | |
| 463 *(wspace+wl) = '.'; | |
| 464 *(wspace+wl+1) = '\0'; | |
| 465 rv = spellsharps(wspace, wspace, 0, 0, tmpword, info, root); | |
| 466 if (!rv) { | |
| 467 memcpy(wspace, cw, wl2); | |
| 468 *(wspace+wl2) = '.'; | |
| 469 *(wspace+wl2+1) = '\0'; | |
| 470 rv = spellsharps(wspace, wspace, 0, 0, tmpword, info, ro
ot); | |
| 471 } | |
| 472 } | |
| 473 if (rv) break; | |
| 474 } | |
| 475 } | |
| 476 case INITCAP: { | |
| 477 *info += SPELL_ORIGCAP; | |
| 478 wl = mkallsmall2(cw, unicw, nc); | |
| 479 memcpy(wspace,cw,(wl+1)); | |
| 480 wl2 = mkinitcap2(cw, unicw, nc); | |
| 481 if (captype == INITCAP) *info += SPELL_INITCAP; | |
| 482 rv = checkword(cw, info, root); | |
| 483 if (captype == INITCAP) *info -= SPELL_INITCAP; | |
| 484 // forbid bad capitalization | |
| 485 // (for example, ijs -> Ijs instead of IJs in Dutch) | |
| 486 // use explicit forms in dic: Ijs/F (F = FORBIDDENWORD flag) | |
| 487 if (*info & SPELL_FORBIDDEN) { | |
| 488 rv = NULL; | |
| 489 break; | |
| 490 } | |
| 491 if (rv && is_keepcase(rv) && (captype == ALLCAP)) rv = NULL; | |
| 492 if (rv) break; | |
| 493 | |
| 494 rv = checkword(wspace, info, root); | |
| 495 if (abbv && !rv) { | |
| 496 | |
| 497 *(wspace+wl) = '.'; | |
| 498 *(wspace+wl+1) = '\0'; | |
| 499 rv = checkword(wspace, info, root); | |
| 500 if (!rv) { | |
| 501 memcpy(wspace, cw, wl2); | |
| 502 *(wspace+wl2) = '.'; | |
| 503 *(wspace+wl2+1) = '\0'; | |
| 504 if (captype == INITCAP) *info += SPELL_INITCAP; | |
| 505 rv = checkword(wspace, info, root); | |
| 506 if (captype == INITCAP) *info -= SPELL_INITCAP; | |
| 507 if (rv && is_keepcase(rv) && (captype == ALLCAP)) rv = NULL; | |
| 508 break; | |
| 509 } | |
| 510 } | |
| 511 if (rv && is_keepcase(rv) && | |
| 512 ((captype == ALLCAP) || | |
| 513 // if CHECKSHARPS: KEEPCASE words with \xDF are allowed | |
| 514 // in INITCAP form, too. | |
| 515 !(pAMgr->get_checksharps() && | |
| 516 ((utf8 && strstr(wspace, "\xC3\x9F")) || | |
| 517 (!utf8 && strchr(wspace, '\xDF')))))) rv = NULL; | |
| 518 break; | |
| 519 } | |
| 520 } | |
| 521 | |
| 522 if (rv) { | |
| 523 if (pAMgr && pAMgr->get_warn() && rv->astr && | |
| 524 TESTAFF(rv->astr, pAMgr->get_warn(), rv->alen)) { | |
| 525 *info += SPELL_WARN; | |
| 526 if (pAMgr->get_forbidwarn()) return 0; | |
| 527 return HUNSPELL_OK_WARN; | |
| 528 } | |
| 529 return HUNSPELL_OK; | |
| 530 } | |
| 531 | |
| 532 // recursive breaking at break points | |
| 533 if (wordbreak) { | |
| 534 char * s; | |
| 535 char r; | |
| 536 int nbr = 0; | |
| 537 wl = strlen(cw); | |
| 538 int numbreak = pAMgr ? pAMgr->get_numbreak() : 0; | |
| 539 | |
| 540 // calculate break points for recursion limit | |
| 541 for (int j = 0; j < numbreak; j++) { | |
| 542 s = cw; | |
| 543 do { | |
| 544 s = (char *) strstr(s, wordbreak[j]); | |
| 545 if (s) { | |
| 546 nbr++; | |
| 547 s++; | |
| 548 } | |
| 549 } while (s); | |
| 550 } | |
| 551 if (nbr >= 10) return 0; | |
| 552 | |
| 553 // check boundary patterns (^begin and end$) | |
| 554 for (int j = 0; j < numbreak; j++) { | |
| 555 int plen = strlen(wordbreak[j]); | |
| 556 if (plen == 1 || plen > wl) continue; | |
| 557 if (wordbreak[j][0] == '^' && strncmp(cw, wordbreak[j] + 1, plen - 1) == 0 | |
| 558 && spell(cw + plen - 1)) return 1; | |
| 559 if (wordbreak[j][plen - 1] == '$' && | |
| 560 strncmp(cw + wl - plen + 1, wordbreak[j], plen - 1) == 0) { | |
| 561 r = cw[wl - plen + 1]; | |
| 562 cw[wl - plen + 1] = '\0'; | |
| 563 if (spell(cw)) return 1; | |
| 564 cw[wl - plen + 1] = r; | |
| 565 } | |
| 566 } | |
| 567 | |
| 568 // other patterns | |
| 569 for (int j = 0; j < numbreak; j++) { | |
| 570 int plen = strlen(wordbreak[j]); | |
| 571 s=(char *) strstr(cw, wordbreak[j]); | |
| 572 if (s && (s > cw) && (s < cw + wl - plen)) { | |
| 573 if (!spell(s + plen)) continue; | |
| 574 r = *s; | |
| 575 *s = '\0'; | |
| 576 // examine 2 sides of the break point | |
| 577 if (spell(cw)) return 1; | |
| 578 *s = r; | |
| 579 | |
| 580 // LANG_hu: spec. dash rule | |
| 581 if (langnum == LANG_hu && strcmp(wordbreak[j], "-") == 0) { | |
| 582 r = s[1]; | |
| 583 s[1] = '\0'; | |
| 584 if (spell(cw)) return 1; // check the first part with dash | |
| 585 s[1] = r; | |
| 586 } | |
| 587 // end of LANG speficic region | |
| 588 | |
| 589 } | |
| 590 } | |
| 591 } | |
| 592 | |
| 593 return 0; | |
| 594 } | |
| 595 | |
| 596 struct hentry * Hunspell::checkword(const char * w, int * info, char ** root) | |
| 597 { | |
| 598 struct hentry * he = NULL; | |
| 599 int len, i; | |
| 600 char w2[MAXWORDUTF8LEN]; | |
| 601 const char * word; | |
| 602 | |
| 603 char * ignoredchars = pAMgr->get_ignore(); | |
| 604 if (ignoredchars != NULL) { | |
| 605 strcpy(w2, w); | |
| 606 if (utf8) { | |
| 607 int ignoredchars_utf16_len; | |
| 608 unsigned short * ignoredchars_utf16 = pAMgr->get_ignore_utf16(&ignoredch
ars_utf16_len); | |
| 609 remove_ignored_chars_utf(w2, ignoredchars_utf16, ignoredchars_utf16_len)
; | |
| 610 } else { | |
| 611 remove_ignored_chars(w2,ignoredchars); | |
| 612 } | |
| 613 word = w2; | |
| 614 } else word = w; | |
| 615 | |
| 616 len = strlen(word); | |
| 617 | |
| 618 if (!len) | |
| 619 return NULL; | |
| 620 | |
| 621 #ifdef HUNSPELL_CHROME_CLIENT | |
| 622 // We need to check if the word length is valid to make coverity (Event | |
| 623 // fixed_size_dest: Possible overrun of N byte fixed size buffer) happy. | |
| 624 if ((utf8 && strlen(word) >= MAXWORDUTF8LEN) || (!utf8 && strlen(word) >= MAXW
ORDLEN)) | |
| 625 return NULL; | |
| 626 #endif | |
| 627 | |
| 628 // word reversing wrapper for complex prefixes | |
| 629 if (complexprefixes) { | |
| 630 if (word != w2) { | |
| 631 strcpy(w2, word); | |
| 632 word = w2; | |
| 633 } | |
| 634 if (utf8) reverseword_utf(w2); else reverseword(w2); | |
| 635 } | |
| 636 | |
| 637 // look word in hash table | |
| 638 for (i = 0; (i < maxdic) && !he; i ++) { | |
| 639 he = (pHMgr[i])->lookup(word); | |
| 640 | |
| 641 // check forbidden and onlyincompound words | |
| 642 if ((he) && (he->astr) && (pAMgr) && TESTAFF(he->astr, pAMgr->get_forbiddenwor
d(), he->alen)) { | |
| 643 if (info) *info += SPELL_FORBIDDEN; | |
| 644 // LANG_hu section: set dash information for suggestions | |
| 645 if (langnum == LANG_hu) { | |
| 646 if (pAMgr->get_compoundflag() && | |
| 647 TESTAFF(he->astr, pAMgr->get_compoundflag(), he->alen)) { | |
| 648 if (info) *info += SPELL_COMPOUND; | |
| 649 } | |
| 650 } | |
| 651 return NULL; | |
| 652 } | |
| 653 | |
| 654 // he = next not needaffix, onlyincompound homonym or onlyupcase word | |
| 655 while (he && (he->astr) && | |
| 656 ((pAMgr->get_needaffix() && TESTAFF(he->astr, pAMgr->get_needaffix(), he->al
en)) || | |
| 657 (pAMgr->get_onlyincompound() && TESTAFF(he->astr, pAMgr->get_onlyincompou
nd(), he->alen)) || | |
| 658 (info && (*info & SPELL_INITCAP) && TESTAFF(he->astr, ONLYUPCASEFLAG, he-
>alen)) | |
| 659 )) he = he->next_homonym; | |
| 660 } | |
| 661 | |
| 662 // check with affixes | |
| 663 if (!he && pAMgr) { | |
| 664 // try stripping off affixes */ | |
| 665 he = pAMgr->affix_check(word, len, 0); | |
| 666 | |
| 667 // check compound restriction and onlyupcase | |
| 668 if (he && he->astr && ( | |
| 669 (pAMgr->get_onlyincompound() && | |
| 670 TESTAFF(he->astr, pAMgr->get_onlyincompound(), he->alen)) || | |
| 671 (info && (*info & SPELL_INITCAP) && | |
| 672 TESTAFF(he->astr, ONLYUPCASEFLAG, he->alen)))) { | |
| 673 he = NULL; | |
| 674 } | |
| 675 | |
| 676 if (he) { | |
| 677 if ((he->astr) && (pAMgr) && TESTAFF(he->astr, pAMgr->get_forbiddenword(
), he->alen)) { | |
| 678 if (info) *info += SPELL_FORBIDDEN; | |
| 679 return NULL; | |
| 680 } | |
| 681 if (root) { | |
| 682 *root = mystrdup(he->word); | |
| 683 if (*root && complexprefixes) { | |
| 684 if (utf8) reverseword_utf(*root); else reverseword(*root); | |
| 685 } | |
| 686 } | |
| 687 // try check compound word | |
| 688 } else if (pAMgr->get_compound()) { | |
| 689 he = pAMgr->compound_check(word, len, 0, 0, 100, 0, NULL, 0, 0, info); | |
| 690 // LANG_hu section: `moving rule' with last dash | |
| 691 if ((!he) && (langnum == LANG_hu) && (word[len-1] == '-')) { | |
| 692 char * dup = mystrdup(word); | |
| 693 if (!dup) return NULL; | |
| 694 dup[len-1] = '\0'; | |
| 695 he = pAMgr->compound_check(dup, len-1, -5, 0, 100, 0, NULL, 1, 0, i
nfo); | |
| 696 free(dup); | |
| 697 } | |
| 698 // end of LANG speficic region | |
| 699 if (he) { | |
| 700 if (root) { | |
| 701 *root = mystrdup(he->word); | |
| 702 if (*root && complexprefixes) { | |
| 703 if (utf8) reverseword_utf(*root); else reverseword(*root
); | |
| 704 } | |
| 705 } | |
| 706 if (info) *info += SPELL_COMPOUND; | |
| 707 } | |
| 708 } | |
| 709 | |
| 710 } | |
| 711 | |
| 712 return he; | |
| 713 } | |
| 714 | |
| 715 int Hunspell::suggest(char*** slst, const char * word) | |
| 716 { | |
| 717 #ifdef HUNSPELL_CHROME_CLIENT | |
| 718 if (pHMgr[0]) pHMgr[0]->EmptyHentryCache(); | |
| 719 #endif | |
| 720 int onlycmpdsug = 0; | |
| 721 char cw[MAXWORDUTF8LEN]; | |
| 722 char wspace[MAXWORDUTF8LEN]; | |
| 723 if (!pSMgr || maxdic == 0) return 0; | |
| 724 w_char unicw[MAXWORDLEN]; | |
| 725 *slst = NULL; | |
| 726 // process XML input of the simplified API (see manual) | |
| 727 if (strncmp(word, SPELL_XML, sizeof(SPELL_XML) - 3) == 0) { | |
| 728 return spellml(slst, word); | |
| 729 } | |
| 730 int nc = strlen(word); | |
| 731 if (utf8) { | |
| 732 if (nc >= MAXWORDUTF8LEN) return 0; | |
| 733 } else { | |
| 734 if (nc >= MAXWORDLEN) return 0; | |
| 735 } | |
| 736 int captype = 0; | |
| 737 int abbv = 0; | |
| 738 int wl = 0; | |
| 739 | |
| 740 // input conversion | |
| 741 RepList * rl = (pAMgr) ? pAMgr->get_iconvtable() : NULL; | |
| 742 if (rl && rl->conv(word, wspace)) wl = cleanword2(cw, wspace, unicw, &nc, &cap
type, &abbv); | |
| 743 else wl = cleanword2(cw, word, unicw, &nc, &captype, &abbv); | |
| 744 | |
| 745 if (wl == 0) return 0; | |
| 746 int ns = 0; | |
| 747 int capwords = 0; | |
| 748 | |
| 749 // check capitalized form for FORCEUCASE | |
| 750 if (pAMgr && captype == NOCAP && pAMgr->get_forceucase()) { | |
| 751 int info = SPELL_ORIGCAP; | |
| 752 char ** wlst; | |
| 753 if (checkword(cw, &info, NULL)) { | |
| 754 if (*slst) { | |
| 755 wlst = *slst; | |
| 756 } else { | |
| 757 wlst = (char **) malloc(MAXSUGGESTION * sizeof(char *)); | |
| 758 if (wlst == NULL) return -1; | |
| 759 *slst = wlst; | |
| 760 for (int i = 0; i < MAXSUGGESTION; i++) { | |
| 761 wlst[i] = NULL; | |
| 762 } | |
| 763 } | |
| 764 wlst[0] = mystrdup(cw); | |
| 765 mkinitcap(wlst[0]); | |
| 766 return 1; | |
| 767 } | |
| 768 } | |
| 769 | |
| 770 switch(captype) { | |
| 771 case NOCAP: { | |
| 772 ns = pSMgr->suggest(slst, cw, ns, &onlycmpdsug); | |
| 773 break; | |
| 774 } | |
| 775 | |
| 776 case INITCAP: { | |
| 777 capwords = 1; | |
| 778 ns = pSMgr->suggest(slst, cw, ns, &onlycmpdsug); | |
| 779 if (ns == -1) break; | |
| 780 memcpy(wspace,cw,(wl+1)); | |
| 781 mkallsmall2(wspace, unicw, nc); | |
| 782 ns = pSMgr->suggest(slst, wspace, ns, &onlycmpdsug); | |
| 783 break; | |
| 784 } | |
| 785 case HUHINITCAP: | |
| 786 capwords = 1; | |
| 787 case HUHCAP: { | |
| 788 ns = pSMgr->suggest(slst, cw, ns, &onlycmpdsug); | |
| 789 if (ns != -1) { | |
| 790 int prevns; | |
| 791 // something.The -> something. The | |
| 792 char * dot = strchr(cw, '.'); | |
| 793 if (dot && (dot > cw)) { | |
| 794 int captype_; | |
| 795 if (utf8) { | |
| 796 w_char w_[MAXWORDLEN]; | |
| 797 int wl_ = u8_u16(w_, MAXWORDLEN, dot + 1); | |
| 798 captype_ = get_captype_utf8(w_, wl_, langnum); | |
| 799 } else captype_ = get_captype(dot+1, strlen(dot+1),
csconv); | |
| 800 if (captype_ == INITCAP) { | |
| 801 char * st = mystrdup(cw); | |
| 802 if (st) st = (char *) realloc(st, wl + 2); | |
| 803 if (st) { | |
| 804 st[(dot - cw) + 1] = ' '; | |
| 805 strcpy(st + (dot - cw) + 2, dot + 1); | |
| 806 ns = insert_sug(slst, st, ns); | |
| 807 free(st); | |
| 808 } | |
| 809 } | |
| 810 } | |
| 811 if (captype == HUHINITCAP) { | |
| 812 // TheOpenOffice.org -> The OpenOffice.org | |
| 813 memcpy(wspace,cw,(wl+1)); | |
| 814 mkinitsmall2(wspace, unicw, nc); | |
| 815 ns = pSMgr->suggest(slst, wspace, ns, &onlycmpdsug); | |
| 816 } | |
| 817 memcpy(wspace,cw,(wl+1)); | |
| 818 mkallsmall2(wspace, unicw, nc); | |
| 819 if (spell(wspace)) ns = insert_sug(slst, wspace, ns); | |
| 820 prevns = ns; | |
| 821 ns = pSMgr->suggest(slst, wspace, ns, &onlycmpdsug); | |
| 822 if (captype == HUHINITCAP) { | |
| 823 mkinitcap2(wspace, unicw, nc); | |
| 824 if (spell(wspace)) ns = insert_sug(slst, wspace, ns)
; | |
| 825 ns = pSMgr->suggest(slst, wspace, ns, &onlycmpdsug); | |
| 826 } | |
| 827 // aNew -> "a New" (instead of "a new") | |
| 828 for (int j = prevns; j < ns; j++) { | |
| 829 char * space = strchr((*slst)[j],' '); | |
| 830 if (space) { | |
| 831 int slen = strlen(space + 1); | |
| 832 // different case after space (need capitalisati
on) | |
| 833 if ((slen < wl) && strcmp(cw + wl - slen, space
+ 1)) { | |
| 834 w_char w[MAXWORDLEN]; | |
| 835 int wc = 0; | |
| 836 char * r = (*slst)[j]; | |
| 837 if (utf8) wc = u8_u16(w, MAXWORDLEN, space +
1); | |
| 838 mkinitcap2(space + 1, w, wc); | |
| 839 // set as first suggestion | |
| 840 for (int k = j; k > 0; k--) (*slst)[k] = (*s
lst)[k - 1]; | |
| 841 (*slst)[0] = r; | |
| 842 } | |
| 843 } | |
| 844 } | |
| 845 } | |
| 846 break; | |
| 847 } | |
| 848 | |
| 849 case ALLCAP: { | |
| 850 memcpy(wspace, cw, (wl+1)); | |
| 851 mkallsmall2(wspace, unicw, nc); | |
| 852 ns = pSMgr->suggest(slst, wspace, ns, &onlycmpdsug); | |
| 853 if (ns == -1) break; | |
| 854 if (pAMgr && pAMgr->get_keepcase() && spell(wspace)) | |
| 855 ns = insert_sug(slst, wspace, ns); | |
| 856 mkinitcap2(wspace, unicw, nc); | |
| 857 ns = pSMgr->suggest(slst, wspace, ns, &onlycmpdsug); | |
| 858 for (int j=0; j < ns; j++) { | |
| 859 mkallcap((*slst)[j]); | |
| 860 if (pAMgr && pAMgr->get_checksharps()) { | |
| 861 char * pos; | |
| 862 if (utf8) { | |
| 863 pos = strstr((*slst)[j], "\xC3\x9F"); | |
| 864 while (pos) { | |
| 865 *pos = 'S'; | |
| 866 *(pos+1) = 'S'; | |
| 867 pos = strstr(pos+2, "\xC3\x9F"); | |
| 868 } | |
| 869 } else { | |
| 870 pos = strchr((*slst)[j], '\xDF'); | |
| 871 while (pos) { | |
| 872 (*slst)[j] = (char *) realloc((*slst)[j], st
rlen((*slst)[j]) + 2); | |
| 873 mystrrep((*slst)[j], "\xDF", "SS"); | |
| 874 pos = strchr((*slst)[j], '\xDF'); | |
| 875 } | |
| 876 } | |
| 877 } | |
| 878 } | |
| 879 break; | |
| 880 } | |
| 881 } | |
| 882 | |
| 883 // LANG_hu section: replace '-' with ' ' in Hungarian | |
| 884 if (langnum == LANG_hu) { | |
| 885 for (int j=0; j < ns; j++) { | |
| 886 char * pos = strchr((*slst)[j],'-'); | |
| 887 if (pos) { | |
| 888 int info; | |
| 889 char w[MAXWORDUTF8LEN]; | |
| 890 *pos = '\0'; | |
| 891 strcpy(w, (*slst)[j]); | |
| 892 strcat(w, pos + 1); | |
| 893 spell(w, &info, NULL); | |
| 894 if ((info & SPELL_COMPOUND) && (info & SPELL_FORBIDDEN)) { | |
| 895 *pos = ' '; | |
| 896 } else *pos = '-'; | |
| 897 } | |
| 898 } | |
| 899 } | |
| 900 // END OF LANG_hu section | |
| 901 | |
| 902 // try ngram approach since found nothing or only compound words | |
| 903 if (pAMgr && (ns == 0 || onlycmpdsug) && (pAMgr->get_maxngramsugs() != 0) && (
*slst)) { | |
| 904 switch(captype) { | |
| 905 case NOCAP: { | |
| 906 ns = pSMgr->ngsuggest(*slst, cw, ns, pHMgr, maxdic); | |
| 907 break; | |
| 908 } | |
| 909 case HUHINITCAP: | |
| 910 capwords = 1; | |
| 911 case HUHCAP: { | |
| 912 memcpy(wspace,cw,(wl+1)); | |
| 913 mkallsmall2(wspace, unicw, nc); | |
| 914 ns = pSMgr->ngsuggest(*slst, wspace, ns, pHMgr, maxdic); | |
| 915 break; | |
| 916 } | |
| 917 case INITCAP: { | |
| 918 capwords = 1; | |
| 919 memcpy(wspace,cw,(wl+1)); | |
| 920 mkallsmall2(wspace, unicw, nc); | |
| 921 ns = pSMgr->ngsuggest(*slst, wspace, ns, pHMgr, maxdic); | |
| 922 break; | |
| 923 } | |
| 924 case ALLCAP: { | |
| 925 memcpy(wspace,cw,(wl+1)); | |
| 926 mkallsmall2(wspace, unicw, nc); | |
| 927 int oldns = ns; | |
| 928 ns = pSMgr->ngsuggest(*slst, wspace, ns, pHMgr, maxdic); | |
| 929 for (int j = oldns; j < ns; j++) | |
| 930 mkallcap((*slst)[j]); | |
| 931 break; | |
| 932 } | |
| 933 } | |
| 934 } | |
| 935 | |
| 936 // try dash suggestion (Afo-American -> Afro-American) | |
| 937 if (char * pos = strchr(cw, '-')) { | |
| 938 char * ppos = cw; | |
| 939 int nodashsug = 1; | |
| 940 char ** nlst = NULL; | |
| 941 int nn = 0; | |
| 942 int last = 0; | |
| 943 if (*slst) { | |
| 944 for (int j = 0; j < ns && nodashsug == 1; j++) { | |
| 945 if (strchr((*slst)[j], '-')) nodashsug = 0; | |
| 946 } | |
| 947 } | |
| 948 while (nodashsug && !last) { | |
| 949 if (*pos == '\0') last = 1; else *pos = '\0'; | |
| 950 if (!spell(ppos)) { | |
| 951 nn = suggest(&nlst, ppos); | |
| 952 for (int j = nn - 1; j >= 0; j--) { | |
| 953 strncpy(wspace, cw, ppos - cw); | |
| 954 strcpy(wspace + (ppos - cw), nlst[j]); | |
| 955 if (!last) { | |
| 956 strcat(wspace, "-"); | |
| 957 strcat(wspace, pos + 1); | |
| 958 } | |
| 959 ns = insert_sug(slst, wspace, ns); | |
| 960 free(nlst[j]); | |
| 961 } | |
| 962 if (nlst != NULL) free(nlst); | |
| 963 nodashsug = 0; | |
| 964 } | |
| 965 if (!last) { | |
| 966 *pos = '-'; | |
| 967 ppos = pos + 1; | |
| 968 pos = strchr(ppos, '-'); | |
| 969 } | |
| 970 if (!pos) pos = cw + strlen(cw); | |
| 971 } | |
| 972 } | |
| 973 | |
| 974 // word reversing wrapper for complex prefixes | |
| 975 if (complexprefixes) { | |
| 976 for (int j = 0; j < ns; j++) { | |
| 977 if (utf8) reverseword_utf((*slst)[j]); else reverseword((*slst)[j]); | |
| 978 } | |
| 979 } | |
| 980 | |
| 981 // capitalize | |
| 982 if (capwords) for (int j=0; j < ns; j++) { | |
| 983 mkinitcap((*slst)[j]); | |
| 984 } | |
| 985 | |
| 986 // expand suggestions with dot(s) | |
| 987 if (abbv && pAMgr && pAMgr->get_sugswithdots()) { | |
| 988 for (int j = 0; j < ns; j++) { | |
| 989 (*slst)[j] = (char *) realloc((*slst)[j], strlen((*slst)[j]) + 1 + abbv); | |
| 990 strcat((*slst)[j], word + strlen(word) - abbv); | |
| 991 } | |
| 992 } | |
| 993 | |
| 994 // remove bad capitalized and forbidden forms | |
| 995 if (pAMgr && (pAMgr->get_keepcase() || pAMgr->get_forbiddenword())) { | |
| 996 switch (captype) { | |
| 997 case INITCAP: | |
| 998 case ALLCAP: { | |
| 999 int l = 0; | |
| 1000 for (int j=0; j < ns; j++) { | |
| 1001 if (!strchr((*slst)[j],' ') && !spell((*slst)[j])) { | |
| 1002 char s[MAXSWUTF8L]; | |
| 1003 w_char w[MAXSWL]; | |
| 1004 int len; | |
| 1005 if (utf8) { | |
| 1006 len = u8_u16(w, MAXSWL, (*slst)[j]); | |
| 1007 } else { | |
| 1008 strcpy(s, (*slst)[j]); | |
| 1009 len = strlen(s); | |
| 1010 } | |
| 1011 mkallsmall2(s, w, len); | |
| 1012 free((*slst)[j]); | |
| 1013 if (spell(s)) { | |
| 1014 (*slst)[l] = mystrdup(s); | |
| 1015 if ((*slst)[l]) l++; | |
| 1016 } else { | |
| 1017 mkinitcap2(s, w, len); | |
| 1018 if (spell(s)) { | |
| 1019 (*slst)[l] = mystrdup(s); | |
| 1020 if ((*slst)[l]) l++; | |
| 1021 } | |
| 1022 } | |
| 1023 } else { | |
| 1024 (*slst)[l] = (*slst)[j]; | |
| 1025 l++; | |
| 1026 } | |
| 1027 } | |
| 1028 ns = l; | |
| 1029 } | |
| 1030 } | |
| 1031 } | |
| 1032 | |
| 1033 // remove duplications | |
| 1034 int l = 0; | |
| 1035 for (int j = 0; j < ns; j++) { | |
| 1036 (*slst)[l] = (*slst)[j]; | |
| 1037 for (int k = 0; k < l; k++) { | |
| 1038 if (strcmp((*slst)[k], (*slst)[j]) == 0) { | |
| 1039 free((*slst)[j]); | |
| 1040 l--; | |
| 1041 break; | |
| 1042 } | |
| 1043 } | |
| 1044 l++; | |
| 1045 } | |
| 1046 ns = l; | |
| 1047 | |
| 1048 // output conversion | |
| 1049 rl = (pAMgr) ? pAMgr->get_oconvtable() : NULL; | |
| 1050 for (int j = 0; rl && j < ns; j++) { | |
| 1051 if (rl->conv((*slst)[j], wspace)) { | |
| 1052 free((*slst)[j]); | |
| 1053 (*slst)[j] = mystrdup(wspace); | |
| 1054 } | |
| 1055 } | |
| 1056 | |
| 1057 // if suggestions removed by nosuggest, onlyincompound parameters | |
| 1058 if (l == 0 && *slst) { | |
| 1059 free(*slst); | |
| 1060 *slst = NULL; | |
| 1061 } | |
| 1062 return l; | |
| 1063 } | |
| 1064 | |
| 1065 void Hunspell::free_list(char *** slst, int n) { | |
| 1066 freelist(slst, n); | |
| 1067 } | |
| 1068 | |
| 1069 char * Hunspell::get_dic_encoding() | |
| 1070 { | |
| 1071 return encoding; | |
| 1072 } | |
| 1073 | |
| 1074 #ifdef HUNSPELL_EXPERIMENTAL | |
| 1075 // XXX need UTF-8 support | |
| 1076 int Hunspell::suggest_auto(char*** slst, const char * word) | |
| 1077 { | |
| 1078 char cw[MAXWORDUTF8LEN]; | |
| 1079 char wspace[MAXWORDUTF8LEN]; | |
| 1080 if (!pSMgr || maxdic == 0) return 0; | |
| 1081 int wl = strlen(word); | |
| 1082 if (utf8) { | |
| 1083 if (wl >= MAXWORDUTF8LEN) return 0; | |
| 1084 } else { | |
| 1085 if (wl >= MAXWORDLEN) return 0; | |
| 1086 } | |
| 1087 int captype = 0; | |
| 1088 int abbv = 0; | |
| 1089 wl = cleanword(cw, word, &captype, &abbv); | |
| 1090 if (wl == 0) return 0; | |
| 1091 int ns = 0; | |
| 1092 *slst = NULL; // HU, nsug in pSMgr->suggest | |
| 1093 | |
| 1094 switch(captype) { | |
| 1095 case NOCAP: { | |
| 1096 ns = pSMgr->suggest_auto(slst, cw, ns); | |
| 1097 if (ns>0) break; | |
| 1098 break; | |
| 1099 } | |
| 1100 | |
| 1101 case INITCAP: { | |
| 1102 memcpy(wspace,cw,(wl+1)); | |
| 1103 mkallsmall(wspace); | |
| 1104 ns = pSMgr->suggest_auto(slst, wspace, ns); | |
| 1105 for (int j=0; j < ns; j++) | |
| 1106 mkinitcap((*slst)[j]); | |
| 1107 ns = pSMgr->suggest_auto(slst, cw, ns); | |
| 1108 break; | |
| 1109 | |
| 1110 } | |
| 1111 | |
| 1112 case HUHINITCAP: | |
| 1113 case HUHCAP: { | |
| 1114 ns = pSMgr->suggest_auto(slst, cw, ns); | |
| 1115 if (ns == 0) { | |
| 1116 memcpy(wspace,cw,(wl+1)); | |
| 1117 mkallsmall(wspace); | |
| 1118 ns = pSMgr->suggest_auto(slst, wspace, ns); | |
| 1119 } | |
| 1120 break; | |
| 1121 } | |
| 1122 | |
| 1123 case ALLCAP: { | |
| 1124 memcpy(wspace,cw,(wl+1)); | |
| 1125 mkallsmall(wspace); | |
| 1126 ns = pSMgr->suggest_auto(slst, wspace, ns); | |
| 1127 | |
| 1128 mkinitcap(wspace); | |
| 1129 ns = pSMgr->suggest_auto(slst, wspace, ns); | |
| 1130 | |
| 1131 for (int j=0; j < ns; j++) | |
| 1132 mkallcap((*slst)[j]); | |
| 1133 break; | |
| 1134 } | |
| 1135 } | |
| 1136 | |
| 1137 // word reversing wrapper for complex prefixes | |
| 1138 if (complexprefixes) { | |
| 1139 for (int j = 0; j < ns; j++) { | |
| 1140 if (utf8) reverseword_utf((*slst)[j]); else reverseword((*slst)[j]); | |
| 1141 } | |
| 1142 } | |
| 1143 | |
| 1144 // expand suggestions with dot(s) | |
| 1145 if (abbv && pAMgr && pAMgr->get_sugswithdots()) { | |
| 1146 for (int j = 0; j < ns; j++) { | |
| 1147 (*slst)[j] = (char *) realloc((*slst)[j], strlen((*slst)[j]) + 1 + abbv); | |
| 1148 strcat((*slst)[j], word + strlen(word) - abbv); | |
| 1149 } | |
| 1150 } | |
| 1151 | |
| 1152 // LANG_hu section: replace '-' with ' ' in Hungarian | |
| 1153 if (langnum == LANG_hu) { | |
| 1154 for (int j=0; j < ns; j++) { | |
| 1155 char * pos = strchr((*slst)[j],'-'); | |
| 1156 if (pos) { | |
| 1157 int info; | |
| 1158 char w[MAXWORDUTF8LEN]; | |
| 1159 *pos = '\0'; | |
| 1160 strcpy(w, (*slst)[j]); | |
| 1161 strcat(w, pos + 1); | |
| 1162 spell(w, &info, NULL); | |
| 1163 if ((info & SPELL_COMPOUND) && (info & SPELL_FORBIDDEN)) { | |
| 1164 *pos = ' '; | |
| 1165 } else *pos = '-'; | |
| 1166 } | |
| 1167 } | |
| 1168 } | |
| 1169 // END OF LANG_hu section | |
| 1170 return ns; | |
| 1171 } | |
| 1172 #endif | |
| 1173 | |
| 1174 int Hunspell::stem(char*** slst, char ** desc, int n) | |
| 1175 { | |
| 1176 char result[MAXLNLEN]; | |
| 1177 char result2[MAXLNLEN]; | |
| 1178 *slst = NULL; | |
| 1179 if (n == 0) return 0; | |
| 1180 *result2 = '\0'; | |
| 1181 for (int i = 0; i < n; i++) { | |
| 1182 *result = '\0'; | |
| 1183 // add compound word parts (except the last one) | |
| 1184 char * s = (char *) desc[i]; | |
| 1185 char * part = strstr(s, MORPH_PART); | |
| 1186 if (part) { | |
| 1187 char * nextpart = strstr(part + 1, MORPH_PART); | |
| 1188 while (nextpart) { | |
| 1189 copy_field(result + strlen(result), part, MORPH_PART); | |
| 1190 part = nextpart; | |
| 1191 nextpart = strstr(part + 1, MORPH_PART); | |
| 1192 } | |
| 1193 s = part; | |
| 1194 } | |
| 1195 | |
| 1196 char **pl; | |
| 1197 char tok[MAXLNLEN]; | |
| 1198 strcpy(tok, s); | |
| 1199 char * alt = strstr(tok, " | "); | |
| 1200 while (alt) { | |
| 1201 alt[1] = MSEP_ALT; | |
| 1202 alt = strstr(alt, " | "); | |
| 1203 } | |
| 1204 int pln = line_tok(tok, &pl, MSEP_ALT); | |
| 1205 for (int k = 0; k < pln; k++) { | |
| 1206 // add derivational suffixes | |
| 1207 if (strstr(pl[k], MORPH_DERI_SFX)) { | |
| 1208 // remove inflectional suffixes | |
| 1209 char * is = strstr(pl[k], MORPH_INFL_SFX); | |
| 1210 if (is) *is = '\0'; | |
| 1211 char * sg = pSMgr->suggest_gen(&(pl[k]), 1, pl[k]); | |
| 1212 if (sg) { | |
| 1213 char ** gen; | |
| 1214 int genl = line_tok(sg, &gen, MSEP_REC); | |
| 1215 free(sg); | |
| 1216 for (int j = 0; j < genl; j++) { | |
| 1217 sprintf(result2 + strlen(result2), "%c%s%s", | |
| 1218 MSEP_REC, result, gen[j]); | |
| 1219 } | |
| 1220 freelist(&gen, genl); | |
| 1221 } | |
| 1222 } else { | |
| 1223 sprintf(result2 + strlen(result2), "%c%s", MSEP_REC, result); | |
| 1224 if (strstr(pl[k], MORPH_SURF_PFX)) { | |
| 1225 copy_field(result2 + strlen(result2), pl[k], MORPH_SURF_PFX); | |
| 1226 } | |
| 1227 copy_field(result2 + strlen(result2), pl[k], MORPH_STEM); | |
| 1228 } | |
| 1229 } | |
| 1230 freelist(&pl, pln); | |
| 1231 } | |
| 1232 int sln = line_tok(result2, slst, MSEP_REC); | |
| 1233 return uniqlist(*slst, sln); | |
| 1234 | |
| 1235 } | |
| 1236 | |
| 1237 int Hunspell::stem(char*** slst, const char * word) | |
| 1238 { | |
| 1239 char ** pl; | |
| 1240 int pln = analyze(&pl, word); | |
| 1241 int pln2 = stem(slst, pl, pln); | |
| 1242 freelist(&pl, pln); | |
| 1243 return pln2; | |
| 1244 } | |
| 1245 | |
| 1246 #ifdef HUNSPELL_EXPERIMENTAL | |
| 1247 int Hunspell::suggest_pos_stems(char*** slst, const char * word) | |
| 1248 { | |
| 1249 char cw[MAXWORDUTF8LEN]; | |
| 1250 char wspace[MAXWORDUTF8LEN]; | |
| 1251 if (! pSMgr || maxdic == 0) return 0; | |
| 1252 int wl = strlen(word); | |
| 1253 if (utf8) { | |
| 1254 if (wl >= MAXWORDUTF8LEN) return 0; | |
| 1255 } else { | |
| 1256 if (wl >= MAXWORDLEN) return 0; | |
| 1257 } | |
| 1258 int captype = 0; | |
| 1259 int abbv = 0; | |
| 1260 wl = cleanword(cw, word, &captype, &abbv); | |
| 1261 if (wl == 0) return 0; | |
| 1262 | |
| 1263 int ns = 0; // ns=0 = normalized input | |
| 1264 | |
| 1265 *slst = NULL; // HU, nsug in pSMgr->suggest | |
| 1266 | |
| 1267 switch(captype) { | |
| 1268 case HUHCAP: | |
| 1269 case NOCAP: { | |
| 1270 ns = pSMgr->suggest_pos_stems(slst, cw, ns); | |
| 1271 | |
| 1272 if ((abbv) && (ns == 0)) { | |
| 1273 memcpy(wspace,cw,wl); | |
| 1274 *(wspace+wl) = '.'; | |
| 1275 *(wspace+wl+1) = '\0'; | |
| 1276 ns = pSMgr->suggest_pos_stems(slst, wspace, ns); | |
| 1277 } | |
| 1278 | |
| 1279 break; | |
| 1280 } | |
| 1281 | |
| 1282 case INITCAP: { | |
| 1283 | |
| 1284 ns = pSMgr->suggest_pos_stems(slst, cw, ns); | |
| 1285 | |
| 1286 if (ns == 0 || ((*slst)[0][0] == '#')) { | |
| 1287 memcpy(wspace,cw,(wl+1)); | |
| 1288 mkallsmall(wspace); | |
| 1289 ns = pSMgr->suggest_pos_stems(slst, wspace, ns); | |
| 1290 } | |
| 1291 | |
| 1292 break; | |
| 1293 | |
| 1294 } | |
| 1295 | |
| 1296 case ALLCAP: { | |
| 1297 ns = pSMgr->suggest_pos_stems(slst, cw, ns); | |
| 1298 if (ns != 0) break; | |
| 1299 | |
| 1300 memcpy(wspace,cw,(wl+1)); | |
| 1301 mkallsmall(wspace); | |
| 1302 ns = pSMgr->suggest_pos_stems(slst, wspace, ns); | |
| 1303 | |
| 1304 if (ns == 0) { | |
| 1305 mkinitcap(wspace); | |
| 1306 ns = pSMgr->suggest_pos_stems(slst, wspace, ns); | |
| 1307 } | |
| 1308 break; | |
| 1309 } | |
| 1310 } | |
| 1311 | |
| 1312 return ns; | |
| 1313 } | |
| 1314 #endif // END OF HUNSPELL_EXPERIMENTAL CODE | |
| 1315 | |
| 1316 const char * Hunspell::get_wordchars() | |
| 1317 { | |
| 1318 return pAMgr->get_wordchars(); | |
| 1319 } | |
| 1320 | |
| 1321 unsigned short * Hunspell::get_wordchars_utf16(int * len) | |
| 1322 { | |
| 1323 return pAMgr->get_wordchars_utf16(len); | |
| 1324 } | |
| 1325 | |
| 1326 void Hunspell::mkinitcap(char * p) | |
| 1327 { | |
| 1328 if (!utf8) { | |
| 1329 if (*p != '\0') *p = csconv[((unsigned char)*p)].cupper; | |
| 1330 } else { | |
| 1331 int len; | |
| 1332 w_char u[MAXWORDLEN]; | |
| 1333 len = u8_u16(u, MAXWORDLEN, p); | |
| 1334 unsigned short i = unicodetoupper((u[0].h << 8) + u[0].l, langnum); | |
| 1335 u[0].h = (unsigned char) (i >> 8); | |
| 1336 u[0].l = (unsigned char) (i & 0x00FF); | |
| 1337 u16_u8(p, MAXWORDUTF8LEN, u, len); | |
| 1338 } | |
| 1339 } | |
| 1340 | |
| 1341 int Hunspell::mkinitcap2(char * p, w_char * u, int nc) | |
| 1342 { | |
| 1343 if (!utf8) { | |
| 1344 if (*p != '\0') *p = csconv[((unsigned char)*p)].cupper; | |
| 1345 } else if (nc > 0) { | |
| 1346 unsigned short i = unicodetoupper((u[0].h << 8) + u[0].l, langnum); | |
| 1347 u[0].h = (unsigned char) (i >> 8); | |
| 1348 u[0].l = (unsigned char) (i & 0x00FF); | |
| 1349 u16_u8(p, MAXWORDUTF8LEN, u, nc); | |
| 1350 return strlen(p); | |
| 1351 } | |
| 1352 return nc; | |
| 1353 } | |
| 1354 | |
| 1355 int Hunspell::mkinitsmall2(char * p, w_char * u, int nc) | |
| 1356 { | |
| 1357 if (!utf8) { | |
| 1358 if (*p != '\0') *p = csconv[((unsigned char)*p)].clower; | |
| 1359 } else if (nc > 0) { | |
| 1360 unsigned short i = unicodetolower((u[0].h << 8) + u[0].l, langnum); | |
| 1361 u[0].h = (unsigned char) (i >> 8); | |
| 1362 u[0].l = (unsigned char) (i & 0x00FF); | |
| 1363 u16_u8(p, MAXWORDUTF8LEN, u, nc); | |
| 1364 return strlen(p); | |
| 1365 } | |
| 1366 return nc; | |
| 1367 } | |
| 1368 | |
| 1369 int Hunspell::add(const char * word) | |
| 1370 { | |
| 1371 if (pHMgr[0]) return (pHMgr[0])->add(word); | |
| 1372 return 0; | |
| 1373 } | |
| 1374 | |
| 1375 int Hunspell::add_with_affix(const char * word, const char * example) | |
| 1376 { | |
| 1377 if (pHMgr[0]) return (pHMgr[0])->add_with_affix(word, example); | |
| 1378 return 0; | |
| 1379 } | |
| 1380 | |
| 1381 int Hunspell::remove(const char * word) | |
| 1382 { | |
| 1383 if (pHMgr[0]) return (pHMgr[0])->remove(word); | |
| 1384 return 0; | |
| 1385 } | |
| 1386 | |
| 1387 const char * Hunspell::get_version() | |
| 1388 { | |
| 1389 return pAMgr->get_version(); | |
| 1390 } | |
| 1391 | |
| 1392 struct cs_info * Hunspell::get_csconv() | |
| 1393 { | |
| 1394 return csconv; | |
| 1395 } | |
| 1396 | |
| 1397 void Hunspell::cat_result(char * result, char * st) | |
| 1398 { | |
| 1399 if (st) { | |
| 1400 if (*result) mystrcat(result, "\n", MAXLNLEN); | |
| 1401 mystrcat(result, st, MAXLNLEN); | |
| 1402 free(st); | |
| 1403 } | |
| 1404 } | |
| 1405 | |
| 1406 int Hunspell::analyze(char*** slst, const char * word) | |
| 1407 { | |
| 1408 char cw[MAXWORDUTF8LEN]; | |
| 1409 char wspace[MAXWORDUTF8LEN]; | |
| 1410 w_char unicw[MAXWORDLEN]; | |
| 1411 int wl2 = 0; | |
| 1412 *slst = NULL; | |
| 1413 if (! pSMgr || maxdic == 0) return 0; | |
| 1414 int nc = strlen(word); | |
| 1415 if (utf8) { | |
| 1416 if (nc >= MAXWORDUTF8LEN) return 0; | |
| 1417 } else { | |
| 1418 if (nc >= MAXWORDLEN) return 0; | |
| 1419 } | |
| 1420 int captype = 0; | |
| 1421 int abbv = 0; | |
| 1422 int wl = 0; | |
| 1423 | |
| 1424 // input conversion | |
| 1425 RepList * rl = (pAMgr) ? pAMgr->get_iconvtable() : NULL; | |
| 1426 if (rl && rl->conv(word, wspace)) wl = cleanword2(cw, wspace, unicw, &nc, &cap
type, &abbv); | |
| 1427 else wl = cleanword2(cw, word, unicw, &nc, &captype, &abbv); | |
| 1428 | |
| 1429 if (wl == 0) { | |
| 1430 if (abbv) { | |
| 1431 for (wl = 0; wl < abbv; wl++) cw[wl] = '.'; | |
| 1432 cw[wl] = '\0'; | |
| 1433 abbv = 0; | |
| 1434 } else return 0; | |
| 1435 } | |
| 1436 | |
| 1437 char result[MAXLNLEN]; | |
| 1438 char * st = NULL; | |
| 1439 | |
| 1440 *result = '\0'; | |
| 1441 | |
| 1442 int n = 0; | |
| 1443 int n2 = 0; | |
| 1444 int n3 = 0; | |
| 1445 | |
| 1446 // test numbers | |
| 1447 // LANG_hu section: set dash information for suggestions | |
| 1448 if (langnum == LANG_hu) { | |
| 1449 while ((n < wl) && | |
| 1450 (((cw[n] <= '9') && (cw[n] >= '0')) || (((cw[n] == '.') || (cw[n] == ','
)) && (n > 0)))) { | |
| 1451 n++; | |
| 1452 if ((cw[n] == '.') || (cw[n] == ',')) { | |
| 1453 if (((n2 == 0) && (n > 3)) || | |
| 1454 ((n2 > 0) && ((cw[n-1] == '.') || (cw[n-1] == ',')))) br
eak; | |
| 1455 n2++; | |
| 1456 n3 = n; | |
| 1457 } | |
| 1458 } | |
| 1459 | |
| 1460 if ((n == wl) && (n3 > 0) && (n - n3 > 3)) return 0; | |
| 1461 if ((n == wl) || ((n>0) && ((cw[n]=='%') || (cw[n]=='\xB0')) && checkword(cw+n
, NULL, NULL))) { | |
| 1462 mystrcat(result, cw, MAXLNLEN); | |
| 1463 result[n - 1] = '\0'; | |
| 1464 if (n == wl) cat_result(result, pSMgr->suggest_morph(cw + n - 1)); | |
| 1465 else { | |
| 1466 char sign = cw[n]; | |
| 1467 cw[n] = '\0'; | |
| 1468 cat_result(result, pSMgr->suggest_morph(cw + n - 1)); | |
| 1469 mystrcat(result, "+", MAXLNLEN); // XXX SPEC. MORPHCODE | |
| 1470 cw[n] = sign; | |
| 1471 cat_result(result, pSMgr->suggest_morph(cw + n)); | |
| 1472 } | |
| 1473 return line_tok(result, slst, MSEP_REC); | |
| 1474 } | |
| 1475 } | |
| 1476 // END OF LANG_hu section | |
| 1477 | |
| 1478 switch(captype) { | |
| 1479 case HUHCAP: | |
| 1480 case HUHINITCAP: | |
| 1481 case NOCAP: { | |
| 1482 cat_result(result, pSMgr->suggest_morph(cw)); | |
| 1483 if (abbv) { | |
| 1484 memcpy(wspace,cw,wl); | |
| 1485 *(wspace+wl) = '.'; | |
| 1486 *(wspace+wl+1) = '\0'; | |
| 1487 cat_result(result, pSMgr->suggest_morph(wspace)); | |
| 1488 } | |
| 1489 break; | |
| 1490 } | |
| 1491 case INITCAP: { | |
| 1492 wl = mkallsmall2(cw, unicw, nc); | |
| 1493 memcpy(wspace,cw,(wl+1)); | |
| 1494 wl2 = mkinitcap2(cw, unicw, nc); | |
| 1495 cat_result(result, pSMgr->suggest_morph(wspace)); | |
| 1496 cat_result(result, pSMgr->suggest_morph(cw)); | |
| 1497 if (abbv) { | |
| 1498 *(wspace+wl) = '.'; | |
| 1499 *(wspace+wl+1) = '\0'; | |
| 1500 cat_result(result, pSMgr->suggest_morph(wspace)); | |
| 1501 | |
| 1502 memcpy(wspace, cw, wl2); | |
| 1503 *(wspace+wl2) = '.'; | |
| 1504 *(wspace+wl2+1) = '\0'; | |
| 1505 | |
| 1506 cat_result(result, pSMgr->suggest_morph(wspace)); | |
| 1507 } | |
| 1508 break; | |
| 1509 } | |
| 1510 case ALLCAP: { | |
| 1511 cat_result(result, pSMgr->suggest_morph(cw)); | |
| 1512 if (abbv) { | |
| 1513 memcpy(wspace,cw,wl); | |
| 1514 *(wspace+wl) = '.'; | |
| 1515 *(wspace+wl+1) = '\0'; | |
| 1516 cat_result(result, pSMgr->suggest_morph(cw)); | |
| 1517 } | |
| 1518 wl = mkallsmall2(cw, unicw, nc); | |
| 1519 memcpy(wspace,cw,(wl+1)); | |
| 1520 wl2 = mkinitcap2(cw, unicw, nc); | |
| 1521 | |
| 1522 cat_result(result, pSMgr->suggest_morph(wspace)); | |
| 1523 cat_result(result, pSMgr->suggest_morph(cw)); | |
| 1524 if (abbv) { | |
| 1525 *(wspace+wl) = '.'; | |
| 1526 *(wspace+wl+1) = '\0'; | |
| 1527 cat_result(result, pSMgr->suggest_morph(wspace)); | |
| 1528 | |
| 1529 memcpy(wspace, cw, wl2); | |
| 1530 *(wspace+wl2) = '.'; | |
| 1531 *(wspace+wl2+1) = '\0'; | |
| 1532 | |
| 1533 cat_result(result, pSMgr->suggest_morph(wspace)); | |
| 1534 } | |
| 1535 break; | |
| 1536 } | |
| 1537 } | |
| 1538 | |
| 1539 if (*result) { | |
| 1540 // word reversing wrapper for complex prefixes | |
| 1541 if (complexprefixes) { | |
| 1542 if (utf8) reverseword_utf(result); else reverseword(result); | |
| 1543 } | |
| 1544 return line_tok(result, slst, MSEP_REC); | |
| 1545 } | |
| 1546 | |
| 1547 // compound word with dash (HU) I18n | |
| 1548 char * dash = NULL; | |
| 1549 int nresult = 0; | |
| 1550 // LANG_hu section: set dash information for suggestions | |
| 1551 if (langnum == LANG_hu) dash = (char *) strchr(cw,'-'); | |
| 1552 if ((langnum == LANG_hu) && dash) { | |
| 1553 *dash='\0'; | |
| 1554 // examine 2 sides of the dash | |
| 1555 if (dash[1] == '\0') { // base word ending with dash | |
| 1556 if (spell(cw)) { | |
| 1557 char * p = pSMgr->suggest_morph(cw); | |
| 1558 if (p) { | |
| 1559 int ret = line_tok(p, slst, MSEP_REC); | |
| 1560 free(p); | |
| 1561 return ret; | |
| 1562 } | |
| 1563 | |
| 1564 } | |
| 1565 } else if ((dash[1] == 'e') && (dash[2] == '\0')) { // XXX (HU) -e hat. | |
| 1566 if (spell(cw) && (spell("-e"))) { | |
| 1567 st = pSMgr->suggest_morph(cw); | |
| 1568 if (st) { | |
| 1569 mystrcat(result, st, MAXLNLEN); | |
| 1570 free(st); | |
| 1571 } | |
| 1572 mystrcat(result,"+", MAXLNLEN); // XXX spec. separator i
n MORPHCODE | |
| 1573 st = pSMgr->suggest_morph("-e"); | |
| 1574 if (st) { | |
| 1575 mystrcat(result, st, MAXLNLEN); | |
| 1576 free(st); | |
| 1577 } | |
| 1578 return line_tok(result, slst, MSEP_REC); | |
| 1579 } | |
| 1580 } else { | |
| 1581 // first word ending with dash: word- XXX ??? | |
| 1582 char r2 = *(dash + 1); | |
| 1583 dash[0]='-'; | |
| 1584 dash[1]='\0'; | |
| 1585 nresult = spell(cw); | |
| 1586 dash[1] = r2; | |
| 1587 dash[0]='\0'; | |
| 1588 if (nresult && spell(dash+1) && ((strlen(dash+1) > 1) || | |
| 1589 ((dash[1] > '0') && (dash[1] < '9')))) { | |
| 1590 st = pSMgr->suggest_morph(cw); | |
| 1591 if (st) { | |
| 1592 mystrcat(result, st, MAXLNLEN); | |
| 1593 free(st); | |
| 1594 mystrcat(result,"+", MAXLNLEN); // XXX spec. sep
arator in MORPHCODE | |
| 1595 } | |
| 1596 st = pSMgr->suggest_morph(dash+1); | |
| 1597 if (st) { | |
| 1598 mystrcat(result, st, MAXLNLEN); | |
| 1599 free(st); | |
| 1600 } | |
| 1601 return line_tok(result, slst, MSEP_REC); | |
| 1602 } | |
| 1603 } | |
| 1604 // affixed number in correct word | |
| 1605 if (nresult && (dash > cw) && (((*(dash-1)<='9') && | |
| 1606 (*(dash-1)>='0')) || (*(dash-1)=='.'))) { | |
| 1607 *dash='-'; | |
| 1608 n = 1; | |
| 1609 if (*(dash - n) == '.') n++; | |
| 1610 // search first not a number character to left from dash | |
| 1611 while (((dash - n)>=cw) && ((*(dash - n)=='0') || (n < 3)) && (n < 6))
{ | |
| 1612 n++; | |
| 1613 } | |
| 1614 if ((dash - n) < cw) n--; | |
| 1615 // numbers: valami1000000-hoz | |
| 1616 // examine 100000-hoz, 10000-hoz 1000-hoz, 10-hoz, | |
| 1617 // 56-hoz, 6-hoz | |
| 1618 for(; n >= 1; n--) { | |
| 1619 if ((*(dash - n) >= '0') && (*(dash - n) <= '9') && checkword(dash -
n, NULL, NULL)) { | |
| 1620 mystrcat(result, cw, MAXLNLEN); | |
| 1621 result[dash - cw - n] = '\0'; | |
| 1622 st = pSMgr->suggest_morph(dash - n); | |
| 1623 if (st) { | |
| 1624 mystrcat(result, st, MAXLNLEN); | |
| 1625 free(st); | |
| 1626 } | |
| 1627 return line_tok(result, slst, MSEP_REC); | |
| 1628 } | |
| 1629 } | |
| 1630 } | |
| 1631 } | |
| 1632 return 0; | |
| 1633 } | |
| 1634 | |
| 1635 int Hunspell::generate(char*** slst, const char * word, char ** pl, int pln) | |
| 1636 { | |
| 1637 *slst = NULL; | |
| 1638 if (!pSMgr || !pln) return 0; | |
| 1639 char **pl2; | |
| 1640 int pl2n = analyze(&pl2, word); | |
| 1641 int captype = 0; | |
| 1642 int abbv = 0; | |
| 1643 char cw[MAXWORDUTF8LEN]; | |
| 1644 cleanword(cw, word, &captype, &abbv); | |
| 1645 char result[MAXLNLEN]; | |
| 1646 *result = '\0'; | |
| 1647 | |
| 1648 for (int i = 0; i < pln; i++) { | |
| 1649 cat_result(result, pSMgr->suggest_gen(pl2, pl2n, pl[i])); | |
| 1650 } | |
| 1651 freelist(&pl2, pl2n); | |
| 1652 | |
| 1653 if (*result) { | |
| 1654 // allcap | |
| 1655 if (captype == ALLCAP) mkallcap(result); | |
| 1656 | |
| 1657 // line split | |
| 1658 int linenum = line_tok(result, slst, MSEP_REC); | |
| 1659 | |
| 1660 // capitalize | |
| 1661 if (captype == INITCAP || captype == HUHINITCAP) { | |
| 1662 for (int j=0; j < linenum; j++) mkinitcap((*slst)[j]); | |
| 1663 } | |
| 1664 | |
| 1665 // temporary filtering of prefix related errors (eg. | |
| 1666 // generate("undrinkable", "eats") --> "undrinkables" and "*undrinks") | |
| 1667 | |
| 1668 int r = 0; | |
| 1669 for (int j=0; j < linenum; j++) { | |
| 1670 if (!spell((*slst)[j])) { | |
| 1671 free((*slst)[j]); | |
| 1672 (*slst)[j] = NULL; | |
| 1673 } else { | |
| 1674 if (r < j) (*slst)[r] = (*slst)[j]; | |
| 1675 r++; | |
| 1676 } | |
| 1677 } | |
| 1678 if (r > 0) return r; | |
| 1679 free(*slst); | |
| 1680 *slst = NULL; | |
| 1681 } | |
| 1682 return 0; | |
| 1683 } | |
| 1684 | |
| 1685 int Hunspell::generate(char*** slst, const char * word, const char * pattern) | |
| 1686 { | |
| 1687 char **pl; | |
| 1688 int pln = analyze(&pl, pattern); | |
| 1689 int n = generate(slst, word, pl, pln); | |
| 1690 freelist(&pl, pln); | |
| 1691 return uniqlist(*slst, n); | |
| 1692 } | |
| 1693 | |
| 1694 // minimal XML parser functions | |
| 1695 int Hunspell::get_xml_par(char * dest, const char * par, int max) | |
| 1696 { | |
| 1697 char * d = dest; | |
| 1698 if (!par) return 0; | |
| 1699 char end = *par; | |
| 1700 char * dmax = dest + max; | |
| 1701 if (end == '>') end = '<'; | |
| 1702 else if (end != '\'' && end != '"') return 0; // bad XML | |
| 1703 for (par++; d < dmax && *par != '\0' && *par != end; par++, d++) *d = *par; | |
| 1704 *d = '\0'; | |
| 1705 mystrrep(dest, "<", "<"); | |
| 1706 mystrrep(dest, "&", "&"); | |
| 1707 return (int)(d - dest); | |
| 1708 } | |
| 1709 | |
| 1710 int Hunspell::get_langnum() const | |
| 1711 { | |
| 1712 return langnum; | |
| 1713 } | |
| 1714 | |
| 1715 // return the beginning of the element (attr == NULL) or the attribute | |
| 1716 const char * Hunspell::get_xml_pos(const char * s, const char * attr) | |
| 1717 { | |
| 1718 const char * end = strchr(s, '>'); | |
| 1719 const char * p = s; | |
| 1720 if (attr == NULL) return end; | |
| 1721 do { | |
| 1722 p = strstr(p, attr); | |
| 1723 if (!p || p >= end) return 0; | |
| 1724 } while (*(p-1) != ' ' && *(p-1) != '\n'); | |
| 1725 return p + strlen(attr); | |
| 1726 } | |
| 1727 | |
| 1728 int Hunspell::check_xml_par(const char * q, const char * attr, const char * valu
e) { | |
| 1729 char cw[MAXWORDUTF8LEN]; | |
| 1730 if (get_xml_par(cw, get_xml_pos(q, attr), MAXWORDUTF8LEN - 1) && | |
| 1731 strcmp(cw, value) == 0) return 1; | |
| 1732 return 0; | |
| 1733 } | |
| 1734 | |
| 1735 int Hunspell::get_xml_list(char ***slst, char * list, const char * tag) { | |
| 1736 int n = 0; | |
| 1737 char * p; | |
| 1738 if (!list) return 0; | |
| 1739 for (p = list; ((p = strstr(p, tag)) != NULL); p++) n++; | |
| 1740 if (n == 0) return 0; | |
| 1741 *slst = (char **) malloc(sizeof(char *) * n); | |
| 1742 if (!*slst) return 0; | |
| 1743 for (p = list, n = 0; ((p = strstr(p, tag)) != NULL); p++, n++) { | |
| 1744 int l = strlen(p); | |
| 1745 (*slst)[n] = (char *) malloc(l + 1); | |
| 1746 if (!(*slst)[n]) return n; | |
| 1747 if (!get_xml_par((*slst)[n], p + strlen(tag) - 1, l)) { | |
| 1748 free((*slst)[n]); | |
| 1749 break; | |
| 1750 } | |
| 1751 } | |
| 1752 return n; | |
| 1753 } | |
| 1754 | |
| 1755 int Hunspell::spellml(char*** slst, const char * word) | |
| 1756 { | |
| 1757 char *q, *q2; | |
| 1758 char cw[MAXWORDUTF8LEN], cw2[MAXWORDUTF8LEN]; | |
| 1759 q = (char *) strstr(word, "<query"); | |
| 1760 if (!q) return 0; // bad XML input | |
| 1761 q2 = strchr(q, '>'); | |
| 1762 if (!q2) return 0; // bad XML input | |
| 1763 q2 = strstr(q2, "<word"); | |
| 1764 if (!q2) return 0; // bad XML input | |
| 1765 if (check_xml_par(q, "type=", "analyze")) { | |
| 1766 int n = 0, s = 0; | |
| 1767 if (get_xml_par(cw, strchr(q2, '>'), MAXWORDUTF8LEN - 10)) n = analyze(sls
t, cw); | |
| 1768 if (n == 0) return 0; | |
| 1769 // convert the result to <code><a>ana1</a><a>ana2</a></code> format | |
| 1770 for (int i = 0; i < n; i++) s+= strlen((*slst)[i]); | |
| 1771 char * r = (char *) malloc(6 + 5 * s + 7 * n + 7 + 1); // XXX 5*s->&->&
; | |
| 1772 if (!r) return 0; | |
| 1773 strcpy(r, "<code>"); | |
| 1774 for (int i = 0; i < n; i++) { | |
| 1775 int l = strlen(r); | |
| 1776 strcpy(r + l, "<a>"); | |
| 1777 strcpy(r + l + 3, (*slst)[i]); | |
| 1778 mystrrep(r + l + 3, "\t", " "); | |
| 1779 mystrrep(r + l + 3, "<", "<"); | |
| 1780 mystrrep(r + l + 3, "&", "&"); | |
| 1781 strcat(r, "</a>"); | |
| 1782 free((*slst)[i]); | |
| 1783 } | |
| 1784 strcat(r, "</code>"); | |
| 1785 (*slst)[0] = r; | |
| 1786 return 1; | |
| 1787 } else if (check_xml_par(q, "type=", "stem")) { | |
| 1788 if (get_xml_par(cw, strchr(q2, '>'), MAXWORDUTF8LEN - 1)) return stem(slst
, cw); | |
| 1789 } else if (check_xml_par(q, "type=", "generate")) { | |
| 1790 int n = get_xml_par(cw, strchr(q2, '>'), MAXWORDUTF8LEN - 1); | |
| 1791 if (n == 0) return 0; | |
| 1792 char * q3 = strstr(q2 + 1, "<word"); | |
| 1793 if (q3) { | |
| 1794 if (get_xml_par(cw2, strchr(q3, '>'), MAXWORDUTF8LEN - 1)) { | |
| 1795 return generate(slst, cw, cw2); | |
| 1796 } | |
| 1797 } else { | |
| 1798 if ((q2 = strstr(q2 + 1, "<code")) != NULL) { | |
| 1799 char ** slst2; | |
| 1800 if ((n = get_xml_list(&slst2, strchr(q2, '>'), "<a>")) != 0) { | |
| 1801 int n2 = generate(slst, cw, slst2, n); | |
| 1802 freelist(&slst2, n); | |
| 1803 return uniqlist(*slst, n2); | |
| 1804 } | |
| 1805 freelist(&slst2, n); | |
| 1806 } | |
| 1807 } | |
| 1808 } | |
| 1809 return 0; | |
| 1810 } | |
| 1811 | |
| 1812 | |
| 1813 #ifdef HUNSPELL_EXPERIMENTAL | |
| 1814 // XXX need UTF-8 support | |
| 1815 char * Hunspell::morph_with_correction(const char * word) | |
| 1816 { | |
| 1817 char cw[MAXWORDUTF8LEN]; | |
| 1818 char wspace[MAXWORDUTF8LEN]; | |
| 1819 if (! pSMgr || maxdic == 0) return NULL; | |
| 1820 int wl = strlen(word); | |
| 1821 if (utf8) { | |
| 1822 if (wl >= MAXWORDUTF8LEN) return NULL; | |
| 1823 } else { | |
| 1824 if (wl >= MAXWORDLEN) return NULL; | |
| 1825 } | |
| 1826 int captype = 0; | |
| 1827 int abbv = 0; | |
| 1828 wl = cleanword(cw, word, &captype, &abbv); | |
| 1829 if (wl == 0) return NULL; | |
| 1830 | |
| 1831 char result[MAXLNLEN]; | |
| 1832 char * st = NULL; | |
| 1833 | |
| 1834 *result = '\0'; | |
| 1835 | |
| 1836 | |
| 1837 switch(captype) { | |
| 1838 case NOCAP: { | |
| 1839 st = pSMgr->suggest_morph_for_spelling_error(cw); | |
| 1840 if (st) { | |
| 1841 mystrcat(result, st, MAXLNLEN); | |
| 1842 free(st); | |
| 1843 } | |
| 1844 if (abbv) { | |
| 1845 memcpy(wspace,cw,wl); | |
| 1846 *(wspace+wl) = '.'; | |
| 1847 *(wspace+wl+1) = '\0'; | |
| 1848 st = pSMgr->suggest_morph_for_spelling_error(wspace); | |
| 1849 if (st) { | |
| 1850 if (*result) mystrcat(result, "\n", MAXLNLEN); | |
| 1851 mystrcat(result, st, MAXLNLEN); | |
| 1852 free(st); | |
| 1853 } | |
| 1854 } | |
| 1855 break; | |
| 1856 } | |
| 1857 case INITCAP: { | |
| 1858 memcpy(wspace,cw,(wl+1)); | |
| 1859 mkallsmall(wspace); | |
| 1860 st = pSMgr->suggest_morph_for_spelling_error(wspace); | |
| 1861 if (st) { | |
| 1862 mystrcat(result, st, MAXLNLEN); | |
| 1863 free(st); | |
| 1864 } | |
| 1865 st = pSMgr->suggest_morph_for_spelling_error(cw); | |
| 1866 if (st) { | |
| 1867 if (*result) mystrcat(result, "\n", MAXLNLEN); | |
| 1868 mystrcat(result, st, MAXLNLEN); | |
| 1869 free(st); | |
| 1870 } | |
| 1871 if (abbv) { | |
| 1872 memcpy(wspace,cw,wl); | |
| 1873 *(wspace+wl) = '.'; | |
| 1874 *(wspace+wl+1) = '\0'; | |
| 1875 mkallsmall(wspace); | |
| 1876 st = pSMgr->suggest_morph_for_spelling_error(wspace); | |
| 1877 if (st) { | |
| 1878 if (*result) mystrcat(result, "\n", MAXLNLEN); | |
| 1879 mystrcat(result, st, MAXLNLEN); | |
| 1880 free(st); | |
| 1881 } | |
| 1882 mkinitcap(wspace); | |
| 1883 st = pSMgr->suggest_morph_for_spelling_error(wspace); | |
| 1884 if (st) { | |
| 1885 if (*result) mystrcat(result, "\n", MAXLNLEN); | |
| 1886 mystrcat(result, st, MAXLNLEN); | |
| 1887 free(st); | |
| 1888 } | |
| 1889 } | |
| 1890 break; | |
| 1891 } | |
| 1892 case HUHCAP: { | |
| 1893 st = pSMgr->suggest_morph_for_spelling_error(cw); | |
| 1894 if (st) { | |
| 1895 mystrcat(result, st, MAXLNLEN); | |
| 1896 free(st); | |
| 1897 } | |
| 1898 memcpy(wspace,cw,(wl+1)); | |
| 1899 mkallsmall(wspace); | |
| 1900 st = pSMgr->suggest_morph_for_spelling_error(wspace); | |
| 1901 if (st) { | |
| 1902 if (*result) mystrcat(result, "\n", MAXLNLEN); | |
| 1903 mystrcat(result, st, MAXLNLEN); | |
| 1904 free(st); | |
| 1905 } | |
| 1906 break; | |
| 1907 } | |
| 1908 case ALLCAP: { | |
| 1909 memcpy(wspace,cw,(wl+1)); | |
| 1910 st = pSMgr->suggest_morph_for_spelling_error(wspace); | |
| 1911 if (st) { | |
| 1912 mystrcat(result, st, MAXLNLEN); | |
| 1913 free(st); | |
| 1914 } | |
| 1915 mkallsmall(wspace); | |
| 1916 st = pSMgr->suggest_morph_for_spelling_error(wspace); | |
| 1917 if (st) { | |
| 1918 if (*result) mystrcat(result, "\n", MAXLNLEN); | |
| 1919 mystrcat(result, st, MAXLNLEN); | |
| 1920 free(st); | |
| 1921 } | |
| 1922 mkinitcap(wspace); | |
| 1923 st = pSMgr->suggest_morph_for_spelling_error(wspace); | |
| 1924 if (st) { | |
| 1925 if (*result) mystrcat(result, "\n", MAXLNLEN); | |
| 1926 mystrcat(result, st, MAXLNLEN); | |
| 1927 free(st); | |
| 1928 } | |
| 1929 if (abbv) { | |
| 1930 memcpy(wspace,cw,(wl+1)); | |
| 1931 *(wspace+wl) = '.'; | |
| 1932 *(wspace+wl+1) = '\0'; | |
| 1933 if (*result) mystrcat(result, "\n", MAXLNLEN); | |
| 1934 st = pSMgr->suggest_morph_for_spelling_error(wspace); | |
| 1935 if (st) { | |
| 1936 mystrcat(result, st, MAXLNLEN); | |
| 1937 free(st); | |
| 1938 } | |
| 1939 mkallsmall(wspace); | |
| 1940 st = pSMgr->suggest_morph_for_spelling_error(wspace); | |
| 1941 if (st) { | |
| 1942 if (*result) mystrcat(result, "\n", MAXLNLEN); | |
| 1943 mystrcat(result, st, MAXLNLEN); | |
| 1944 free(st); | |
| 1945 } | |
| 1946 mkinitcap(wspace); | |
| 1947 st = pSMgr->suggest_morph_for_spelling_error(wspace); | |
| 1948 if (st) { | |
| 1949 if (*result) mystrcat(result, "\n", MAXLNLEN); | |
| 1950 mystrcat(result, st, MAXLNLEN); | |
| 1951 free(st); | |
| 1952 } | |
| 1953 } | |
| 1954 break; | |
| 1955 } | |
| 1956 } | |
| 1957 | |
| 1958 if (*result) return mystrdup(result); | |
| 1959 return NULL; | |
| 1960 } | |
| 1961 | |
| 1962 #endif // END OF HUNSPELL_EXPERIMENTAL CODE | |
| 1963 | |
| 1964 Hunhandle *Hunspell_create(const char * affpath, const char * dpath) | |
| 1965 { | |
| 1966 #ifdef HUNSPELL_CHROME_CLIENT | |
| 1967 return NULL; | |
| 1968 #else | |
| 1969 return (Hunhandle*)(new Hunspell(affpath, dpath)); | |
| 1970 #endif | |
| 1971 } | |
| 1972 | |
| 1973 Hunhandle *Hunspell_create_key(const char * affpath, const char * dpath, | |
| 1974 const char * key) | |
| 1975 { | |
| 1976 #ifdef HUNSPELL_CHROME_CLIENT | |
| 1977 return NULL; | |
| 1978 #else | |
| 1979 return (Hunhandle*)(new Hunspell(affpath, dpath, key)); | |
| 1980 #endif | |
| 1981 } | |
| 1982 | |
| 1983 void Hunspell_destroy(Hunhandle *pHunspell) | |
| 1984 { | |
| 1985 delete (Hunspell*)(pHunspell); | |
| 1986 } | |
| 1987 | |
| 1988 int Hunspell_spell(Hunhandle *pHunspell, const char *word) | |
| 1989 { | |
| 1990 return ((Hunspell*)pHunspell)->spell(word); | |
| 1991 } | |
| 1992 | |
| 1993 char *Hunspell_get_dic_encoding(Hunhandle *pHunspell) | |
| 1994 { | |
| 1995 return ((Hunspell*)pHunspell)->get_dic_encoding(); | |
| 1996 } | |
| 1997 | |
| 1998 int Hunspell_suggest(Hunhandle *pHunspell, char*** slst, const char * word) | |
| 1999 { | |
| 2000 return ((Hunspell*)pHunspell)->suggest(slst, word); | |
| 2001 } | |
| 2002 | |
| 2003 int Hunspell_analyze(Hunhandle *pHunspell, char*** slst, const char * word) | |
| 2004 { | |
| 2005 return ((Hunspell*)pHunspell)->analyze(slst, word); | |
| 2006 } | |
| 2007 | |
| 2008 int Hunspell_stem(Hunhandle *pHunspell, char*** slst, const char * word) | |
| 2009 { | |
| 2010 return ((Hunspell*)pHunspell)->stem(slst, word); | |
| 2011 } | |
| 2012 | |
| 2013 int Hunspell_stem2(Hunhandle *pHunspell, char*** slst, char** desc, int n) | |
| 2014 { | |
| 2015 return ((Hunspell*)pHunspell)->stem(slst, desc, n); | |
| 2016 } | |
| 2017 | |
| 2018 int Hunspell_generate(Hunhandle *pHunspell, char*** slst, const char * word, | |
| 2019 const char * word2) | |
| 2020 { | |
| 2021 return ((Hunspell*)pHunspell)->generate(slst, word, word2); | |
| 2022 } | |
| 2023 | |
| 2024 int Hunspell_generate2(Hunhandle *pHunspell, char*** slst, const char * word, | |
| 2025 char** desc, int n) | |
| 2026 { | |
| 2027 return ((Hunspell*)pHunspell)->generate(slst, word, desc, n); | |
| 2028 } | |
| 2029 | |
| 2030 /* functions for run-time modification of the dictionary */ | |
| 2031 | |
| 2032 /* add word to the run-time dictionary */ | |
| 2033 | |
| 2034 int Hunspell_add(Hunhandle *pHunspell, const char * word) { | |
| 2035 return ((Hunspell*)pHunspell)->add(word); | |
| 2036 } | |
| 2037 | |
| 2038 /* add word to the run-time dictionary with affix flags of | |
| 2039 * the example (a dictionary word): Hunspell will recognize | |
| 2040 * affixed forms of the new word, too. | |
| 2041 */ | |
| 2042 | |
| 2043 int Hunspell_add_with_affix(Hunhandle *pHunspell, const char * word, | |
| 2044 const char * example) { | |
| 2045 return ((Hunspell*)pHunspell)->add_with_affix(word, example); | |
| 2046 } | |
| 2047 | |
| 2048 /* remove word from the run-time dictionary */ | |
| 2049 | |
| 2050 int Hunspell_remove(Hunhandle *pHunspell, const char * word) { | |
| 2051 return ((Hunspell*)pHunspell)->remove(word); | |
| 2052 } | |
| 2053 | |
| 2054 void Hunspell_free_list(Hunhandle *, char *** slst, int n) { | |
| 2055 freelist(slst, n); | |
| 2056 } | |
| OLD | NEW |
|---|
Chromium Code Reviews
Issue 1135173004:
Rename third_party/hunspell_new back to third_party/hunspell. (Closed)
Base URL: https://chromium.googlesource.com/chromium/src.git@master