locale/xlocale.c

   1 /*
   2  * Copyright (c) 2005, 2008 Apple Inc. All rights reserved.
   3  *
   4  * @APPLE_LICENSE_HEADER_START@
   5  *
   6  * This file contains Original Code and/or Modifications of Original Code
   7  * as defined in and that are subject to the Apple Public Source License
   8  * Version 2.0 (the 'License'). You may not use this file except in
   9  * compliance with the License. Please obtain a copy of the License at
  10  * http://www.opensource.apple.com/apsl/ and read it before using this
  11  * file.
  12  *
  13  * The Original Code and all software distributed under the License are
  14  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  15  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  16  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  17  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  18  * Please see the License for the specific language governing rights and
  19  * limitations under the License.
  20  *
  21  * @APPLE_LICENSE_HEADER_END@
  22  */
  23
  24 #include "xlocale_private.h"
  25 #include <errno.h>
  26 #include <stddef.h>
  27 #include <string.h>
  28 #include "ldpart.h"
  29
  30 #define NMBSTATET       10
  31 #define C_LOCALE_INITIALIZER    {       \
  32         0, NULL,                        \
  33         {}, {}, {}, {}, {},             \
  34         {}, {}, {}, {}, {},             \
  35         LOCK_INITIALIZER,               \
  36         XMAGIC,                         \
  37         1, 0, 0, 0, 0, 0, 1, 1, 0,      \
  38         NULL,                           \
  39         &_DefaultRuneXLocale,           \
  40 }
  41
  42 static char C[] = "C";
  43 static const struct _xlocale __c_locale = C_LOCALE_INITIALIZER;
  44 const locale_t _c_locale = (const locale_t)&__c_locale;
  45 __private_extern__ struct _xlocale __global_locale = C_LOCALE_INITIALIZER;
  46 __private_extern__ pthread_key_t __locale_key = (pthread_key_t)-1;
  47
  48 extern int __collate_load_tables(const char *, locale_t);
  49 extern int __detect_path_locale(void);
  50 extern const char *__get_locale_env(int);
  51 extern int __messages_load_locale(const char *, locale_t);
  52 extern int __monetary_load_locale(const char *, locale_t);
  53 extern int __numeric_load_locale(const char *, locale_t);
  54 extern int __setrunelocale(const char *, locale_t);
  55 extern int __time_load_locale(const char *, locale_t);
  56
  57 static void _releaselocale(locale_t loc);
  58
  59 /*
  60  * check that the encoding is the right size, isn't . or .. and doesn't
  61  * contain any slashes
  62  */
  63 static inline __attribute__((always_inline)) int
  64 _checkencoding(const char *encoding)
  65 {
  66         return (encoding && (strlen(encoding) > ENCODING_LEN
  67           || (encoding[0] == '.' && (encoding[1] == 0
  68           || (encoding[1] == '.' && encoding[2] == 0)))
  69           || strchr(encoding, '/') != NULL)) ? -1 : 0;
  70 }
  71
  72 /*
  73  * check that the locale has the right magic number
  74  */
  75 static inline __attribute__((always_inline)) int
  76 _checklocale(const locale_t loc)
  77 {
  78         if (!loc)
  79                 return 0;
  80         return (loc == LC_GLOBAL_LOCALE || loc->__magic == XMAGIC) ? 0 : -1;
  81 }
  82
  83 /*
  84  * copy a locale_t except anything before the magic value
  85  */
  86 static inline __attribute__((always_inline)) void
  87 _copylocale(locale_t dst, const locale_t src)
  88 {
  89         memcpy(&dst->__magic, &src->__magic, sizeof(*dst) - offsetof(struct _xlocale, __magic));
  90 }
  91
  92 /*
  93  * Make a copy of a locale_t, locking/unlocking the source.
  94  * A NULL locale_t means to make a copy of the current
  95  * locale while LC_GLOBAL_LOCALE means to copy the global locale.  If
  96  * &__c_locale is passed (meaning a C locale is desired), just make
  97  * a copy.
  98  */
  99 static locale_t
 100 _duplocale(locale_t loc)
 101 {
 102         locale_t new;
 103
 104         if ((new = (locale_t)malloc(sizeof(struct _xlocale))) == NULL)
 105                 return NULL;
 106         new->__refcount = 1;
 107         new->__free_extra = (__free_extra_t)_releaselocale;
 108         new->__lock = LOCK_INITIALIZER;
 109         if (loc == NULL)
 110                 loc = __current_locale();
 111         else if (loc == LC_GLOBAL_LOCALE)
 112                 loc = &__global_locale;
 113         else if (loc == &__c_locale) {
 114                 *new = __c_locale;
 115                 new->__refcount = 1;
 116                 new->__free_extra = (__free_extra_t)_releaselocale;
 117                 new->__lock = LOCK_INITIALIZER;
 118                 return new;
 119         }
 120         XL_LOCK(loc);
 121         _copylocale(new, loc);
 122         XL_UNLOCK(loc);
 123         /* __mbs_mblen is the first of NMBSTATET mbstate_t buffers */
 124         bzero(&new->__mbs_mblen, offsetof(struct _xlocale, __magic)
 125             - offsetof(struct _xlocale, __mbs_mblen));
 126         /* collate */
 127         XL_RETAIN(new->__lc_collate);
 128         /* ctype */
 129         XL_RETAIN(new->__lc_ctype);
 130         /* messages */
 131         XL_RETAIN(new->__lc_messages);
 132         /* monetary */
 133         XL_RETAIN(new->__lc_monetary);
 134         /* numeric */
 135         XL_RETAIN(new->__lc_numeric);
 136         XL_RETAIN(new->__lc_numeric_loc);
 137         /* time */
 138         XL_RETAIN(new->__lc_time);
 139
 140         return new;
 141 }
 142
 143 /*
 144  * Modify a locale_t, setting the parts specified in the mask
 145  * to the locale specified by the string.  If the string is NULL, the C
 146  * locale is used.  If the string is empty, the value is determined from
 147  * the environment.  -1 is returned on error, and loc is in a partial state.
 148  */
 149 static int
 150 _modifylocale(locale_t loc, int mask, __const char *locale)
 151 {
 152         int m, ret;
 153         const char *enc = NULL;
 154         char *oenc;
 155
 156         if (!locale)
 157                 locale = C;
 158
 159         ret = __detect_path_locale();
 160         if (ret) {
 161                 errno = ret;
 162                 return -1;
 163         }
 164
 165         if (*locale)
 166                 enc = locale;
 167         for(m = 1; m <= _LC_LAST_MASK; m <<= 1) {
 168                 if (m & mask) {
 169                         switch(m) {
 170                         case LC_COLLATE_MASK:
 171                                 if (!*locale) {
 172                                         enc = __get_locale_env(LC_COLLATE);
 173                                         if (_checkencoding(enc) < 0) {
 174                                                 errno = EINVAL;
 175                                                 return -1;
 176                                         }
 177                                 }
 178                                 oenc = (loc->__collate_load_error ? C : loc->__lc_collate->__encoding);
 179                                 if (strcmp(enc, oenc) != 0 && __collate_load_tables(enc, loc) == _LDP_ERROR)
 180                                         return -1;
 181                                 break;
 182                         case LC_CTYPE_MASK:
 183                                 if (!*locale) {
 184                                         enc = __get_locale_env(LC_CTYPE);
 185                                         if (_checkencoding(enc) < 0) {
 186                                                 errno = EINVAL;
 187                                                 return -1;
 188                                         }
 189                                 }
 190                                 if (strcmp(enc, loc->__lc_ctype->__ctype_encoding) != 0) {
 191                                         if ((ret = __setrunelocale(enc, loc)) != 0) {
 192                                                 errno = ret;
 193                                                 return -1;
 194                                         }
 195                                         if (loc->__numeric_fp_cvt == LC_NUMERIC_FP_SAME_LOCALE)
 196                                                 loc->__numeric_fp_cvt = LC_NUMERIC_FP_UNINITIALIZED;
 197                                 }
 198                                 break;
 199                         case LC_MESSAGES_MASK:
 200                                 if (!*locale) {
 201                                         enc = __get_locale_env(LC_MESSAGES);
 202                                         if (_checkencoding(enc) < 0) {
 203                                                 errno = EINVAL;
 204                                                 return -1;
 205                                         }
 206                                 }
 207                                 oenc = (loc->_messages_using_locale ? loc->__lc_messages->_messages_locale_buf : C);
 208                                 if (strcmp(enc, oenc) != 0 && __messages_load_locale(enc, loc) == _LDP_ERROR)
 209                                         return -1;
 210                                 break;
 211                         case LC_MONETARY_MASK:
 212                                 if (!*locale) {
 213                                         enc = __get_locale_env(LC_MONETARY);
 214                                         if (_checkencoding(enc) < 0) {
 215                                                 errno = EINVAL;
 216                                                 return -1;
 217                                         }
 218                                 }
 219                                 oenc = (loc->_monetary_using_locale ? loc->__lc_monetary->_monetary_locale_buf : C);
 220                                 if (strcmp(enc, oenc) != 0 && __monetary_load_locale(enc, loc) == _LDP_ERROR)
 221                                         return -1;
 222                                 break;
 223                         case LC_NUMERIC_MASK:
 224                                 if (!*locale) {
 225                                         enc = __get_locale_env(LC_NUMERIC);
 226                                         if (_checkencoding(enc) < 0) {
 227                                                 errno = EINVAL;
 228                                                 return -1;
 229                                         }
 230                                 }
 231                                 oenc = (loc->_numeric_using_locale ? loc->__lc_numeric->_numeric_locale_buf : C);
 232                                 if (strcmp(enc, oenc) != 0) {
 233                                         if (__numeric_load_locale(enc, loc) == _LDP_ERROR)
 234                                                 return -1;
 235                                         loc->__numeric_fp_cvt = LC_NUMERIC_FP_UNINITIALIZED;
 236                                         XL_RELEASE(loc->__lc_numeric_loc);
 237                                         loc->__lc_numeric_loc = NULL;
 238                                 }
 239                                 break;
 240                         case LC_TIME_MASK:
 241                                 if (!*locale) {
 242                                         enc = __get_locale_env(LC_TIME);
 243                                         if (_checkencoding(enc) < 0) {
 244                                                 errno = EINVAL;
 245                                                 return -1;
 246                                         }
 247                                 }
 248                                 oenc = (loc->_time_using_locale ? loc->__lc_time->_time_locale_buf : C);
 249                                 if (strcmp(enc, oenc) != 0 && __time_load_locale(enc, loc) == _LDP_ERROR)
 250                                         return -1;
 251                                 break;
 252                         }
 253                 }
 254         }
 255         return 0;
 256 }
 257
 258 /*
 259  * release all the memory objects (the memory will be freed when the refcount
 260  * becomes zero)
 261  */
 262 static void
 263 _releaselocale(locale_t loc)
 264 {
 265         /* collate */
 266         XL_RELEASE(loc->__lc_collate);
 267         /* ctype */
 268         XL_RELEASE(loc->__lc_ctype);
 269         /* messages */
 270         XL_RELEASE(loc->__lc_messages);
 271         /* monetary */
 272         XL_RELEASE(loc->__lc_monetary);
 273         /* numeric */
 274         XL_RELEASE(loc->__lc_numeric);
 275         XL_RELEASE(loc->__lc_numeric_loc);
 276         /* time */
 277         XL_RELEASE(loc->__lc_time);
 278 }
 279
 280 /*
 281  * EXTERNAL: Duplicate a (non-NULL) locale_t.  LC_GLOBAL_LOCALE means the
 282  * global locale, while NULL means the current locale.  NULL is returned
 283  * on error.
 284  */
 285 locale_t
 286 duplocale(locale_t loc)
 287 {
 288         if (_checklocale(loc) < 0) {
 289                 errno = EINVAL;
 290                 return NULL;
 291         }
 292         return _duplocale(loc);
 293 }
 294
 295 /*
 296  * EXTERNAL: Free a locale_t, releasing all memory objects.  Don't free
 297  * illegal locale_t's or the global locale.
 298  */
 299 int
 300 freelocale(locale_t loc)
 301 {
 302         if (!loc || _checklocale(loc) < 0 || loc == &__global_locale
 303             || loc == LC_GLOBAL_LOCALE || loc == &__c_locale) {
 304                 errno = EINVAL;
 305                 return -1;
 306         }
 307         XL_RELEASE(loc);
 308         return 0;
 309 }
 310
 311 /*
 312  * EXTERNAL: Create a new locale_t, based on the base locale_t, and modified
 313  * by the mask and locale string.  If the base is NULL, the current locale
 314  * is used as the base.  If locale is NULL, changes are made from the C locale
 315  * for categories set in mask.
 316  */
 317 locale_t
 318 newlocale(int mask, __const char *locale, locale_t base)
 319 {
 320         locale_t new;
 321         int lcmask = (mask & LC_ALL_MASK);
 322
 323         if (_checkencoding(locale) < 0) {
 324                 errno = EINVAL;
 325                 return NULL;
 326         }
 327         if (lcmask == LC_ALL_MASK)
 328                 base = (locale_t)&__c_locale;
 329         else if (_checklocale(base) < 0) {
 330                 errno = EINVAL;
 331                 return NULL;
 332         }
 333         new = _duplocale(base);
 334         if (new == NULL)
 335                 return NULL;
 336         if (lcmask == 0 || (lcmask == LC_ALL_MASK && locale == NULL))
 337                 return new;
 338         if (_modifylocale(new, lcmask, locale) < 0) {
 339                 freelocale(new);
 340                 return NULL;
 341         }
 342         return new;
 343 }
 344
 345 /*
 346  * PRIVATE EXTERNAL: Returns the locale that can be used by wcstod and
 347  * family, to convert the wide character string to a multi-byte string
 348  * (the LC_NUMERIC and LC_CTYPE locales may be different).
 349  */
 350 __private_extern__ locale_t
 351 __numeric_ctype(locale_t loc)
 352 {
 353         switch(loc->__numeric_fp_cvt) {
 354         case LC_NUMERIC_FP_UNINITIALIZED: {
 355                 const char *ctype = loc->__lc_ctype->__ctype_encoding;
 356                 const char *numeric = (loc->_numeric_using_locale ? loc->__lc_numeric->_numeric_locale_buf : C);
 357                 if (strcmp(ctype, numeric) == 0) {
 358                         loc->__numeric_fp_cvt = LC_NUMERIC_FP_SAME_LOCALE;
 359                         return loc;
 360                 } else {
 361                         loc->__lc_numeric_loc = newlocale(LC_CTYPE_MASK, numeric, (locale_t)&__c_locale);
 362                         if (loc->__lc_numeric_loc) {
 363                                 loc->__numeric_fp_cvt = LC_NUMERIC_FP_USE_LOCALE;
 364                                 return loc->__lc_numeric_loc;
 365                         } else { /* shouldn't happen, but just use the same locale */
 366                                 loc->__numeric_fp_cvt = LC_NUMERIC_FP_SAME_LOCALE;
 367                                 return loc;
 368                         }
 369                 }
 370         }
 371         case LC_NUMERIC_FP_SAME_LOCALE:
 372                 return loc;
 373         case LC_NUMERIC_FP_USE_LOCALE:
 374                 return loc->__lc_numeric_loc;
 375         }
 376         return loc;     /* shouldn't happen */
 377 }
 378
 379 /*
 380  * EXTERNAL: Returns the locale string for the part specified in mask.  The
 381  * least significant bit is used.  If loc is NULL, the current per-thread
 382  * locale is used.
 383  */
 384 const char *
 385 querylocale(int mask, locale_t loc)
 386 {
 387         int m;
 388         const char *ret;
 389
 390         if (_checklocale(loc) < 0 || (mask & LC_ALL_MASK) == 0) {
 391                 errno = EINVAL;
 392                 return NULL;
 393         }
 394         if (loc == NULL)
 395                 loc = __current_locale();
 396         else if (loc == LC_GLOBAL_LOCALE)
 397                 loc = &__global_locale;
 398         m = ffs(mask);
 399         if (m == 0 || m > _LC_NUM_MASK) {
 400                 errno = EINVAL;
 401                 return NULL;
 402         }
 403         XL_LOCK(loc);
 404         switch(1 << (m - 1)) {
 405         case LC_COLLATE_MASK:
 406                 ret = (loc->__collate_load_error ? C : loc->__lc_collate->__encoding);
 407                 break;
 408         case LC_CTYPE_MASK:
 409                 ret = loc->__lc_ctype->__ctype_encoding;
 410                 break;
 411         case LC_MESSAGES_MASK:
 412                 ret = (loc->_messages_using_locale ? loc->__lc_messages->_messages_locale_buf : C);
 413                 break;
 414         case LC_MONETARY_MASK:
 415                 ret = (loc->_monetary_using_locale ? loc->__lc_monetary->_monetary_locale_buf : C);
 416                 break;
 417         case LC_NUMERIC_MASK:
 418                 ret = (loc->_numeric_using_locale ? loc->__lc_numeric->_numeric_locale_buf : C);
 419                 break;
 420         case LC_TIME_MASK:
 421                 ret = (loc->_time_using_locale ? loc->__lc_time->_time_locale_buf : C);
 422                 break;
 423         default:
 424                 /* should never get here */
 425                 XL_UNLOCK(loc);
 426                 errno = EINVAL;
 427                 return NULL;
 428         }
 429         XL_UNLOCK(loc);
 430         return ret;
 431 }
 432
 433 /*
 434  * EXTERNAL: Set the thread-specific locale.  The previous locale is returned.
 435  * Use LC_GLOBAL_LOCALE to set the global locale.  LC_GLOBAL_LOCALE
 436  * may also be returned if there was no previous thread-specific locale in
 437  * effect.  If loc is NULL, the current locale is returned, but no locale
 438  * chance is made.  NULL is returned on error.
 439  */
 440 locale_t
 441 uselocale(locale_t loc)
 442 {
 443         locale_t orig;
 444
 445         if (loc == NULL)
 446                 orig = (locale_t)pthread_getspecific(__locale_key);
 447         else {
 448                 if (_checklocale(loc) < 0) {
 449                         errno = EINVAL;
 450                         return NULL;
 451                 }
 452                 if (loc == LC_GLOBAL_LOCALE ||
 453                     loc == &__global_locale)    /* should never happen */
 454                         loc = NULL;
 455                 XL_RETAIN(loc);
 456                 orig = pthread_getspecific(__locale_key);
 457                 pthread_setspecific(__locale_key, loc);
 458                 XL_RELEASE(orig);
 459         }
 460         return (orig ? orig : LC_GLOBAL_LOCALE);
 461 }
 462
 463 /*
 464  * EXTERNAL: Used by the MB_CUR_MAX macro to determine the thread-specific
 465  * value.
 466  */
 467 int
 468 ___mb_cur_max(void)
 469 {
 470         return __current_locale()->__lc_ctype->__mb_cur_max;
 471 }
 472
 473 /*
 474  * EXTERNAL: Used by the MB_CUR_MAX_L macro to determine the thread-specific
 475  * value, from the given locale_t.
 476  */
 477 int
 478 ___mb_cur_max_l(locale_t loc)
 479 {
 480         return __locale_ptr(loc)->__lc_ctype->__mb_cur_max;
 481 }
 482
 483 static void
 484 __xlocale_release(void *loc)
 485 {
 486         XL_RELEASE((locale_t)loc);
 487 }
 488
 489 /*
 490  * Called from the Libc initializer to setup the thread-specific key.
 491  */
 492 /*
 493  * Partition _pthread_keys in a lower part that dyld can use, and an upper
 494  * part for libSystem.  The libSystem part starts at __pthread_tsd_first = 10.
 495  * dyld will set this value to 1.
 496  */
 497
 498 __private_extern__ void
 499 __xlocale_init(void)
 500 {
 501         if (__locale_key == (pthread_key_t)-1) {
 502                 __locale_key = __LIBC_PTHREAD_KEY_XLOCALE;
 503                 pthread_key_init_np(__locale_key, __xlocale_release);
 504         }
 505 }
 506