locale/findlocale.c

   1 /* Copyright (C) 1996-2023 Free Software Foundation, Inc.
   2    This file is part of the GNU C Library.
   3
   4    The GNU C Library is free software; you can redistribute it and/or
   5    modify it under the terms of the GNU Lesser General Public
   6    License as published by the Free Software Foundation; either
   7    version 2.1 of the License, or (at your option) any later version.
   8
   9    The GNU C Library is distributed in the hope that it will be useful,
  10    but WITHOUT ANY WARRANTY; without even the implied warranty of
  11    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  12    Lesser General Public License for more details.
  13
  14    You should have received a copy of the GNU Lesser General Public
  15    License along with the GNU C Library; if not, see
  16    <https://www.gnu.org/licenses/>.  */
  17
  18 #include <assert.h>
  19 #include <errno.h>
  20 #include <locale.h>
  21 #include <stdlib.h>
  22 #include <string.h>
  23 #include <unistd.h>
  24 #ifdef _POSIX_MAPPED_FILES
  25 # include <sys/mman.h>
  26 #endif
  27
  28 #include "localeinfo.h"
  29 #include "../iconv/gconv_charset.h"
  30 #include "../iconv/gconv_int.h"
  31
  32
  33 #ifdef NL_CURRENT_INDIRECT
  34 # define DEFINE_CATEGORY(category, category_name, items, a) \
  35 extern struct __locale_data _nl_C_##category; \
  36 weak_extern (_nl_C_##category)
  37 # include "categories.def"
  38 # undef DEFINE_CATEGORY
  39
  40 /* Array indexed by category of pointers to _nl_C_CATEGORY slots.
  41    Elements are zero for categories whose data is never used.  */
  42 struct __locale_data *const _nl_C[] attribute_hidden =
  43   {
  44 # define DEFINE_CATEGORY(category, category_name, items, a) \
  45     [category] = &_nl_C_##category,
  46 # include "categories.def"
  47 # undef DEFINE_CATEGORY
  48   };
  49 #else
  50 # define _nl_C          (_nl_C_locobj.__locales)
  51 #endif
  52
  53
  54 /* For each category we keep a list of records for the locale files
  55    which are somehow addressed.  */
  56 struct loaded_l10nfile *_nl_locale_file_list[__LC_LAST];
  57
  58 const char _nl_default_locale_path[] attribute_hidden = COMPLOCALEDIR;
  59
  60 /* Checks if the name is actually present, that is, not NULL and not
  61    empty.  */
  62 static inline int
  63 name_present (const char *name)
  64 {
  65   return name != NULL && name[0] != '\0';
  66 }
  67
  68 /* Checks that the locale name neither extremely long, nor contains a
  69    ".." path component (to prevent directory traversal).  */
  70 static inline int
  71 valid_locale_name (const char *name)
  72 {
  73   /* Not set.  */
  74   size_t namelen = strlen (name);
  75   /* Name too long.  The limit is arbitrary and prevents stack overflow
  76      issues later.  */
  77   if (__glibc_unlikely (namelen > 255))
  78     return 0;
  79   /* Directory traversal attempt.  */
  80   static const char slashdot[4] = {'/', '.', '.', '/'};
  81   if (__glibc_unlikely (__memmem (name, namelen,
  82                                   slashdot, sizeof (slashdot)) != NULL))
  83     return 0;
  84   if (namelen == 2 && __glibc_unlikely (name[0] == '.' && name [1] == '.'))
  85     return 0;
  86   if (namelen >= 3
  87       && __glibc_unlikely (((name[0] == '.'
  88                              && name[1] == '.'
  89                              && name[2] == '/')
  90                             || (name[namelen - 3] == '/'
  91                                 && name[namelen - 2] == '.'
  92                                 && name[namelen - 1] == '.'))))
  93     return 0;
  94   /* If there is a slash in the name, it must start with one.  */
  95   if (__glibc_unlikely (memchr (name, '/', namelen) != NULL) && name[0] != '/')
  96     return 0;
  97   return 1;
  98 }
  99
 100 struct __locale_data *
 101 _nl_find_locale (const char *locale_path, size_t locale_path_len,
 102                  int category, const char **name)
 103 {
 104   int mask;
 105   /* Name of the locale for this category.  */
 106   const char *cloc_name = *name;
 107   const char *language;
 108   const char *modifier;
 109   const char *territory;
 110   const char *codeset;
 111   const char *normalized_codeset;
 112   struct loaded_l10nfile *locale_file;
 113
 114   if (cloc_name[0] == '\0')
 115     {
 116       /* The user decides which locale to use by setting environment
 117          variables.  */
 118       cloc_name = getenv ("LC_ALL");
 119       if (!name_present (cloc_name))
 120         cloc_name = getenv (_nl_category_names_get (category));
 121       if (!name_present (cloc_name))
 122         cloc_name = getenv ("LANG");
 123       if (!name_present (cloc_name))
 124         cloc_name = _nl_C_name;
 125     }
 126
 127   /* We used to fall back to the C locale if the name contains a slash
 128      character '/', but we now check for directory traversal in
 129      valid_locale_name, so this is no longer necessary.  */
 130
 131   if (__builtin_expect (strcmp (cloc_name, _nl_C_name), 1) == 0
 132       || __builtin_expect (strcmp (cloc_name, _nl_POSIX_name), 1) == 0)
 133     {
 134       /* We need not load anything.  The needed data is contained in
 135          the library itself.  */
 136       *name = _nl_C_name;
 137       return _nl_C[category];
 138     }
 139   else if (!valid_locale_name (cloc_name))
 140     {
 141       __set_errno (EINVAL);
 142       return NULL;
 143     }
 144
 145   *name = cloc_name;
 146
 147   /* We really have to load some data.  First we try the archive,
 148      but only if there was no LOCPATH environment variable specified.  */
 149   if (__glibc_likely (locale_path == NULL))
 150     {
 151       struct __locale_data *data
 152         = _nl_load_locale_from_archive (category, name);
 153       if (__glibc_likely (data != NULL))
 154         return data;
 155
 156       /* Nothing in the archive with the given name.  Expanding it as
 157          an alias and retry.  */
 158       cloc_name = _nl_expand_alias (*name);
 159       if (cloc_name != NULL)
 160         {
 161           data = _nl_load_locale_from_archive (category, &cloc_name);
 162           if (__builtin_expect (data != NULL, 1))
 163             return data;
 164         }
 165
 166       /* Nothing in the archive.  Set the default path to search below.  */
 167       locale_path = _nl_default_locale_path;
 168       locale_path_len = sizeof _nl_default_locale_path;
 169     }
 170   else
 171     /* We really have to load some data.  First see whether the name is
 172        an alias.  Please note that this makes it impossible to have "C"
 173        or "POSIX" as aliases.  */
 174     cloc_name = _nl_expand_alias (*name);
 175
 176   if (cloc_name == NULL)
 177     /* It is no alias.  */
 178     cloc_name = *name;
 179
 180   /* Make a writable copy of the locale name.  */
 181   char *loc_name = strdupa (cloc_name);
 182
 183   /* LOCALE can consist of up to four recognized parts for the XPG syntax:
 184
 185                 language[_territory[.codeset]][@modifier]
 186
 187      Beside the first all of them are allowed to be missing.  If the
 188      full specified locale is not found, the less specific one are
 189      looked for.  The various part will be stripped off according to
 190      the following order:
 191                 (1) codeset
 192                 (2) normalized codeset
 193                 (3) territory
 194                 (4) modifier
 195    */
 196   mask = _nl_explode_name (loc_name, &language, &modifier, &territory,
 197                            &codeset, &normalized_codeset);
 198   if (mask == -1)
 199     /* Memory allocate problem.  */
 200     return NULL;
 201
 202   /* If exactly this locale was already asked for we have an entry with
 203      the complete name.  */
 204   locale_file = _nl_make_l10nflist (&_nl_locale_file_list[category],
 205                                     locale_path, locale_path_len, mask,
 206                                     language, territory, codeset,
 207                                     normalized_codeset, modifier,
 208                                     _nl_category_names_get (category), 0);
 209
 210   if (locale_file == NULL)
 211     {
 212       /* Find status record for addressed locale file.  We have to search
 213          through all directories in the locale path.  */
 214       locale_file = _nl_make_l10nflist (&_nl_locale_file_list[category],
 215                                         locale_path, locale_path_len, mask,
 216                                         language, territory, codeset,
 217                                         normalized_codeset, modifier,
 218                                         _nl_category_names_get (category), 1);
 219       if (locale_file == NULL)
 220         /* This means we are out of core.  */
 221         return NULL;
 222     }
 223
 224   /* The space for normalized_codeset is dynamically allocated.  Free it.  */
 225   if (mask & XPG_NORM_CODESET)
 226     free ((void *) normalized_codeset);
 227
 228   if (locale_file->decided == 0)
 229     _nl_load_locale (locale_file, category);
 230
 231   if (locale_file->data == NULL)
 232     {
 233       int cnt;
 234       for (cnt = 0; locale_file->successor[cnt] != NULL; ++cnt)
 235         {
 236           if (locale_file->successor[cnt]->decided == 0)
 237             _nl_load_locale (locale_file->successor[cnt], category);
 238           if (locale_file->successor[cnt]->data != NULL)
 239             break;
 240         }
 241       /* Move the entry we found (or NULL) to the first place of
 242          successors.  */
 243       locale_file->successor[0] = locale_file->successor[cnt];
 244       locale_file = locale_file->successor[cnt];
 245
 246       if (locale_file == NULL)
 247         return NULL;
 248     }
 249
 250   /* The LC_CTYPE category allows to check whether a locale is really
 251      usable.  If the locale name contains a charset name and the
 252      charset name used in the locale (present in the LC_CTYPE data) is
 253      not the same (after resolving aliases etc) we reject the locale
 254      since using it would irritate users expecting the charset named
 255      in the locale name.  */
 256   if (codeset != NULL)
 257     {
 258       /* Get the codeset information from the locale file.  */
 259       static const int codeset_idx[] =
 260         {
 261           [__LC_CTYPE] = _NL_ITEM_INDEX (CODESET),
 262           [__LC_NUMERIC] = _NL_ITEM_INDEX (_NL_NUMERIC_CODESET),
 263           [__LC_TIME] = _NL_ITEM_INDEX (_NL_TIME_CODESET),
 264           [__LC_COLLATE] = _NL_ITEM_INDEX (_NL_COLLATE_CODESET),
 265           [__LC_MONETARY] = _NL_ITEM_INDEX (_NL_MONETARY_CODESET),
 266           [__LC_MESSAGES] = _NL_ITEM_INDEX (_NL_MESSAGES_CODESET),
 267           [__LC_PAPER] = _NL_ITEM_INDEX (_NL_PAPER_CODESET),
 268           [__LC_NAME] = _NL_ITEM_INDEX (_NL_NAME_CODESET),
 269           [__LC_ADDRESS] = _NL_ITEM_INDEX (_NL_ADDRESS_CODESET),
 270           [__LC_TELEPHONE] = _NL_ITEM_INDEX (_NL_TELEPHONE_CODESET),
 271           [__LC_MEASUREMENT] = _NL_ITEM_INDEX (_NL_MEASUREMENT_CODESET),
 272           [__LC_IDENTIFICATION] = _NL_ITEM_INDEX (_NL_IDENTIFICATION_CODESET)
 273         };
 274       const struct __locale_data *data;
 275       const char *locale_codeset;
 276       char *clocale_codeset;
 277       char *ccodeset;
 278
 279       data = (const struct __locale_data *) locale_file->data;
 280       locale_codeset =
 281         (const char *) data->values[codeset_idx[category]].string;
 282       assert (locale_codeset != NULL);
 283       /* Note the length of the allocated memory: +3 for up to two slashes
 284          and the NUL byte.  */
 285       clocale_codeset = (char *) alloca (strlen (locale_codeset) + 3);
 286       strip (clocale_codeset, locale_codeset);
 287
 288       ccodeset = (char *) alloca (strlen (codeset) + 3);
 289       strip (ccodeset, codeset);
 290
 291       if (__gconv_compare_alias (upstr (ccodeset, ccodeset),
 292                                  upstr (clocale_codeset,
 293                                         clocale_codeset)) != 0)
 294         /* The codesets are not identical, don't use the locale.  */
 295         return NULL;
 296     }
 297
 298   /* Determine the locale name for which loading succeeded.  This
 299      information comes from the file name.  The form is
 300      <path>/<locale>/LC_foo.  We must extract the <locale> part.  */
 301   if (((const struct __locale_data *) locale_file->data)->name == NULL)
 302     {
 303       char *cp, *endp;
 304
 305       endp = strrchr (locale_file->filename, '/');
 306       cp = endp - 1;
 307       while (cp[-1] != '/')
 308         --cp;
 309       ((struct __locale_data *) locale_file->data)->name
 310         = __strndup (cp, endp - cp);
 311     }
 312
 313   /* Determine whether the user wants transliteration or not.  */
 314   if (modifier != NULL
 315       && __strcasecmp_l (modifier, "TRANSLIT", _nl_C_locobj_ptr) == 0)
 316     ((struct __locale_data *) locale_file->data)->use_translit = 1;
 317
 318   /* Increment the usage count.  */
 319   if (((const struct __locale_data *) locale_file->data)->usage_count
 320       < MAX_USAGE_COUNT)
 321     ++((struct __locale_data *) locale_file->data)->usage_count;
 322
 323   return (struct __locale_data *) locale_file->data;
 324 }
 325
 326
 327 /* Calling this function assumes the lock for handling global locale data
 328    is acquired.  */
 329 void
 330 _nl_remove_locale (int locale, struct __locale_data *data)
 331 {
 332   if (--data->usage_count == 0)
 333     {
 334       if (data->alloc != ld_archive)
 335         {
 336           /* First search the entry in the list of loaded files.  */
 337           struct loaded_l10nfile *ptr = _nl_locale_file_list[locale];
 338
 339           /* Search for the entry.  It must be in the list.  Otherwise it
 340              is a bug and we crash badly.  */
 341           while ((struct __locale_data *) ptr->data != data)
 342             ptr = ptr->next;
 343
 344           /* Mark the data as not available anymore.  So when the data has
 345              to be used again it is reloaded.  */
 346           ptr->decided = 0;
 347           ptr->data = NULL;
 348         }
 349
 350       /* This does the real work.  */
 351       _nl_unload_locale (locale, data);
 352     }
 353 }