wcsmbs/wcsmbsload.c

   1 /* Copyright (C) 1998-2002,2004,2005,2008,2010,2011,2012
   2    Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998.
   5
   6    The GNU C Library is free software; you can redistribute it and/or
   7    modify it under the terms of the GNU Lesser General Public
   8    License as published by the Free Software Foundation; either
   9    version 2.1 of the License, or (at your option) any later version.
  10
  11    The GNU C Library is distributed in the hope that it will be useful,
  12    but WITHOUT ANY WARRANTY; without even the implied warranty of
  13    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14    Lesser General Public License for more details.
  15
  16    You should have received a copy of the GNU Lesser General Public
  17    License along with the GNU C Library; if not, see
  18    <http://www.gnu.org/licenses/>.  */
  19
  20 #include <ctype.h>
  21 #include <langinfo.h>
  22 #include <limits.h>
  23 #include <stdlib.h>
  24 #include <string.h>
  25
  26 #include <locale/localeinfo.h>
  27 #include <wcsmbsload.h>
  28 #include <bits/libc-lock.h>
  29
  30
  31 /* These are the descriptions for the default conversion functions.  */
  32 static const struct __gconv_step to_wc =
  33 {
  34   .__shlib_handle = NULL,
  35   .__modname = NULL,
  36   .__counter = INT_MAX,
  37   .__from_name = (char *) "ANSI_X3.4-1968//TRANSLIT",
  38   .__to_name = (char *) "INTERNAL",
  39   .__fct = __gconv_transform_ascii_internal,
  40   .__btowc_fct = __gconv_btwoc_ascii,
  41   .__init_fct = NULL,
  42   .__end_fct = NULL,
  43   .__min_needed_from = 1,
  44   .__max_needed_from = 1,
  45   .__min_needed_to = 4,
  46   .__max_needed_to = 4,
  47   .__stateful = 0,
  48   .__data = NULL
  49 };
  50
  51 static const struct __gconv_step to_mb =
  52 {
  53   .__shlib_handle = NULL,
  54   .__modname = NULL,
  55   .__counter = INT_MAX,
  56   .__from_name = (char *) "INTERNAL",
  57   .__to_name = (char *) "ANSI_X3.4-1968//TRANSLIT",
  58   .__fct = __gconv_transform_internal_ascii,
  59   .__btowc_fct = NULL,
  60   .__init_fct = NULL,
  61   .__end_fct = NULL,
  62   .__min_needed_from = 4,
  63   .__max_needed_from = 4,
  64   .__min_needed_to = 1,
  65   .__max_needed_to = 1,
  66   .__stateful = 0,
  67   .__data = NULL
  68 };
  69
  70
  71 /* For the default locale we only have to handle ANSI_X3.4-1968.  */
  72 const struct gconv_fcts __wcsmbs_gconv_fcts_c =
  73 {
  74   .towc = (struct __gconv_step *) &to_wc,
  75   .towc_nsteps = 1,
  76   .tomb = (struct __gconv_step *) &to_mb,
  77   .tomb_nsteps = 1,
  78 };
  79
  80
  81 attribute_hidden
  82 struct __gconv_step *
  83 __wcsmbs_getfct (const char *to, const char *from, size_t *nstepsp)
  84 {
  85   size_t nsteps;
  86   struct __gconv_step *result;
  87 #if 0
  88   size_t nstateful;
  89   size_t cnt;
  90 #endif
  91
  92   if (__gconv_find_transform (to, from, &result, &nsteps, 0) != __GCONV_OK)
  93     /* Loading the conversion step is not possible.  */
  94     return NULL;
  95
  96   /* Maybe it is someday necessary to allow more than one step.
  97      Currently this is not the case since the conversions handled here
  98      are from and to INTERNAL and there always is a converted for
  99      that.  It the directly following code is enabled the libio
 100      functions will have to allocate appropriate __gconv_step_data
 101      elements instead of only one.  */
 102 #if 0
 103   /* Count the number of stateful conversions.  Since we will only
 104      have one 'mbstate_t' object available we can only deal with one
 105      stateful conversion.  */
 106   nstateful = 0;
 107   for (cnt = 0; cnt < nsteps; ++cnt)
 108     if (result[cnt].__stateful)
 109       ++nstateful;
 110   if (nstateful > 1)
 111 #else
 112   if (nsteps > 1)
 113 #endif
 114     {
 115       /* We cannot handle this case.  */
 116       __gconv_close_transform (result, nsteps);
 117       result = NULL;
 118     }
 119   else
 120     *nstepsp = nsteps;
 121
 122   return result;
 123 }
 124
 125
 126 /* Extract from the given locale name the character set portion.  Since
 127    only the XPG form of the name includes this information we don't have
 128    to take care for the CEN form.  */
 129 #define extract_charset_name(str) \
 130   ({                                                                          \
 131     const char *cp = str;                                                     \
 132     char *result = NULL;                                                      \
 133                                                                               \
 134     cp += strcspn (cp, "@.+,");                                               \
 135     if (*cp == '.')                                                           \
 136       {                                                                       \
 137         const char *endp = ++cp;                                              \
 138         while (*endp != '\0' && *endp != '@')                                 \
 139           ++endp;                                                             \
 140         if (endp != cp)                                                       \
 141           result = strndupa (cp, endp - cp);                                  \
 142       }                                                                       \
 143     result;                                                                   \
 144   })
 145
 146
 147 /* Some of the functions here must not be used while setlocale is called.  */
 148 __libc_rwlock_define (extern, __libc_setlocale_lock attribute_hidden)
 149
 150 /* Load conversion functions for the currently selected locale.  */
 151 void
 152 internal_function
 153 __wcsmbs_load_conv (struct __locale_data *new_category)
 154 {
 155   /* Acquire the lock.  */
 156   __libc_rwlock_wrlock (__libc_setlocale_lock);
 157
 158   /* We should repeat the test since while we waited some other thread
 159      might have run this function.  */
 160   if (__builtin_expect (new_category->private.ctype == NULL, 1))
 161     {
 162       /* We must find the real functions.  */
 163       const char *charset_name;
 164       const char *complete_name;
 165       struct gconv_fcts *new_fcts;
 166       int use_translit;
 167
 168       /* Allocate the gconv_fcts structure.  */
 169       new_fcts = calloc (1, sizeof *new_fcts);
 170       if (new_fcts == NULL)
 171         goto failed;
 172
 173       /* Get name of charset of the locale.  */
 174       charset_name = new_category->values[_NL_ITEM_INDEX(CODESET)].string;
 175
 176       /* Does the user want transliteration?  */
 177       use_translit = new_category->use_translit;
 178
 179       /* Normalize the name and add the slashes necessary for a
 180          complete lookup.  */
 181       complete_name = norm_add_slashes (charset_name,
 182                                         use_translit ? "TRANSLIT" : "");
 183
 184       /* It is not necessary to use transliteration in this direction
 185          since the internal character set is supposed to be able to
 186          represent all others.  */
 187       new_fcts->towc = __wcsmbs_getfct ("INTERNAL", complete_name,
 188                                         &new_fcts->towc_nsteps);
 189       if (new_fcts->towc != NULL)
 190         new_fcts->tomb = __wcsmbs_getfct (complete_name, "INTERNAL",
 191                                           &new_fcts->tomb_nsteps);
 192
 193       /* If any of the conversion functions is not available we don't
 194          use any since this would mean we cannot convert back and
 195          forth.  NB: NEW_FCTS was allocated with calloc.  */
 196       if (new_fcts->tomb == NULL)
 197         {
 198           if (new_fcts->towc != NULL)
 199             __gconv_close_transform (new_fcts->towc, new_fcts->towc_nsteps);
 200
 201           free (new_fcts);
 202
 203         failed:
 204           new_category->private.ctype = &__wcsmbs_gconv_fcts_c;
 205         }
 206       else
 207         {
 208           new_category->private.ctype = new_fcts;
 209           new_category->private.cleanup = &_nl_cleanup_ctype;
 210         }
 211     }
 212
 213   __libc_rwlock_unlock (__libc_setlocale_lock);
 214 }
 215
 216
 217 /* Clone the current conversion function set.  */
 218 void
 219 internal_function
 220 __wcsmbs_clone_conv (struct gconv_fcts *copy)
 221 {
 222   const struct gconv_fcts *orig;
 223
 224   orig = get_gconv_fcts (_NL_CURRENT_DATA (LC_CTYPE));
 225
 226   /* Copy the data.  */
 227   *copy = *orig;
 228
 229   /* Now increment the usage counters.
 230      Note: This assumes copy->*_nsteps == 1.  */
 231   if (copy->towc->__shlib_handle != NULL)
 232     ++copy->towc->__counter;
 233   if (copy->tomb->__shlib_handle != NULL)
 234     ++copy->tomb->__counter;
 235 }
 236
 237
 238 /* Get converters for named charset.  */
 239 int
 240 internal_function
 241 __wcsmbs_named_conv (struct gconv_fcts *copy, const char *name)
 242 {
 243   copy->towc = __wcsmbs_getfct ("INTERNAL", name, &copy->towc_nsteps);
 244   if (copy->towc == NULL)
 245     return 1;
 246
 247   copy->tomb = __wcsmbs_getfct (name, "INTERNAL", &copy->tomb_nsteps);
 248   if (copy->tomb == NULL)
 249     {
 250       __gconv_close_transform (copy->towc, copy->towc_nsteps);
 251       return 1;
 252     }
 253
 254   return 0;
 255 }
 256
 257 void internal_function
 258 _nl_cleanup_ctype (struct __locale_data *locale)
 259 {
 260   const struct gconv_fcts *const data = locale->private.ctype;
 261   if (data != NULL)
 262     {
 263       locale->private.ctype = NULL;
 264       locale->private.cleanup = NULL;
 265
 266       /* Free the old conversions.  */
 267       __gconv_close_transform (data->tomb, data->tomb_nsteps);
 268       __gconv_close_transform (data->towc, data->towc_nsteps);
 269       free ((char *) data);
 270     }
 271 }