wcsmbs/wcsmbsload.c

   1 /* Copyright (C) 1998, 1999 Free Software Foundation, Inc.
   2    This file is part of the GNU C Library.
   3    Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998.
   4
   5    The GNU C Library is free software; you can redistribute it and/or
   6    modify it under the terms of the GNU Library General Public License as
   7    published by the Free Software Foundation; either version 2 of the
   8    License, or (at your option) any later version.
   9
  10    The GNU C Library is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13    Library General Public License for more details.
  14
  15    You should have received a copy of the GNU Library General Public
  16    License along with the GNU C Library; see the file COPYING.LIB.  If not,
  17    write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  18    Boston, MA 02111-1307, USA.  */
  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 #include <iconv/gconv_int.h>
  30
  31
  32 /* Last loaded locale for LC_CTYPE.  We initialize for the C locale
  33    which is enabled at startup.  */
  34 extern const struct locale_data _nl_C_LC_CTYPE;
  35 const struct locale_data *__wcsmbs_last_locale = &_nl_C_LC_CTYPE;
  36
  37
  38 /* These are the descriptions for the default conversion functions.  */
  39 static struct gconv_step to_wc =
  40 {
  41   shlib_handle: NULL,
  42   modname: NULL,
  43   counter: INT_MAX,
  44   from_name: "ANSI_X3.4-1968//",
  45   to_name: "INTERNAL",
  46   fct: __gconv_transform_ascii_internal,
  47   init_fct: NULL,
  48   end_fct: NULL,
  49   min_needed_from: 1,
  50   max_needed_from: 1,
  51   min_needed_to: 4,
  52   max_needed_to: 4,
  53   stateful: 0,
  54   data: NULL
  55 };
  56
  57 static struct gconv_step to_mb =
  58 {
  59   shlib_handle: NULL,
  60   modname: NULL,
  61   counter: INT_MAX,
  62   from_name: "INTERNAL",
  63   to_name: "ANSI_X3.4-1968//",
  64   fct: __gconv_transform_internal_ascii,
  65   init_fct: NULL,
  66   end_fct: NULL,
  67   min_needed_from: 4,
  68   max_needed_from: 4,
  69   min_needed_to: 1,
  70   max_needed_to: 1,
  71   stateful: 0,
  72   data: NULL
  73 };
  74
  75
  76 /* For the default locale we only have to handle ANSI_X3.4-1968.  */
  77 struct gconv_fcts __wcsmbs_gconv_fcts =
  78 {
  79   towc: &to_wc,
  80   tomb: &to_mb
  81 };
  82
  83
  84 static inline struct gconv_step *
  85 getfct (const char *to, const char *from)
  86 {
  87   size_t nsteps;
  88   struct gconv_step *result;
  89
  90   if (__gconv_find_transform (to, from, &result, &nsteps) != GCONV_OK)
  91     /* Loading the conversion step is not possible.  */
  92     return NULL;
  93
  94   /* We must only have one step in this conversion.  */
  95   if (nsteps != 1)
  96     return NULL;
  97
  98   return result;
  99 }
 100
 101
 102 /* Extract from the given locale name the character set portion.  Since
 103    only the XPG form of the name includes this information we don't have
 104    to take care for the CEN form.  */
 105 #define extract_charset_name(str) \
 106   ({                                                                          \
 107     const char *cp = str;                                                     \
 108     char *result = NULL;                                                      \
 109                                                                               \
 110     cp += strcspn (cp, "@.+,");                                               \
 111     if (*cp == '.')                                                           \
 112       {                                                                       \
 113         const char *endp = ++cp;                                              \
 114         while (*endp != '\0' && *endp != '@')                                 \
 115           ++endp;                                                             \
 116         if (endp != cp)                                                       \
 117           result = strndupa (cp, endp - cp);                                  \
 118       }                                                                       \
 119     result;                                                                   \
 120   })
 121
 122
 123 /* The gconv functions expects the name to be complete, including the
 124    trailing shashes if necessary.  */
 125 #define norm_add_slashes(str) \
 126   ({                                                                          \
 127     const char *cp = str;                                                     \
 128     char *result;                                                             \
 129     char *tmp;                                                                \
 130     size_t cnt = 0;                                                           \
 131                                                                               \
 132     while (*cp != '\0')                                                       \
 133       if (*cp++ == '/')                                                       \
 134         ++cnt;                                                                \
 135                                                                               \
 136     tmp = result = alloca (cp - str + 3);                                     \
 137     cp = str;                                                                 \
 138     while (*cp != '\0')                                                       \
 139       *tmp++ = _toupper (*cp++);                                              \
 140     if (cnt < 2)                                                              \
 141       {                                                                       \
 142         *tmp++ = '/';                                                         \
 143         if (cnt < 1)                                                          \
 144           *tmp++ = '/';                                                       \
 145       }                                                                       \
 146     *tmp = '\0';                                                              \
 147     result;                                                                   \
 148   })
 149
 150
 151 /* Load conversion functions for the currently selected locale.  */
 152 void
 153 internal_function
 154 __wcsmbs_load_conv (const struct locale_data *new_category)
 155 {
 156   /* We must modify global data.  */
 157   __libc_lock_define_initialized (static, lock)
 158
 159   /* Acquire the lock.  */
 160   __libc_lock_lock (lock);
 161
 162   /* We should repest the test since while we waited some other thread
 163      might have run this function.  */
 164   if (__wcsmbs_last_locale != new_category)
 165     {
 166       if (new_category->name == _nl_C_name)     /* Yes, pointer comparison.  */
 167         {
 168         failed:
 169           __wcsmbs_gconv_fcts.towc = &to_wc;
 170           __wcsmbs_gconv_fcts.tomb = &to_mb;
 171         }
 172       else
 173         {
 174           /* We must find the real functions.  */
 175           const char *charset_name;
 176           const char *complete_name;
 177
 178           /* Get name of charset of the locale.  We first examine
 179              whether we have a character set mentioned in the locale
 180              name.  If this isn't the case we use the information from
 181              the locale files.  */
 182           charset_name = extract_charset_name (setlocale (LC_CTYPE, NULL));
 183           if (charset_name == NULL)
 184             charset_name =
 185               new_category->values[_NL_ITEM_INDEX(CODESET)].string;
 186
 187           /* Normalize the name and add the slashes necessary for a
 188              complete lookup.  */
 189           complete_name = norm_add_slashes (charset_name);
 190
 191           __wcsmbs_gconv_fcts.tomb = getfct (complete_name, "INTERNAL");
 192           __wcsmbs_gconv_fcts.towc = getfct ("INTERNAL", complete_name);
 193
 194           /* If any of the conversion functions is not available we don't
 195              use any since this would mean we cannot convert back and
 196              forth.*/
 197           if (__wcsmbs_gconv_fcts.towc == NULL
 198               || __wcsmbs_gconv_fcts.tomb == NULL)
 199             goto failed;
 200         }
 201
 202       /* Set last-used variable for current locale.  */
 203       __wcsmbs_last_locale = new_category;
 204     }
 205
 206   __libc_lock_unlock (lock);
 207 }