iconv/gconv_charset.h

   1 /* Charset name normalization.
   2    Copyright (C) 2001-2020 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4    Contributed by Ulrich Drepper <drepper@cygnus.com>, 2001.
   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    <https://www.gnu.org/licenses/>.  */
  19
  20 #include <ctype.h>
  21 #include <locale.h>
  22 #include <stdbool.h>
  23 #include <string.h>
  24 #include <sys/stat.h>
  25 #include <stdlib.h>
  26 #include "gconv_int.h"
  27
  28
  29 /* An iconv encoding is in the form of a triplet, with parts separated by
  30    a '/' character.  The first part is the standard name, the second part is
  31    the character set, and the third part is the error handler.  If the first
  32    part is sufficient to identify both the standard and the character set
  33    then the second part can be empty e.g. UTF-8//.  If the first part is not
  34    sufficient to identify both the standard and the character set then the
  35    second part is required e.g. ISO-10646/UTF8/.  If neither the first or
  36    second parts are provided e.g. //, then the current locale is used.
  37    The actual values used in the first and second parts are not entirely
  38    relevant to the implementation.  The values themselves are used in a hash
  39    table to lookup modules and so the naming convention of the first two parts
  40    is somewhat arbitrary and only helps locate the entries in the cache.
  41    The third part is the error handler and is comprised of a ',' or '/'
  42    separated list of suffixes.  Currently, we support "TRANSLIT" for
  43    transliteration and "IGNORE" for ignoring conversion errors due to
  44    unrecognized input characters.  */
  45 #define GCONV_TRIPLE_SEPARATOR "/"
  46 #define GCONV_SUFFIX_SEPARATOR ","
  47 #define GCONV_TRANSLIT_SUFFIX "TRANSLIT"
  48 #define GCONV_IGNORE_ERRORS_SUFFIX "IGNORE"
  49
  50
  51 /* This function accepts the charset names of the source and destination of the
  52    conversion and populates *conv_spec with an equivalent conversion
  53    specification that may later be used by __gconv_open.  The charset names
  54    might contain options in the form of suffixes that alter the conversion,
  55    e.g. "ISO-10646/UTF-8/TRANSLIT".  It processes the charset names, ignoring
  56    and truncating any suffix options in fromcode, and processing and truncating
  57    any suffix options in tocode.  Supported suffix options ("TRANSLIT" or
  58    "IGNORE") when found in tocode lead to the corresponding flag in *conv_spec
  59    to be set to true.  Unrecognized suffix options are silently discarded.  If
  60    the function succeeds, it returns conv_spec back to the caller.  It returns
  61    NULL upon failure.  */
  62 struct gconv_spec *
  63 __gconv_create_spec (struct gconv_spec *conv_spec, const char *fromcode,
  64                      const char *tocode);
  65 libc_hidden_proto (__gconv_create_spec)
  66
  67
  68 /* This function frees all heap memory allocated by __gconv_create_spec.  */
  69 static void __attribute__ ((unused))
  70 gconv_destroy_spec (struct gconv_spec *conv_spec)
  71 {
  72   free (conv_spec->fromcode);
  73   free (conv_spec->tocode);
  74   return;
  75 }
  76
  77
  78 /* This function copies in-order, characters from the source 's' that are
  79    either alpha-numeric or one in one of these: "_-.,:/" - into the destination
  80    'wp' while dropping all other characters.  In the process, it converts all
  81    alphabetical characters to upper case.  It then appends up to two '/'
  82    characters so that the total number of '/'es in the destination is 2.  */
  83 static inline void __attribute__ ((unused, always_inline))
  84 strip (char *wp, const char *s)
  85 {
  86   int slash_count = 0;
  87
  88   while (*s != '\0')
  89     {
  90       if (__isalnum_l (*s, _nl_C_locobj_ptr)
  91           || *s == '_' || *s == '-' || *s == '.' || *s == ',' || *s == ':')
  92         *wp++ = __toupper_l (*s, _nl_C_locobj_ptr);
  93       else if (*s == '/')
  94         {
  95           if (++slash_count == 3)
  96             break;
  97           *wp++ = '/';
  98         }
  99       ++s;
 100     }
 101
 102   while (slash_count++ < 2)
 103     *wp++ = '/';
 104
 105   *wp = '\0';
 106 }
 107
 108
 109 static inline char * __attribute__ ((unused, always_inline))
 110 upstr (char *dst, const char *str)
 111 {
 112   char *cp = dst;
 113   while ((*cp++ = __toupper_l (*str++, _nl_C_locobj_ptr)) != '\0')
 114     /* nothing */;
 115   return dst;
 116 }