[BACK]Return to localcharset.c CVS log [TXT] [DIR] Up to [ELWIX - Embedded LightWeight unIX -] / embedaddon / libiconv / libcharset / lib

Diff for /embedaddon/libiconv/libcharset/lib/localcharset.c between versions 1.1.1.2 and 1.1.1.3

version 1.1.1.2, 2012/05/29 09:29:42 version 1.1.1.3, 2021/03/17 13:38:46
Line 1 Line 1
 /* Determine a canonical name for the current locale's character encoding.  /* Determine a canonical name for the current locale's character encoding.
   
   Copyright (C) 2000-2006, 2008-2010 Free Software Foundation, Inc.   Copyright (C) 2000-2006, 2008-2018 Free Software Foundation, Inc.
   
    This program is free software; you can redistribute it and/or modify it     This program is free software; you can redistribute it and/or modify it
    under the terms of the GNU Library General Public License as published     under the terms of the GNU Library General Public License as published
Line 12 Line 12
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
    Library General Public License for more details.     Library General Public License for more details.
   
   You should have received a copy of the GNU Library General Public   You should have received a copy of the GNU Library General Public License
   License along with this program; if not, write to the Free Software   along with this program; if not, see <https://www.gnu.org/licenses/>.  */
   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, 
   USA.  */ 
   
 /* Written by Bruno Haible <bruno@clisp.org>.  */  /* Written by Bruno Haible <bruno@clisp.org>.  */
   
Line 24 Line 22
 /* Specification.  */  /* Specification.  */
 #include "localcharset.h"  #include "localcharset.h"
   
 #include <fcntl.h>  
 #include <stddef.h>  #include <stddef.h>
 #include <stdio.h>  #include <stdio.h>
 #include <string.h>  #include <string.h>
 #include <stdlib.h>  #include <stdlib.h>
   
 #if defined __APPLE__ && defined __MACH__ && HAVE_LANGINFO_CODESET  #if defined __APPLE__ && defined __MACH__ && HAVE_LANGINFO_CODESET
# define DARWIN7 /* Darwin 7 or newer, i.e. MacOS X 10.3 or newer */# define DARWIN7 /* Darwin 7 or newer, i.e. Mac OS X 10.3 or newer */
 #endif  #endif
   
#if defined _WIN32 || defined __WIN32__#if defined _WIN32 && !defined __CYGWIN__
# define WIN32_NATIVE# define WINDOWS_NATIVE
 # include <locale.h>
 #endif  #endif
   
 #if defined __EMX__  #if defined __EMX__
Line 45 Line 43
 # endif  # endif
 #endif  #endif
   
#if !defined WIN32_NATIVE#if !defined WINDOWS_NATIVE
# include <unistd.h> 
 # if HAVE_LANGINFO_CODESET  # if HAVE_LANGINFO_CODESET
 #  include <langinfo.h>  #  include <langinfo.h>
 # else  # else
#  if 0 /* see comment below */#  if 0 /* see comment regarding use of setlocale(), below */
 #   include <locale.h>  #   include <locale.h>
 #  endif  #  endif
 # endif  # endif
Line 58 Line 55
 #  define WIN32_LEAN_AND_MEAN  #  define WIN32_LEAN_AND_MEAN
 #  include <windows.h>  #  include <windows.h>
 # endif  # endif
#elif defined WIN32_NATIVE#elif defined WINDOWS_NATIVE
 # define WIN32_LEAN_AND_MEAN  # define WIN32_LEAN_AND_MEAN
 # include <windows.h>  # include <windows.h>
 #endif  #endif
Line 67 Line 64
 # include <os2.h>  # include <os2.h>
 #endif  #endif
   
#if ENABLE_RELOCATABLE/* For MB_CUR_MAX_L */
# include "relocatable.h"#if defined DARWIN7
#else# include <xlocale.h>
# define relocate(pathname) (pathname) 
 #endif  #endif
   
 /* Get LIBDIR.  */  
 #ifndef LIBDIR  
 # include "configmake.h"  
 #endif  
   
/* Define O_NOFOLLOW to 0 on platforms where it does not exist.  */#if HAVE_LANGINFO_CODESET || defined WINDOWS_NATIVE || defined OS2
#ifndef O_NOFOLLOW 
# define O_NOFOLLOW 0 
#endif 
   
#if defined _WIN32 || defined __WIN32__ || defined __CYGWIN__ || defined __EMX__ || defined __DJGPP__/* On these platforms, we use a mapping from non-canonical encoding name
  /* Win32, Cygwin, OS/2, DOS */   to GNU canonical encoding name.  */
# define ISSLASH(C) ((C) == '/' || (C) == '\\') 
#endif 
   
#ifndef DIRECTORY_SEPARATOR/* With glibc-2.1 or newer, we don't need any canonicalization,
# define DIRECTORY_SEPARATOR '/'   because glibc has iconv and both glibc and libiconv support all
#endif   GNU canonical names directly.  */
 # if !((defined __GNU_LIBRARY__ && __GLIBC__ >= 2) || defined __UCLIBC__)
   
#ifndef ISSLASHstruct table_entry
# define ISSLASH(C) ((C) == DIRECTORY_SEPARATOR) 
#endif 
 
#if HAVE_DECL_GETC_UNLOCKED 
# undef getc 
# define getc getc_unlocked 
#endif 
 
/* The following static variable is declared 'volatile' to avoid a 
   possible multithread problem in the function get_charset_aliases. If we 
   are running in a threaded environment, and if two threads initialize 
   'charset_aliases' simultaneously, both will produce the same value, 
   and everything will be ok if the two assignments to 'charset_aliases' 
   are atomic. But I don't know what will happen if the two assignments mix.  */ 
#if __STDC__ != 1 
# define volatile /* empty */ 
#endif 
/* Pointer to the contents of the charset.alias file, if it has already been 
   read, else NULL.  Its format is: 
   ALIAS_1 '\0' CANONICAL_1 '\0' ... ALIAS_n '\0' CANONICAL_n '\0' '\0'  */ 
static const char * volatile charset_aliases; 
 
/* Return a pointer to the contents of the charset.alias file.  */ 
static const char * 
get_charset_aliases (void) 
 {  {
  const char *cp;  const char alias[11+1];
   const char canonical[11+1];
 };
   
  cp = charset_aliases;/* Table of platform-dependent mappings, sorted in ascending order.  */
  if (cp == NULL)static const struct table_entry alias_table[] =
    {  {
#if !(defined DARWIN7 || defined VMS || defined WIN32_NATIVE || defined __CYGWIN__)#  if defined __FreeBSD__                                   /* FreeBSD */
      const char *dir;  /*{ "ARMSCII-8",  "ARMSCII-8" },*/
      const char *base = "charset.alias";    { "Big5",       "BIG5" },
      char *file_name;    { "Big5HKSCS",  "BIG5-HKSCS" },
     { "C",          "ASCII" },
   /*{ "CP1131",     "CP1131" },*/
   /*{ "CP1251",     "CP1251" },*/
   /*{ "CP866",      "CP866" },*/
   /*{ "GB18030",    "GB18030" },*/
   /*{ "GB2312",     "GB2312" },*/
   /*{ "GBK",        "GBK" },*/
   /*{ "ISCII-DEV",  "?" },*/
     { "ISO8859-1",  "ISO-8859-1" },
     { "ISO8859-13", "ISO-8859-13" },
     { "ISO8859-15", "ISO-8859-15" },
     { "ISO8859-2",  "ISO-8859-2" },
     { "ISO8859-4",  "ISO-8859-4" },
     { "ISO8859-5",  "ISO-8859-5" },
     { "ISO8859-7",  "ISO-8859-7" },
     { "ISO8859-9",  "ISO-8859-9" },
   /*{ "KOI8-R",     "KOI8-R" },*/
   /*{ "KOI8-U",     "KOI8-U" },*/
     { "SJIS",       "SHIFT_JIS" },
     { "US-ASCII",   "ASCII" },
     { "eucCN",      "GB2312" },
     { "eucJP",      "EUC-JP" },
     { "eucKR",      "EUC-KR" }
 #   define alias_table_defined
 #  endif
 #  if defined __NetBSD__                                    /* NetBSD */
     { "646",        "ASCII" },
   /*{ "ARMSCII-8",  "ARMSCII-8" },*/
   /*{ "BIG5",       "BIG5" },*/
     { "Big5-HKSCS", "BIG5-HKSCS" },
   /*{ "CP1251",     "CP1251" },*/
   /*{ "CP866",      "CP866" },*/
   /*{ "GB18030",    "GB18030" },*/
   /*{ "GB2312",     "GB2312" },*/
     { "ISO8859-1",  "ISO-8859-1" },
     { "ISO8859-13", "ISO-8859-13" },
     { "ISO8859-15", "ISO-8859-15" },
     { "ISO8859-2",  "ISO-8859-2" },
     { "ISO8859-4",  "ISO-8859-4" },
     { "ISO8859-5",  "ISO-8859-5" },
     { "ISO8859-7",  "ISO-8859-7" },
   /*{ "KOI8-R",     "KOI8-R" },*/
   /*{ "KOI8-U",     "KOI8-U" },*/
   /*{ "PT154",      "PT154" },*/
     { "SJIS",       "SHIFT_JIS" },
     { "eucCN",      "GB2312" },
     { "eucJP",      "EUC-JP" },
     { "eucKR",      "EUC-KR" },
     { "eucTW",      "EUC-TW" }
 #   define alias_table_defined
 #  endif
 #  if defined __OpenBSD__                                   /* OpenBSD */
     { "646",        "ASCII" },
     { "ISO8859-1",  "ISO-8859-1" },
     { "ISO8859-13", "ISO-8859-13" },
     { "ISO8859-15", "ISO-8859-15" },
     { "ISO8859-2",  "ISO-8859-2" },
     { "ISO8859-4",  "ISO-8859-4" },
     { "ISO8859-5",  "ISO-8859-5" },
     { "ISO8859-7",  "ISO-8859-7" }
 #   define alias_table_defined
 #  endif
 #  if defined __APPLE__ && defined __MACH__                 /* Mac OS X */
     /* Darwin 7.5 has nl_langinfo(CODESET), but sometimes its value is
        useless:
        - It returns the empty string when LANG is set to a locale of the
          form ll_CC, although ll_CC/LC_CTYPE is a symlink to an UTF-8
          LC_CTYPE file.
        - The environment variables LANG, LC_CTYPE, LC_ALL are not set by
          the system; nl_langinfo(CODESET) returns "US-ASCII" in this case.
        - The documentation says:
            "... all code that calls BSD system routines should ensure
             that the const *char parameters of these routines are in UTF-8
             encoding. All BSD system functions expect their string
             parameters to be in UTF-8 encoding and nothing else."
          It also says
            "An additional caveat is that string parameters for files,
             paths, and other file-system entities must be in canonical
             UTF-8. In a canonical UTF-8 Unicode string, all decomposable
             characters are decomposed ..."
          but this is not true: You can pass non-decomposed UTF-8 strings
          to file system functions, and it is the OS which will convert
          them to decomposed UTF-8 before accessing the file system.
        - The Apple Terminal application displays UTF-8 by default.
        - However, other applications are free to use different encodings:
          - xterm uses ISO-8859-1 by default.
          - TextEdit uses MacRoman by default.
        We prefer UTF-8 over decomposed UTF-8-MAC because one should
        minimize the use of decomposed Unicode. Unfortunately, through the
        Darwin file system, decomposed UTF-8 strings are leaked into user
        space nevertheless.
        Then there are also the locales with encodings other than US-ASCII
        and UTF-8. These locales can be occasionally useful to users (e.g.
        when grepping through ISO-8859-1 encoded text files), when all their
        file names are in US-ASCII.
      */
     { "ARMSCII-8",  "ARMSCII-8" },
     { "Big5",       "BIG5" },
     { "Big5HKSCS",  "BIG5-HKSCS" },
     { "CP1131",     "CP1131" },
     { "CP1251",     "CP1251" },
     { "CP866",      "CP866" },
     { "CP949",      "CP949" },
     { "GB18030",    "GB18030" },
     { "GB2312",     "GB2312" },
     { "GBK",        "GBK" },
   /*{ "ISCII-DEV",  "?" },*/
     { "ISO8859-1",  "ISO-8859-1" },
     { "ISO8859-13", "ISO-8859-13" },
     { "ISO8859-15", "ISO-8859-15" },
     { "ISO8859-2",  "ISO-8859-2" },
     { "ISO8859-4",  "ISO-8859-4" },
     { "ISO8859-5",  "ISO-8859-5" },
     { "ISO8859-7",  "ISO-8859-7" },
     { "ISO8859-9",  "ISO-8859-9" },
     { "KOI8-R",     "KOI8-R" },
     { "KOI8-U",     "KOI8-U" },
     { "PT154",      "PT154" },
     { "SJIS",       "SHIFT_JIS" },
     { "eucCN",      "GB2312" },
     { "eucJP",      "EUC-JP" },
     { "eucKR",      "EUC-KR" }
 #   define alias_table_defined
 #  endif
 #  if defined _AIX                                          /* AIX */
   /*{ "GBK",        "GBK" },*/
     { "IBM-1046",   "CP1046" },
     { "IBM-1124",   "CP1124" },
     { "IBM-1129",   "CP1129" },
     { "IBM-1252",   "CP1252" },
     { "IBM-850",    "CP850" },
     { "IBM-856",    "CP856" },
     { "IBM-921",    "ISO-8859-13" },
     { "IBM-922",    "CP922" },
     { "IBM-932",    "CP932" },
     { "IBM-943",    "CP943" },
     { "IBM-eucCN",  "GB2312" },
     { "IBM-eucJP",  "EUC-JP" },
     { "IBM-eucKR",  "EUC-KR" },
     { "IBM-eucTW",  "EUC-TW" },
     { "ISO8859-1",  "ISO-8859-1" },
     { "ISO8859-15", "ISO-8859-15" },
     { "ISO8859-2",  "ISO-8859-2" },
     { "ISO8859-5",  "ISO-8859-5" },
     { "ISO8859-6",  "ISO-8859-6" },
     { "ISO8859-7",  "ISO-8859-7" },
     { "ISO8859-8",  "ISO-8859-8" },
     { "ISO8859-9",  "ISO-8859-9" },
     { "TIS-620",    "TIS-620" },
   /*{ "UTF-8",      "UTF-8" },*/
     { "big5",       "BIG5" }
 #   define alias_table_defined
 #  endif
 #  if defined __hpux                                        /* HP-UX */
     { "SJIS",      "SHIFT_JIS" },
     { "arabic8",   "HP-ARABIC8" },
     { "big5",      "BIG5" },
     { "cp1251",    "CP1251" },
     { "eucJP",     "EUC-JP" },
     { "eucKR",     "EUC-KR" },
     { "eucTW",     "EUC-TW" },
     { "gb18030",   "GB18030" },
     { "greek8",    "HP-GREEK8" },
     { "hebrew8",   "HP-HEBREW8" },
     { "hkbig5",    "BIG5-HKSCS" },
     { "hp15CN",    "GB2312" },
     { "iso88591",  "ISO-8859-1" },
     { "iso885913", "ISO-8859-13" },
     { "iso885915", "ISO-8859-15" },
     { "iso88592",  "ISO-8859-2" },
     { "iso88594",  "ISO-8859-4" },
     { "iso88595",  "ISO-8859-5" },
     { "iso88596",  "ISO-8859-6" },
     { "iso88597",  "ISO-8859-7" },
     { "iso88598",  "ISO-8859-8" },
     { "iso88599",  "ISO-8859-9" },
     { "kana8",     "HP-KANA8" },
     { "koi8r",     "KOI8-R" },
     { "roman8",    "HP-ROMAN8" },
     { "tis620",    "TIS-620" },
     { "turkish8",  "HP-TURKISH8" },
     { "utf8",      "UTF-8" }
 #   define alias_table_defined
 #  endif
 #  if defined __sgi                                         /* IRIX */
     { "ISO8859-1",  "ISO-8859-1" },
     { "ISO8859-15", "ISO-8859-15" },
     { "ISO8859-2",  "ISO-8859-2" },
     { "ISO8859-5",  "ISO-8859-5" },
     { "ISO8859-7",  "ISO-8859-7" },
     { "ISO8859-9",  "ISO-8859-9" },
     { "eucCN",      "GB2312" },
     { "eucJP",      "EUC-JP" },
     { "eucKR",      "EUC-KR" },
     { "eucTW",      "EUC-TW" }
 #   define alias_table_defined
 #  endif
 #  if defined __osf__                                       /* OSF/1 */
   /*{ "GBK",        "GBK" },*/
     { "ISO8859-1",  "ISO-8859-1" },
     { "ISO8859-15", "ISO-8859-15" },
     { "ISO8859-2",  "ISO-8859-2" },
     { "ISO8859-4",  "ISO-8859-4" },
     { "ISO8859-5",  "ISO-8859-5" },
     { "ISO8859-7",  "ISO-8859-7" },
     { "ISO8859-8",  "ISO-8859-8" },
     { "ISO8859-9",  "ISO-8859-9" },
     { "KSC5601",    "CP949" },
     { "SJIS",       "SHIFT_JIS" },
     { "TACTIS",     "TIS-620" },
   /*{ "UTF-8",      "UTF-8" },*/
     { "big5",       "BIG5" },
     { "cp850",      "CP850" },
     { "dechanyu",   "DEC-HANYU" },
     { "dechanzi",   "GB2312" },
     { "deckanji",   "DEC-KANJI" },
     { "deckorean",  "EUC-KR" },
     { "eucJP",      "EUC-JP" },
     { "eucKR",      "EUC-KR" },
     { "eucTW",      "EUC-TW" },
     { "sdeckanji",  "EUC-JP" }
 #   define alias_table_defined
 #  endif
 #  if defined __sun                                         /* Solaris */
     { "5601",        "EUC-KR" },
     { "646",         "ASCII" },
   /*{ "BIG5",        "BIG5" },*/
     { "Big5-HKSCS",  "BIG5-HKSCS" },
     { "GB18030",     "GB18030" },
   /*{ "GBK",         "GBK" },*/
     { "ISO8859-1",   "ISO-8859-1" },
     { "ISO8859-11",  "TIS-620" },
     { "ISO8859-13",  "ISO-8859-13" },
     { "ISO8859-15",  "ISO-8859-15" },
     { "ISO8859-2",   "ISO-8859-2" },
     { "ISO8859-3",   "ISO-8859-3" },
     { "ISO8859-4",   "ISO-8859-4" },
     { "ISO8859-5",   "ISO-8859-5" },
     { "ISO8859-6",   "ISO-8859-6" },
     { "ISO8859-7",   "ISO-8859-7" },
     { "ISO8859-8",   "ISO-8859-8" },
     { "ISO8859-9",   "ISO-8859-9" },
     { "PCK",         "SHIFT_JIS" },
     { "TIS620.2533", "TIS-620" },
   /*{ "UTF-8",       "UTF-8" },*/
     { "ansi-1251",   "CP1251" },
     { "cns11643",    "EUC-TW" },
     { "eucJP",       "EUC-JP" },
     { "gb2312",      "GB2312" },
     { "koi8-r",      "KOI8-R" }
 #   define alias_table_defined
 #  endif
 #  if defined __minix                                       /* Minix */
     { "646", "ASCII" }
 #   define alias_table_defined
 #  endif
 #  if defined WINDOWS_NATIVE || defined __CYGWIN__          /* Windows */
     { "CP1361",  "JOHAB" },
     { "CP20127", "ASCII" },
     { "CP20866", "KOI8-R" },
     { "CP20936", "GB2312" },
     { "CP21866", "KOI8-RU" },
     { "CP28591", "ISO-8859-1" },
     { "CP28592", "ISO-8859-2" },
     { "CP28593", "ISO-8859-3" },
     { "CP28594", "ISO-8859-4" },
     { "CP28595", "ISO-8859-5" },
     { "CP28596", "ISO-8859-6" },
     { "CP28597", "ISO-8859-7" },
     { "CP28598", "ISO-8859-8" },
     { "CP28599", "ISO-8859-9" },
     { "CP28605", "ISO-8859-15" },
     { "CP38598", "ISO-8859-8" },
     { "CP51932", "EUC-JP" },
     { "CP51936", "GB2312" },
     { "CP51949", "EUC-KR" },
     { "CP51950", "EUC-TW" },
     { "CP54936", "GB18030" },
     { "CP65001", "UTF-8" },
     { "CP936",   "GBK" }
 #   define alias_table_defined
 #  endif
 #  if defined OS2                                           /* OS/2 */
     /* The list of encodings is taken from "List of OS/2 Codepages"
        by Alex Taylor:
        <http://altsan.org/os2/toolkits/uls/index.html#codepages>.
        See also "IBM Globalization - Code page identifiers":
        <https://www-01.ibm.com/software/globalization/cp/cp_cpgid.html>.  */
     { "CP1089", "ISO-8859-6" },
     { "CP1208", "UTF-8" },
     { "CP1381", "GB2312" },
     { "CP1386", "GBK" },
     { "CP3372", "EUC-JP" },
     { "CP813",  "ISO-8859-7" },
     { "CP819",  "ISO-8859-1" },
     { "CP878",  "KOI8-R" },
     { "CP912",  "ISO-8859-2" },
     { "CP913",  "ISO-8859-3" },
     { "CP914",  "ISO-8859-4" },
     { "CP915",  "ISO-8859-5" },
     { "CP916",  "ISO-8859-8" },
     { "CP920",  "ISO-8859-9" },
     { "CP921",  "ISO-8859-13" },
     { "CP923",  "ISO-8859-15" },
     { "CP954",  "EUC-JP" },
     { "CP964",  "EUC-TW" },
     { "CP970",  "EUC-KR" }
 #   define alias_table_defined
 #  endif
 #  if defined VMS                                           /* OpenVMS */
     /* The list of encodings is taken from the OpenVMS 7.3-1 documentation
        "Compaq C Run-Time Library Reference Manual for OpenVMS systems"
        section 10.7 "Handling Different Character Sets".  */
     { "DECHANYU",  "DEC-HANYU" },
     { "DECHANZI",  "GB2312" },
     { "DECKANJI",  "DEC-KANJI" },
     { "DECKOREAN", "EUC-KR" },
     { "ISO8859-1", "ISO-8859-1" },
     { "ISO8859-2", "ISO-8859-2" },
     { "ISO8859-5", "ISO-8859-5" },
     { "ISO8859-7", "ISO-8859-7" },
     { "ISO8859-8", "ISO-8859-8" },
     { "ISO8859-9", "ISO-8859-9" },
     { "SDECKANJI", "EUC-JP" },
     { "SJIS",      "SHIFT_JIS" },
     { "eucJP",     "EUC-JP" },
     { "eucTW",     "EUC-TW" }
 #   define alias_table_defined
 #  endif
 #  ifndef alias_table_defined
     /* Just a dummy entry, to avoid a C syntax error.  */
     { "", "" }
 #  endif
   };
   
      /* Make it possible to override the charset.alias location.  This is# endif
         necessary for running the testsuite before "make install".  */ 
      dir = getenv ("CHARSETALIASDIR"); 
      if (dir == NULL || dir[0] == '\0') 
        dir = relocate (LIBDIR); 
   
      /* Concatenate dir and base into freshly allocated file_name.  */#else
      { 
        size_t dir_len = strlen (dir); 
        size_t base_len = strlen (base); 
        int add_slash = (dir_len > 0 && !ISSLASH (dir[dir_len - 1])); 
        file_name = (char *) malloc (dir_len + add_slash + base_len + 1); 
        if (file_name != NULL) 
          { 
            memcpy (file_name, dir, dir_len); 
            if (add_slash) 
              file_name[dir_len] = DIRECTORY_SEPARATOR; 
            memcpy (file_name + dir_len + add_slash, base, base_len + 1); 
          } 
      } 
   
      if (file_name == NULL)/* On these platforms, we use a mapping from locale name to GNU canonical
        /* Out of memory.  Treat the file as empty.  */   encoding name.  */
        cp = ""; 
      else 
        { 
          int fd; 
   
          /* Open the file.  Reject symbolic links on platforms that supportstruct table_entry
             O_NOFOLLOW.  This is a security feature.  Without it, an attacker{
             could retrieve parts of the contents (namely, the tail of the  const char locale[17+1];
             first line that starts with "* ") of an arbitrary file by placing  const char canonical[11+1];
             a symbolic link to that file under the name "charset.alias" in};
             some writable directory and defining the environment variable 
             CHARSETALIASDIR to point to that directory.  */ 
          fd = open (file_name, 
                     O_RDONLY | (HAVE_WORKING_O_NOFOLLOW ? O_NOFOLLOW : 0)); 
          if (fd < 0) 
            /* File not found.  Treat it as empty.  */ 
            cp = ""; 
          else 
            { 
              FILE *fp; 
   
              fp = fdopen (fd, "r");/* Table of platform-dependent mappings, sorted in ascending order.  */
              if (fp == NULL)static const struct table_entry locale_table[] =
                {  {
                  /* Out of memory.  Treat the file as empty.  */# if defined __FreeBSD__                                    /* FreeBSD 4.2 */
                  close (fd);    { "cs_CZ.ISO_8859-2",  "ISO-8859-2" },
                  cp = "";    { "da_DK.DIS_8859-15", "ISO-8859-15" },
                }    { "da_DK.ISO_8859-1",  "ISO-8859-1" },
              else    { "de_AT.DIS_8859-15", "ISO-8859-15" },
                {    { "de_AT.ISO_8859-1",  "ISO-8859-1" },
                  /* Parse the file's contents.  */    { "de_CH.DIS_8859-15", "ISO-8859-15" },
                  char *res_ptr = NULL;    { "de_CH.ISO_8859-1",  "ISO-8859-1" },
                  size_t res_size = 0;    { "de_DE.DIS_8859-15", "ISO-8859-15" },
    { "de_DE.ISO_8859-1",  "ISO-8859-1" },
                  for (;;)    { "en_AU.DIS_8859-15", "ISO-8859-15" },
                    {    { "en_AU.ISO_8859-1",  "ISO-8859-1" },
                      int c;    { "en_CA.DIS_8859-15", "ISO-8859-15" },
                      char buf1[50+1];    { "en_CA.ISO_8859-1",  "ISO-8859-1" },
                      char buf2[50+1];    { "en_GB.DIS_8859-15", "ISO-8859-15" },
                      size_t l1, l2;    { "en_GB.ISO_8859-1",  "ISO-8859-1" },
                      char *old_res_ptr;    { "en_US.DIS_8859-15", "ISO-8859-15" },
    { "en_US.ISO_8859-1",  "ISO-8859-1" },
                      c = getc (fp);    { "es_ES.DIS_8859-15", "ISO-8859-15" },
                      if (c == EOF)    { "es_ES.ISO_8859-1",  "ISO-8859-1" },
                        break;    { "fi_FI.DIS_8859-15", "ISO-8859-15" },
                      if (c == '\n' || c == ' ' || c == '\t')    { "fi_FI.ISO_8859-1",  "ISO-8859-1" },
                        continue;    { "fr_BE.DIS_8859-15", "ISO-8859-15" },
                      if (c == '#')    { "fr_BE.ISO_8859-1",  "ISO-8859-1" },
                        {    { "fr_CA.DIS_8859-15", "ISO-8859-15" },
                          /* Skip comment, to end of line.  */    { "fr_CA.ISO_8859-1",  "ISO-8859-1" },
                          do    { "fr_CH.DIS_8859-15", "ISO-8859-15" },
                            c = getc (fp);    { "fr_CH.ISO_8859-1",  "ISO-8859-1" },
                          while (!(c == EOF || c == '\n'));    { "fr_FR.DIS_8859-15", "ISO-8859-15" },
                          if (c == EOF)    { "fr_FR.ISO_8859-1",  "ISO-8859-1" },
                            break;    { "hr_HR.ISO_8859-2",  "ISO-8859-2" },
                          continue;    { "hu_HU.ISO_8859-2",  "ISO-8859-2" },
                        }    { "is_IS.DIS_8859-15", "ISO-8859-15" },
                      ungetc (c, fp);    { "is_IS.ISO_8859-1",  "ISO-8859-1" },
                      if (fscanf (fp, "%50s %50s", buf1, buf2) < 2)    { "it_CH.DIS_8859-15", "ISO-8859-15" },
                        break;    { "it_CH.ISO_8859-1",  "ISO-8859-1" },
                      l1 = strlen (buf1);    { "it_IT.DIS_8859-15", "ISO-8859-15" },
                      l2 = strlen (buf2);    { "it_IT.ISO_8859-1",  "ISO-8859-1" },
                      old_res_ptr = res_ptr;    { "ja_JP.EUC",         "EUC-JP" },
                      if (res_size == 0)    { "ja_JP.SJIS",        "SHIFT_JIS" },
                        {    { "ja_JP.Shift_JIS",   "SHIFT_JIS" },
                          res_size = l1 + 1 + l2 + 1;    { "ko_KR.EUC",         "EUC-KR" },
                          res_ptr = (char *) malloc (res_size + 1);    { "la_LN.ASCII",       "ASCII" },
                        }    { "la_LN.DIS_8859-15", "ISO-8859-15" },
                      else    { "la_LN.ISO_8859-1",  "ISO-8859-1" },
                        {    { "la_LN.ISO_8859-2",  "ISO-8859-2" },
                          res_size += l1 + 1 + l2 + 1;    { "la_LN.ISO_8859-4",  "ISO-8859-4" },
                          res_ptr = (char *) realloc (res_ptr, res_size + 1);    { "lt_LN.ASCII",       "ASCII" },
                        }    { "lt_LN.DIS_8859-15", "ISO-8859-15" },
                      if (res_ptr == NULL)    { "lt_LN.ISO_8859-1",  "ISO-8859-1" },
                        {    { "lt_LN.ISO_8859-2",  "ISO-8859-2" },
                          /* Out of memory. */    { "lt_LT.ISO_8859-4",  "ISO-8859-4" },
                          res_size = 0;    { "nl_BE.DIS_8859-15", "ISO-8859-15" },
                          free (old_res_ptr);    { "nl_BE.ISO_8859-1",  "ISO-8859-1" },
                          break;    { "nl_NL.DIS_8859-15", "ISO-8859-15" },
                        }    { "nl_NL.ISO_8859-1",  "ISO-8859-1" },
                      strcpy (res_ptr + res_size - (l2 + 1) - (l1 + 1), buf1);    { "no_NO.DIS_8859-15", "ISO-8859-15" },
                      strcpy (res_ptr + res_size - (l2 + 1), buf2);    { "no_NO.ISO_8859-1",  "ISO-8859-1" },
                    }    { "pl_PL.ISO_8859-2",  "ISO-8859-2" },
                  fclose (fp);    { "pt_PT.DIS_8859-15", "ISO-8859-15" },
                  if (res_size == 0)    { "pt_PT.ISO_8859-1",  "ISO-8859-1" },
                    cp = "";    { "ru_RU.CP866",       "CP866" },
                  else    { "ru_RU.ISO_8859-5",  "ISO-8859-5" },
                    {    { "ru_RU.KOI8-R",      "KOI8-R" },
                      *(res_ptr + res_size) = '\0';    { "ru_SU.CP866",       "CP866" },
                      cp = res_ptr;    { "ru_SU.ISO_8859-5",  "ISO-8859-5" },
                    }    { "ru_SU.KOI8-R",      "KOI8-R" },
                }    { "sl_SI.ISO_8859-2",  "ISO-8859-2" },
            }    { "sv_SE.DIS_8859-15", "ISO-8859-15" },
    { "sv_SE.ISO_8859-1",  "ISO-8859-1" },
          free (file_name);    { "uk_UA.KOI8-U",      "KOI8-U" },
        }    { "zh_CN.EUC",         "GB2312" },
    { "zh_TW.BIG5",        "BIG5" },
#else    { "zh_TW.Big5",        "BIG5" }
#  define locale_table_defined
# if defined DARWIN7 
      /* To avoid the trouble of installing a file that is shared by many 
         GNU packages -- many packaging systems have problems with this --, 
         simply inline the aliases here.  */ 
      cp = "ISO8859-1" "\0" "ISO-8859-1" "\0" 
           "ISO8859-2" "\0" "ISO-8859-2" "\0" 
           "ISO8859-4" "\0" "ISO-8859-4" "\0" 
           "ISO8859-5" "\0" "ISO-8859-5" "\0" 
           "ISO8859-7" "\0" "ISO-8859-7" "\0" 
           "ISO8859-9" "\0" "ISO-8859-9" "\0" 
           "ISO8859-13" "\0" "ISO-8859-13" "\0" 
           "ISO8859-15" "\0" "ISO-8859-15" "\0" 
           "KOI8-R" "\0" "KOI8-R" "\0" 
           "KOI8-U" "\0" "KOI8-U" "\0" 
           "CP866" "\0" "CP866" "\0" 
           "CP949" "\0" "CP949" "\0" 
           "CP1131" "\0" "CP1131" "\0" 
           "CP1251" "\0" "CP1251" "\0" 
           "eucCN" "\0" "GB2312" "\0" 
           "GB2312" "\0" "GB2312" "\0" 
           "eucJP" "\0" "EUC-JP" "\0" 
           "eucKR" "\0" "EUC-KR" "\0" 
           "Big5" "\0" "BIG5" "\0" 
           "Big5HKSCS" "\0" "BIG5-HKSCS" "\0" 
           "GBK" "\0" "GBK" "\0" 
           "GB18030" "\0" "GB18030" "\0" 
           "SJIS" "\0" "SHIFT_JIS" "\0" 
           "ARMSCII-8" "\0" "ARMSCII-8" "\0" 
           "PT154" "\0" "PT154" "\0" 
         /*"ISCII-DEV" "\0" "?" "\0"*/ 
           "*" "\0" "UTF-8" "\0"; 
 # endif  # endif
# if defined __DJGPP__                                      /* DOS / DJGPP 2.03 */
# if defined VMS    /* The encodings given here may not all be correct.
      /* To avoid the troubles of an extra file charset.alias_vms in the       If you find that the encoding given for your language and
         sources of many GNU packages, simply inline the aliases here.  */       country is not the one your DOS machine actually uses, just
      /* The list of encodings is taken from the OpenVMS 7.3-1 documentation       correct it in this file, and send a mail to
         "Compaq C Run-Time Library Reference Manual for OpenVMS systems"       Juan Manuel Guerrero <juan.guerrero@gmx.de>
         section 10.7 "Handling Different Character Sets".  */       and <bug-gnulib@gnu.org>.  */
      cp = "ISO8859-1" "\0" "ISO-8859-1" "\0"    { "C",     "ASCII" },
           "ISO8859-2" "\0" "ISO-8859-2" "\0"    { "ar",    "CP864" },
           "ISO8859-5" "\0" "ISO-8859-5" "\0"    { "ar_AE", "CP864" },
           "ISO8859-7" "\0" "ISO-8859-7" "\0"    { "ar_DZ", "CP864" },
           "ISO8859-8" "\0" "ISO-8859-8" "\0"    { "ar_EG", "CP864" },
           "ISO8859-9" "\0" "ISO-8859-9" "\0"    { "ar_IQ", "CP864" },
           /* Japanese */    { "ar_IR", "CP864" },
           "eucJP" "\0" "EUC-JP" "\0"    { "ar_JO", "CP864" },
           "SJIS" "\0" "SHIFT_JIS" "\0"    { "ar_KW", "CP864" },
           "DECKANJI" "\0" "DEC-KANJI" "\0"    { "ar_MA", "CP864" },
           "SDECKANJI" "\0" "EUC-JP" "\0"    { "ar_OM", "CP864" },
           /* Chinese */    { "ar_QA", "CP864" },
           "eucTW" "\0" "EUC-TW" "\0"    { "ar_SA", "CP864" },
           "DECHANYU" "\0" "DEC-HANYU" "\0"    { "ar_SY", "CP864" },
           "DECHANZI" "\0" "GB2312" "\0"    { "be",    "CP866" },
           /* Korean */    { "be_BE", "CP866" },
           "DECKOREAN" "\0" "EUC-KR" "\0";    { "bg",    "CP866" }, /* not CP855 ?? */
     { "bg_BG", "CP866" }, /* not CP855 ?? */
     { "ca",    "CP850" },
     { "ca_ES", "CP850" },
     { "cs",    "CP852" },
     { "cs_CZ", "CP852" },
     { "da",    "CP865" }, /* not CP850 ?? */
     { "da_DK", "CP865" }, /* not CP850 ?? */
     { "de",    "CP850" },
     { "de_AT", "CP850" },
     { "de_CH", "CP850" },
     { "de_DE", "CP850" },
     { "el",    "CP869" },
     { "el_GR", "CP869" },
     { "en",    "CP850" },
     { "en_AU", "CP850" }, /* not CP437 ?? */
     { "en_CA", "CP850" },
     { "en_GB", "CP850" },
     { "en_NZ", "CP437" },
     { "en_US", "CP437" },
     { "en_ZA", "CP850" }, /* not CP437 ?? */
     { "eo",    "CP850" },
     { "eo_EO", "CP850" },
     { "es",    "CP850" },
     { "es_AR", "CP850" },
     { "es_BO", "CP850" },
     { "es_CL", "CP850" },
     { "es_CO", "CP850" },
     { "es_CR", "CP850" },
     { "es_CU", "CP850" },
     { "es_DO", "CP850" },
     { "es_EC", "CP850" },
     { "es_ES", "CP850" },
     { "es_GT", "CP850" },
     { "es_HN", "CP850" },
     { "es_MX", "CP850" },
     { "es_NI", "CP850" },
     { "es_PA", "CP850" },
     { "es_PE", "CP850" },
     { "es_PY", "CP850" },
     { "es_SV", "CP850" },
     { "es_UY", "CP850" },
     { "es_VE", "CP850" },
     { "et",    "CP850" },
     { "et_EE", "CP850" },
     { "eu",    "CP850" },
     { "eu_ES", "CP850" },
     { "fi",    "CP850" },
     { "fi_FI", "CP850" },
     { "fr",    "CP850" },
     { "fr_BE", "CP850" },
     { "fr_CA", "CP850" },
     { "fr_CH", "CP850" },
     { "fr_FR", "CP850" },
     { "ga",    "CP850" },
     { "ga_IE", "CP850" },
     { "gd",    "CP850" },
     { "gd_GB", "CP850" },
     { "gl",    "CP850" },
     { "gl_ES", "CP850" },
     { "he",    "CP862" },
     { "he_IL", "CP862" },
     { "hr",    "CP852" },
     { "hr_HR", "CP852" },
     { "hu",    "CP852" },
     { "hu_HU", "CP852" },
     { "id",    "CP850" }, /* not CP437 ?? */
     { "id_ID", "CP850" }, /* not CP437 ?? */
     { "is",    "CP861" }, /* not CP850 ?? */
     { "is_IS", "CP861" }, /* not CP850 ?? */
     { "it",    "CP850" },
     { "it_CH", "CP850" },
     { "it_IT", "CP850" },
     { "ja",    "CP932" },
     { "ja_JP", "CP932" },
     { "kr",    "CP949" }, /* not CP934 ?? */
     { "kr_KR", "CP949" }, /* not CP934 ?? */
     { "lt",    "CP775" },
     { "lt_LT", "CP775" },
     { "lv",    "CP775" },
     { "lv_LV", "CP775" },
     { "mk",    "CP866" }, /* not CP855 ?? */
     { "mk_MK", "CP866" }, /* not CP855 ?? */
     { "mt",    "CP850" },
     { "mt_MT", "CP850" },
     { "nb",    "CP865" }, /* not CP850 ?? */
     { "nb_NO", "CP865" }, /* not CP850 ?? */
     { "nl",    "CP850" },
     { "nl_BE", "CP850" },
     { "nl_NL", "CP850" },
     { "nn",    "CP865" }, /* not CP850 ?? */
     { "nn_NO", "CP865" }, /* not CP850 ?? */
     { "no",    "CP865" }, /* not CP850 ?? */
     { "no_NO", "CP865" }, /* not CP850 ?? */
     { "pl",    "CP852" },
     { "pl_PL", "CP852" },
     { "pt",    "CP850" },
     { "pt_BR", "CP850" },
     { "pt_PT", "CP850" },
     { "ro",    "CP852" },
     { "ro_RO", "CP852" },
     { "ru",    "CP866" },
     { "ru_RU", "CP866" },
     { "sk",    "CP852" },
     { "sk_SK", "CP852" },
     { "sl",    "CP852" },
     { "sl_SI", "CP852" },
     { "sq",    "CP852" },
     { "sq_AL", "CP852" },
     { "sr",    "CP852" }, /* CP852 or CP866 or CP855 ?? */
     { "sr_CS", "CP852" }, /* CP852 or CP866 or CP855 ?? */
     { "sr_YU", "CP852" }, /* CP852 or CP866 or CP855 ?? */
     { "sv",    "CP850" },
     { "sv_SE", "CP850" },
     { "th",    "CP874" },
     { "th_TH", "CP874" },
     { "tr",    "CP857" },
     { "tr_TR", "CP857" },
     { "uk",    "CP1125" },
     { "uk_UA", "CP1125" },
     { "zh_CN", "GBK" },
     { "zh_TW", "CP950" } /* not CP938 ?? */
 #  define locale_table_defined
 # endif  # endif
# ifndef locale_table_defined
# if defined WIN32_NATIVE || defined __CYGWIN__    /* Just a dummy entry, to avoid a C syntax error.  */
      /* To avoid the troubles of installing a separate file in the same    { "", "" }
         directory as the DLL and of retrieving the DLL's directory at 
         runtime, simply inline the aliases here.  */ 
 
      cp = "CP936" "\0" "GBK" "\0" 
           "CP1361" "\0" "JOHAB" "\0" 
           "CP20127" "\0" "ASCII" "\0" 
           "CP20866" "\0" "KOI8-R" "\0" 
           "CP20936" "\0" "GB2312" "\0" 
           "CP21866" "\0" "KOI8-RU" "\0" 
           "CP28591" "\0" "ISO-8859-1" "\0" 
           "CP28592" "\0" "ISO-8859-2" "\0" 
           "CP28593" "\0" "ISO-8859-3" "\0" 
           "CP28594" "\0" "ISO-8859-4" "\0" 
           "CP28595" "\0" "ISO-8859-5" "\0" 
           "CP28596" "\0" "ISO-8859-6" "\0" 
           "CP28597" "\0" "ISO-8859-7" "\0" 
           "CP28598" "\0" "ISO-8859-8" "\0" 
           "CP28599" "\0" "ISO-8859-9" "\0" 
           "CP28605" "\0" "ISO-8859-15" "\0" 
           "CP38598" "\0" "ISO-8859-8" "\0" 
           "CP51932" "\0" "EUC-JP" "\0" 
           "CP51936" "\0" "GB2312" "\0" 
           "CP51949" "\0" "EUC-KR" "\0" 
           "CP51950" "\0" "EUC-TW" "\0" 
           "CP54936" "\0" "GB18030" "\0" 
           "CP65001" "\0" "UTF-8" "\0"; 
 # endif  # endif
     };
   
 #endif  #endif
   
       charset_aliases = cp;  
     }  
   
   return cp;  
 }  
   
 /* Determine the current locale's character encoding, and canonicalize it  /* Determine the current locale's character encoding, and canonicalize it
   into one of the canonical names listed in config.charset.   into one of the canonical names listed in localcharset.h.
    The result must not be freed; it is statically allocated.     The result must not be freed; it is statically allocated.
    If the canonical name cannot be determined, the result is a non-canonical     If the canonical name cannot be determined, the result is a non-canonical
    name.  */     name.  */
Line 359  const char * Line 689  const char *
 locale_charset (void)  locale_charset (void)
 {  {
   const char *codeset;    const char *codeset;
   const char *aliases;  
   
#if !(defined WIN32_NATIVE || defined OS2)#if HAVE_LANGINFO_CODESET || defined WINDOWS_NATIVE || defined OS2
   
 # if HAVE_LANGINFO_CODESET  # if HAVE_LANGINFO_CODESET
   
Line 408  locale_charset (void) Line 737  locale_charset (void)
             }              }
         }          }
   
      /* Woe32 has a function returning the locale's codepage as a number:      /* The Windows API has a function returning the locale's codepage as a
         GetACP().  This encoding is used by Cygwin, unless the user has set         number: GetACP().  This encoding is used by Cygwin, unless the user
         the environment variable CYGWIN=codepage:oem (which very few people         has set the environment variable CYGWIN=codepage:oem (which very few
         do).         people do).
          Output directed to console windows needs to be converted (to           Output directed to console windows needs to be converted (to
          GetOEMCP() if the console is using a raster font, or to           GetOEMCP() if the console is using a raster font, or to
          GetConsoleOutputCP() if it is using a TrueType font).  Cygwin does           GetConsoleOutputCP() if it is using a TrueType font).  Cygwin does
Line 424  locale_charset (void) Line 753  locale_charset (void)
     }      }
 #  endif  #  endif
   
# else  if (codeset == NULL)
     /* The canonical name cannot be determined.  */
     codeset = "";
   
  /* On old systems which lack it, use setlocale or getenv.  */# elif defined WINDOWS_NATIVE
  const char *locale = NULL; 
   
  /* But most old systems don't have a complete set of locales.  Some  static char buf[2 + 10 + 1];
     (like SunOS 4 or DJGPP) have only the C locale.  Therefore we don't 
     use setlocale here; it would return "C" when it doesn't support the 
     locale name the user has set.  */ 
#  if 0 
  locale = setlocale (LC_CTYPE, NULL); 
#  endif 
  if (locale == NULL || locale[0] == '\0') 
    { 
      locale = getenv ("LC_ALL"); 
      if (locale == NULL || locale[0] == '\0') 
        { 
          locale = getenv ("LC_CTYPE"); 
          if (locale == NULL || locale[0] == '\0') 
            locale = getenv ("LANG"); 
        } 
    } 
   
  /* On some old systems, one used to set locale = "iso8859_1". On others,  /* The Windows API has a function returning the locale's codepage as
     you set it to "language_COUNTRY.charset". In any case, we resolve it     a number, but the value doesn't change according to what the
     through the charset.alias file.  */     'setlocale' call specified.  So we use it as a last resort, in
  codeset = locale;     case the string returned by 'setlocale' doesn't specify the
      codepage.  */
   char *current_locale = setlocale (LC_ALL, NULL);
   char *pdot;
   
# endif  /* If they set different locales for different categories,
      'setlocale' will return a semi-colon separated list of locale
      values.  To make sure we use the correct one, we choose LC_CTYPE.  */
   if (strchr (current_locale, ';'))
     current_locale = setlocale (LC_CTYPE, NULL);
   
#elif defined WIN32_NATIVE  pdot = strrchr (current_locale, '.');
  if (pdot && 2 + strlen (pdot + 1) + 1 <= sizeof (buf))
  static char buf[2 + 10 + 1];    sprintf (buf, "CP%s", pdot + 1);
  else
  /* Woe32 has a function returning the locale's codepage as a number:    {
     GetACP().      /* The Windows API has a function returning the locale's codepage as a
     When the output goes to a console window, it needs to be provided in        number: GetACP().
     GetOEMCP() encoding if the console is using a raster font, or in        When the output goes to a console window, it needs to be provided in
     GetConsoleOutputCP() encoding if it is using a TrueType font.        GetOEMCP() encoding if the console is using a raster font, or in
     But in GUI programs and for output sent to files and pipes, GetACP()        GetConsoleOutputCP() encoding if it is using a TrueType font.
     encoding is the best bet.  */        But in GUI programs and for output sent to files and pipes, GetACP()
  sprintf (buf, "CP%u", GetACP ());        encoding is the best bet.  */
       sprintf (buf, "CP%u", GetACP ());
     }
   codeset = buf;    codeset = buf;
   
#elif defined OS2# elif defined OS2
   
   const char *locale;    const char *locale;
   static char buf[2 + 10 + 1];    static char buf[2 + 10 + 1];
   ULONG cp[3];    ULONG cp[3];
   ULONG cplen;    ULONG cplen;
   
     codeset = NULL;
   
   /* Allow user to override the codeset, as set in the operating system,    /* Allow user to override the codeset, as set in the operating system,
      with standard language environment variables.  */       with standard language environment variables.  */
   locale = getenv ("LC_ALL");    locale = getenv ("LC_ALL");
Line 506  locale_charset (void) Line 831  locale_charset (void)
             }              }
         }          }
   
      /* Resolve through the charset.alias file.  */      /* For the POSIX locale, don't use the system's codepage.  */
      codeset = locale;      if (strcmp (locale, "C") == 0 || strcmp (locale, "POSIX") == 0)
         codeset = "";
     }      }
  else
   if (codeset == NULL)
     {      {
       /* OS/2 has a function returning the locale's codepage as a number.  */        /* OS/2 has a function returning the locale's codepage as a number.  */
       if (DosQueryCp (sizeof (cp), cp, &cplen))        if (DosQueryCp (sizeof (cp), cp, &cplen))
Line 521  locale_charset (void) Line 848  locale_charset (void)
         }          }
     }      }
   
#endif# else
   
  if (codeset == NULL)#  error "Add code for other platforms here."
    /* The canonical name cannot be determined.  */ 
    codeset = ""; 
   
  /* Resolve alias. */# endif
  for (aliases = get_charset_aliases ();
       *aliases != '\0';  /* Resolve alias.  */
       aliases += strlen (aliases) + 1, aliases += strlen (aliases) + 1)  {
    if (strcmp (codeset, aliases) == 0# ifdef alias_table_defined
        || (aliases[0] == '*' && aliases[1] == '\0'))    /* On some platforms, UTF-8 locales are the most frequently used ones.
        Speed up the common case and slow down the less common cases by
        testing for this case first.  */
 #  if defined __OpenBSD__ || (defined __APPLE__ && defined __MACH__) || defined __sun || defined __CYGWIN__
     if (strcmp (codeset, "UTF-8") == 0)
       goto done_table_lookup;
     else
 #  endif
       {        {
        codeset = aliases + strlen (aliases) + 1;        const struct table_entry * const table = alias_table;
        break;        size_t const table_size =
           sizeof (alias_table) / sizeof (struct table_entry);
         /* The table is sorted.  Perform a binary search.  */
         size_t hi = table_size;
         size_t lo = 0;
         while (lo < hi)
           {
             /* Invariant:
                for i < lo, strcmp (table[i].alias, codeset) < 0,
                for i >= hi, strcmp (table[i].alias, codeset) > 0.  */
             size_t mid = (hi + lo) >> 1; /* >= lo, < hi */
             int cmp = strcmp (table[mid].alias, codeset);
             if (cmp < 0)
               lo = mid + 1;
             else if (cmp > 0)
               hi = mid;
             else
               {
                 /* Found an i with
                      strcmp (table[i].alias, codeset) == 0.  */
                 codeset = table[mid].canonical;
                 goto done_table_lookup;
               }
           }
       }        }
       if (0)
         done_table_lookup: ;
       else
   # endif
         {
           /* Did not find it in the table.  */
           /* On Mac OS X, all modern locales use the UTF-8 encoding.
              BeOS and Haiku have a single locale, and it has UTF-8 encoding.  */
   # if (defined __APPLE__ && defined __MACH__) || defined __BEOS__ || defined __HAIKU__
           codeset = "UTF-8";
   # else
           /* Don't return an empty string.  GNU libc and GNU libiconv interpret
              the empty string as denoting "the locale's character encoding",
              thus GNU libiconv would call this function a second time.  */
           if (codeset[0] == '\0')
             codeset = "ASCII";
   # endif
         }
     }
   
  /* Don't return an empty string.  GNU libc and GNU libiconv interpret#else
     the empty string as denoting "the locale's character encoding",
     thus GNU libiconv would call this function a second time.  */  /* On old systems which lack it, use setlocale or getenv.  */
  if (codeset[0] == '\0')  const char *locale = NULL;
 
   /* But most old systems don't have a complete set of locales.  Some
      (like DJGPP) have only the C locale.  Therefore we don't use setlocale
      here; it would return "C" when it doesn't support the locale name the
      user has set.  */
 # if 0
   locale = setlocale (LC_CTYPE, NULL);
 # endif
   if (locale == NULL || locale[0] == '\0')
     {
       locale = getenv ("LC_ALL");
       if (locale == NULL || locale[0] == '\0')
         {
           locale = getenv ("LC_CTYPE");
           if (locale == NULL || locale[0] == '\0')
             locale = getenv ("LANG");
             if (locale == NULL)
               locale = "";
         }
     }
 
   /* Map locale name to canonical encoding name.  */
   {
 # ifdef locale_table_defined
     const struct table_entry * const table = locale_table;
     size_t const table_size =
       sizeof (locale_table) / sizeof (struct table_entry);
     /* The table is sorted.  Perform a binary search.  */
     size_t hi = table_size;
     size_t lo = 0;
     while (lo < hi)
       {
         /* Invariant:
            for i < lo, strcmp (table[i].locale, locale) < 0,
            for i >= hi, strcmp (table[i].locale, locale) > 0.  */
         size_t mid = (hi + lo) >> 1; /* >= lo, < hi */
         int cmp = strcmp (table[mid].locale, locale);
         if (cmp < 0)
           lo = mid + 1;
         else if (cmp > 0)
           hi = mid;
         else
           {
             /* Found an i with
                  strcmp (table[i].locale, locale) == 0.  */
             codeset = table[mid].canonical;
             goto done_table_lookup;
           }
       }
     if (0)
       done_table_lookup: ;
     else
 # endif
       {
         /* Did not find it in the table.  */
         /* On Mac OS X, all modern locales use the UTF-8 encoding.
            BeOS and Haiku have a single locale, and it has UTF-8 encoding.  */
 # if (defined __APPLE__ && defined __MACH__) || defined __BEOS__ || defined __HAIKU__
         codeset = "UTF-8";
 # else
         /* The canonical name cannot be determined.  */
         /* Don't return an empty string.  GNU libc and GNU libiconv interpret
            the empty string as denoting "the locale's character encoding",
            thus GNU libiconv would call this function a second time.  */
         codeset = "ASCII";
 # endif
       }
   }
 
 #endif
 
 #ifdef DARWIN7
   /* Mac OS X sets MB_CUR_MAX to 1 when LC_ALL=C, and "UTF-8"
      (the default codeset) does not work when MB_CUR_MAX is 1.  */
   if (strcmp (codeset, "UTF-8") == 0 && MB_CUR_MAX_L (uselocale (NULL)) <= 1)
     codeset = "ASCII";      codeset = "ASCII";
   #endif
   
   return codeset;    return codeset;
 }  }
Removed from v.1.1.1.2  
changed lines
  Added in v.1.1.1.3

FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>