localcharset: Move mapping tables into the code.

* lib/localcharset.h: Document the GNU canonical names for character
encodings here.
* lib/localcharset.c: Don't include <fcntl.h>, <unistd.h>,
relocatable.h, configmake.h.
(O_NOFOLLOW, ISSLASH, DIRECTORY_SEPARATOR, getc, volatile): Remove
macros.
(charset_aliases): Remove variable.
(get_charset_aliases): Remove function.
(struct table_entry): New type.
(alias_table, locale_table): New constants.
(locale_charset): Use the alias_table or locale_table to get the
canonicalized encoding name.
* lib/config.charset: Remove file.
* lib/ref-add.sin: Remove file.
* lib/ref-del.sin: Remove file.
* m4/localcharset.m4 (gl_LOCALCHARSET): Don't require gl_FCNTL_O_FLAGS,
AC_CANONICAL_HOST, gl_GLIBC21. Don't check for getc_unlocked.
* modules/localcharset (Notice): Remove.
(Files): Remove config.charset, ref-add.sin, ref-del.sin, fcntl-o.m4,
glibc21.m4.
(Depends-on): Remove configmake.
(configure.ac): Define LOCALCHARSET_TESTS_ENVIRONMENT to empty.
(Makefile.am): Simplify.
* build-aux/prefix-gnulib-mk: Remove special code for the removed files.

1 files changed, 95 insertions, 1 deletions

diff --git a/lib/localcharset.h b/lib/localcharset.h
index 917e473e8b..56d5aaa176 100644
--- a/lib/localcharset.h
+++ b/lib/localcharset.h
@@ -25,12 +25,106 @@ extern "C" {
 
 
 /* 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 below.
    The result must not be freed; it is statically allocated.
    If the canonical name cannot be determined, the result is a non-canonical
    name.  */
 extern const char * locale_charset (void);
 
+/* About GNU canonical names for character encodings:
+
+   Every canonical name must be supported by GNU libiconv.  Support by GNU libc
+   is also desirable.
+
+   The name is case insensitive.  Usually an upper case MIME charset name is
+   preferred.
+
+   The current list of these GNU canonical names is:
+
+       name              MIME?             used by which systems
+                                    (darwin = Mac OS X, windows = native Windows)
+
+   ASCII, ANSI_X3.4-1968       glibc solaris freebsd netbsd darwin cygwin
+   ISO-8859-1              Y   glibc aix hpux irix osf solaris freebsd netbsd openbsd darwin cygwin
+   ISO-8859-2              Y   glibc aix hpux irix osf solaris freebsd netbsd openbsd darwin cygwin
+   ISO-8859-3              Y   glibc solaris cygwin
+   ISO-8859-4              Y   osf solaris freebsd netbsd openbsd darwin
+   ISO-8859-5              Y   glibc aix hpux irix osf solaris freebsd netbsd openbsd darwin cygwin
+   ISO-8859-6              Y   glibc aix hpux solaris cygwin
+   ISO-8859-7              Y   glibc aix hpux irix osf solaris netbsd openbsd darwin cygwin
+   ISO-8859-8              Y   glibc aix hpux osf solaris cygwin
+   ISO-8859-9              Y   glibc aix hpux irix osf solaris darwin cygwin
+   ISO-8859-13                 glibc netbsd openbsd darwin cygwin
+   ISO-8859-14                 glibc cygwin
+   ISO-8859-15                 glibc aix osf solaris freebsd netbsd openbsd darwin cygwin
+   KOI8-R                  Y   glibc solaris freebsd netbsd openbsd darwin
+   KOI8-U                  Y   glibc freebsd netbsd openbsd darwin cygwin
+   KOI8-T                      glibc
+   CP437                       dos
+   CP775                       dos
+   CP850                       aix osf dos
+   CP852                       dos
+   CP855                       dos
+   CP856                       aix
+   CP857                       dos
+   CP861                       dos
+   CP862                       dos
+   CP864                       dos
+   CP865                       dos
+   CP866                       freebsd netbsd openbsd darwin dos
+   CP869                       dos
+   CP874                       windows dos
+   CP922                       aix
+   CP932                       aix cygwin windows dos
+   CP943                       aix
+   CP949                       osf darwin windows dos
+   CP950                       windows dos
+   CP1046                      aix
+   CP1124                      aix
+   CP1125                      dos
+   CP1129                      aix
+   CP1131                      darwin
+   CP1250                      windows
+   CP1251                      glibc solaris netbsd openbsd darwin cygwin windows
+   CP1252                      aix windows
+   CP1253                      windows
+   CP1254                      windows
+   CP1255                      glibc windows
+   CP1256                      windows
+   CP1257                      windows
+   GB2312                  Y   glibc aix hpux irix solaris freebsd netbsd darwin
+   EUC-JP                  Y   glibc aix hpux irix osf solaris freebsd netbsd darwin
+   EUC-KR                  Y   glibc aix hpux irix osf solaris freebsd netbsd darwin cygwin
+   EUC-TW                      glibc aix hpux irix osf solaris netbsd
+   BIG5                    Y   glibc aix hpux osf solaris freebsd netbsd darwin cygwin
+   BIG5-HKSCS                  glibc solaris darwin
+   GBK                         glibc aix osf solaris darwin cygwin windows dos
+   GB18030                     glibc solaris netbsd darwin
+   SHIFT_JIS               Y   hpux osf solaris freebsd netbsd darwin
+   JOHAB                       glibc solaris windows
+   TIS-620                     glibc aix hpux osf solaris cygwin
+   VISCII                  Y   glibc
+   TCVN5712-1                  glibc
+   ARMSCII-8                   glibc darwin
+   GEORGIAN-PS                 glibc cygwin
+   PT154                       glibc
+   HP-ROMAN8                   hpux
+   HP-ARABIC8                  hpux
+   HP-GREEK8                   hpux
+   HP-HEBREW8                  hpux
+   HP-TURKISH8                 hpux
+   HP-KANA8                    hpux
+   DEC-KANJI                   osf
+   DEC-HANYU                   osf
+   UTF-8                   Y   glibc aix hpux osf solaris netbsd darwin cygwin
+
+   Note: Names which are not marked as being a MIME name should not be used in
+   Internet protocols for information interchange (mail, news, etc.).
+
+   Note: ASCII and ANSI_X3.4-1968 are synonymous canonical names.  Applications
+   must understand both names and treat them as equivalent.
+ */
+
 
 #ifdef __cplusplus
 }