Move Encode alias definitions to charset.cfg, register also our non-recommended but recognized aliases.

2 files changed, 20 insertions, 34 deletions

diff --git a/htdocs/config/charset.cfg b/htdocs/config/charset.cfg
index 22b9358..8e47c85 100644
--- a/htdocs/config/charset.cfg
+++ b/htdocs/config/charset.cfg
@@ -1,7 +1,7 @@
 #
 # list of accepted/preferred character encodings 
 #
-# $Id: charset.cfg,v 1.16 2007-07-20 02:58:10 ot Exp $
+# $Id: charset.cfg,v 1.17 2009-06-29 19:57:51 ville Exp $
 #
 # Syntax:
 #
@@ -11,11 +11,13 @@
 #
 # ? indicates the action to take:
 # 1: OK, character supported
+# A: OK, character supported, Encode::Alias it to result
 # X: frequent error, e.g. starting with x-; ask user to replace with result
 # ERR: a charset we refuse, per some policy. Reason stated after ERR
 
 #e.g:
 # utf-8         = 1
+# some-alias    = A perl-Encode-name
 # odd-alias     = X good-alias
 # bad_charset   = ERR explain reason
 
@@ -30,11 +32,11 @@ iso-8859-4                      = 1
 iso-8859-5                      = 1
 iso-8859-6                      = 1
 # implicit bidi, but character encoding is the same
-iso-8859-6-i                    = 1
+iso-8859-6-i                    = A iso-8859-6
 iso-8859-7                      = 1
 iso-8859-8                      = 1
 # implicit bidi, but character encoding is the same
-iso-8859-8-i                    = 1
+iso-8859-8-i                    = A iso-8859-8
 iso-8859-9                      = 1
 iso-8859-10                     = 1
 iso-8859-11                     = 1
@@ -53,7 +55,9 @@ big5-hkscs                      = 1
 iso-2022-kr                     = 1
 euc-kr                          = 1
 gb18030                         = 1
-tis-620                         = 1
+# 0xA0 is U+00A0 in ISO-8859-11 but undefined in tis-620
+# other than that the character encodings are equivalent
+tis-620                         = A iso-8859-11
 koi8-r                          = 1
 koi8-u                          = 1
 iso-ir-111                      = 1
@@ -66,9 +70,11 @@ windows-1255                    = 1
 windows-1256                    = 1
 windows-1257                    = 1
 # windows-1258                  = 1
-macintosh                       = 1
+# Encode::Byte does not know 'macintosh' but MacRoman
+macintosh                       = A MacRoman
 ks_c_5601-1987                  = 1
-ksc_5601                        = 1
+# Encode only knows the long hand version of 'ksc_5601'
+ksc_5601                        = A KS_C_5601-1987
 
 x-mac-roman                     = X macintosh
 x-sjis                          = X shift_jis
diff --git a/httpd/cgi-bin/check b/httpd/cgi-bin/check
index 8a12b4b..6dcb610 100755
--- a/httpd/cgi-bin/check
+++ b/httpd/cgi-bin/check
@@ -14,7 +14,7 @@
 # This source code is available under the license at:
 #     http://www.w3.org/Consortium/Legal/copyright-software
 #
-# $Id: check,v 1.664 2009-06-29 18:21:16 ville Exp $
+# $Id: check,v 1.665 2009-06-29 19:57:51 ville Exp $
 #
 # Disable buffering on STDOUT!
 $| = 1;
@@ -191,6 +191,12 @@ Directory not readable (permission denied): @_r
   }
 
   #
+  # Register Encode aliases.
+  while (my ($key, $value) = each %{$CFG->{Charsets}}) {
+    Encode::Alias::define_alias($key, $1) if ($value =~ /^[AX] (\S+)/);
+  }
+
+  #
   # Set debug flag.
   if ($CFG->{'Allow Debug'}) {
     $DEBUG = TRUE if $ENV{W3C_VALIDATOR_DEBUG} || $CFG->{'Enable Debug'};
@@ -200,7 +206,7 @@ Directory not readable (permission denied): @_r
 
   #
   # Strings
-  $VERSION =  q$Revision: 1.664 $;
+  $VERSION =  q$Revision: 1.665 $;
   $VERSION =~ s/Revision: ([\d\.]+) /$1/;
 
   #
@@ -568,31 +574,6 @@ unless ($File->{Charset}->{XML} || $File->{Charset}->{META}){ #suggest character
 # Abort if an error was flagged while finding the encoding.
 &abort_if_error_flagged($File, O_CHARSET|O_DOCTYPE);
 
-#
-# Encode alias definitions. This might not be the best
-# place for them, feel free to move them elsewhere.
-
-# implicit bidi, but character encoding is the same
-Encode::Alias::define_alias('iso-8859-6-i', 'iso-8859-6');
-
-# implicit bidi, but character encoding is the same
-Encode::Alias::define_alias('iso-8859-8-i', 'iso-8859-8');
-
-# 0xA0 is U+00A0 in ISO-8859-11 but undefined in tis-620
-# other than that the character encodings are equivalent
-Encode::Alias::define_alias('tis-620', 'iso-8859-11');
-
-# Encode::Byte does not know 'macintosh' but MacRoman
-Encode::Alias::define_alias('macintosh', 'MacRoman');
-
-# x-mac-roman is the non-standard version of 'macintosh'
-Encode::Alias::define_alias('x-mac-roman', 'MacRoman');
-
-# Encode only knows the long hand version of 'ksc_5601'
-Encode::Alias::define_alias('ksc_5601', 'KS_C_5601-1987');
-
-# gb18030 requires Encode::HanExtra but no additional alias
-
 $File->{Charset}->{Default} = FALSE;
 unless ($File->{Charset}->{Use}) { # No charset given...
      $File->{Charset}->{Use} = 'utf-8';
@@ -601,7 +582,6 @@ unless ($File->{Charset}->{Use}) { # No charset given...
     &add_warning('W04', {W04_charset => "UTF-8"});
 }
 
-
 # Always transcode, even if the content claims to be UTF-8
 $File = transcode($File);
 if (($File->{ContentType} eq "text/html") and ($File->{Charset}->{Default}) and $File->{'Error Flagged'}) {