"The last pieces of Chinese puzzle" from Autrijus.

p4raw-id: //depot/perl@15029

7 files changed, 74 insertions, 17 deletions

diff --git a/ext/Encode/CN/CN.pm b/ext/Encode/CN/CN.pm
index 7f828d35bd..b2d1795e30 100644
--- a/ext/Encode/CN/CN.pm
+++ b/ext/Encode/CN/CN.pm
@@ -1,9 +1,14 @@
 package Encode::CN;
-use Encode;
 our $VERSION = '0.02';
+
+use Encode;
+use Encode::CN::HZ;
 use XSLoader;
 XSLoader::load('Encode::CN',$VERSION);
 
+local $@;
+eval "use Encode::HanExtra"; # load extra encodings if they exist
+
 1;
 __END__
 =head1 NAME
@@ -25,7 +30,8 @@ Encodings supported are as follows.
   gb2312	The raw (low-bit) GB2312 character map
   gb12345	Traditional chinese counterpart to GB2312 (raw)
   iso-ir-165	GB2312 + GB6345 + GB8565 + additions
-  cp936	Code Page 936, also known as GBK (Extended GuoBiao)
+  cp936		Code Page 936, also known as GBK (Extended GuoBiao)
+  hz		7-bit escaped GB2312 encoding
 
 To find how to use this module in detail, see L<Encode>.
 
@@ -35,9 +41,10 @@ Due to size concerns, C<GB 18030> (an extension to C<GBK>) is distributed
 separately on CPAN, under the name L<Encode::HanExtra>. That module
 also contains extra Taiwan-based encodings.
 
-=head1 BUGS
+This module will automatically load L<Encode::HanExtra> if you have it on
+your machine.
 
-The C<HZ> (Hanzi) escaped encoding is not supported.
+=head1 BUGS
 
 ASCII part (0x00-0x7f) is preserved for all encodings, even though it
 conflicts with mappings by the Unicode Consortium.  See
diff --git a/ext/Encode/Encode.pm b/ext/Encode/Encode.pm
index d0bb788b06..445dd2401b 100644
--- a/ext/Encode/Encode.pm
+++ b/ext/Encode/Encode.pm
@@ -173,7 +173,6 @@ define_alias( qr/^gbk$/i => '"cp936"');
 # TODO: HP-UX '8' encodings arabic8 greek8 hebrew8 kana8 thai8 turkish8
 # TODO: HP-UX '15' encodings japanese15 korean15 roi15
 # TODO: Cyrillic encoding ISO-IR-111 (useful?)
-# TODO: Chinese encodings HZ
 # TODO: Armenian encoding ARMSCII-8
 # TODO: Hebrew encoding ISO-8859-8-1
 # TODO: Thai encoding TCVN
diff --git a/ext/Encode/Encode/HZ.enc b/ext/Encode/Encode/HZ.enc
deleted file mode 100644
index 748ee0bd20..0000000000
--- a/ext/Encode/Encode/HZ.enc
+++ /dev/null
@@ -1,7 +0,0 @@
-# Encoding file: HZ, HanZi
-H
-name		HZ
-init		{}
-final		{}
-ascii		\x7e\x7d
-gb2312		\x7e\x7b
diff --git a/ext/Encode/KR/KR.pm b/ext/Encode/KR/KR.pm
index aa2428128d..9936c5d22a 100644
--- a/ext/Encode/KR/KR.pm
+++ b/ext/Encode/KR/KR.pm
@@ -1,6 +1,7 @@
 package Encode::KR;
-use Encode;
 our $VERSION = '0.02';
+
+use Encode;
 use XSLoader;
 XSLoader::load('Encode::KR',$VERSION);
 
@@ -23,7 +24,7 @@ are as follows.
 
   euc-kr	EUC (Extended Unix Character)
   ksc5601	Korean standard code set
-  cp949	Code Page 949 (EUC-KR + Unified Hangul Code)
+  cp949		Code Page 949 (EUC-KR + Unified Hangul Code)
   
 To find how to use this module in detail, see L<Encode>.
 
diff --git a/ext/Encode/MANIFEST b/ext/Encode/MANIFEST
index bf34b59bf4..6300a735c2 100644
--- a/ext/Encode/MANIFEST
+++ b/ext/Encode/MANIFEST
@@ -95,7 +95,6 @@ Encode/gb12345.enc
 Encode/gb1988.enc
 Encode/gb2312.enc
 Encode/gsm0338.enc
-Encode/HZ.enc
 Encode/iso-ir-165.enc
 Encode/ir-197.enc
 Encode/jis0201.enc
@@ -155,6 +154,7 @@ lib/Encode/ucs2_le.pm
 lib/Encode/Unicode.pm
 lib/Encode/utf8.pm
 lib/Encode/XS.pm
+lib/Encode/CN/HZ.pm
 lib/Encode/Tcl/Escape.pm
 lib/Encode/Tcl/Extended.pm
 lib/Encode/Tcl/HanZi.pm
diff --git a/ext/Encode/TW/TW.pm b/ext/Encode/TW/TW.pm
index 90b046041b..c3f64fdd25 100644
--- a/ext/Encode/TW/TW.pm
+++ b/ext/Encode/TW/TW.pm
@@ -1,9 +1,13 @@
 package Encode::TW;
-use Encode;
 our $VERSION = '0.02';
+
+use Encode;
 use XSLoader;
 XSLoader::load('Encode::TW',$VERSION);
 
+local $@;
+eval "use Encode::HanExtra"; # load extra encodings if they exist
+
 1;
 __END__
 =head1 NAME
@@ -23,7 +27,7 @@ Encodings supported are as follows.
 
   big5		The original Big5 encoding
   big5-hkscs	Big5 plus Cantonese characters in Hong Kong
-  cp950	Code Page 950 (Big5 + Microsoft vendor mappings)
+  cp950		Code Page 950 (Big5 + Microsoft vendor mappings)
   
 To find how to use this module in detail, see L<Encode>.
 
@@ -33,6 +37,9 @@ Due to size concerns, C<EUC-TW> (Extended Unix Character) and C<BIG5PLUS>
 (CMEX's Big5+) are distributed separately on CPAN, under the name
 L<Encode::HanExtra>. That module also contains extra China-based encodings.
 
+This module will automatically load L<Encode::HanExtra> if you have it on
+your machine.
+
 =head1 BUGS
 
 The C<CNS11643> encoding files are not complete (only the first two planes,
diff --git a/ext/Encode/lib/Encode/CN/HZ.pm b/ext/Encode/lib/Encode/CN/HZ.pm
new file mode 100644
index 0000000000..a57ae8a971
--- /dev/null
+++ b/ext/Encode/lib/Encode/CN/HZ.pm
@@ -0,0 +1,50 @@
+package Encode::CN::HZ;
+
+use Encode::CN;
+use Encode qw|encode decode|;
+use base 'Encode::Encoding';
+
+use strict;
+
+# HZ is but escaped GB, so we implement it with the
+# GB2312(raw) encoding here. Cf. RFC 1842 & 1843.
+
+my $canon = 'hz';
+my $obj = bless {name => $canon}, __PACKAGE__;
+$obj->Define($canon);
+
+sub decode
+{
+    my ($obj,$str,$chk) = @_;
+    my $gb = Encode::find_encoding('gb2312');
+
+    $str =~ s{~(?:(~)|\n|{([^~]*)~}|)}
+             {$1 ? '~' : defined $2 ? $gb->decode($2, $chk) : ''}eg;
+
+    return $str;
+}
+
+sub encode
+{
+    my ($obj,$str,$chk) = @_;
+    my $gb = Encode::find_encoding('gb2312');
+
+    $str =~ s/~/~~/g;
+    $str =~ s/((?:	
+	\p{InCJKCompatibility}|
+	\p{InCJKCompatibilityForms}|
+	\p{InCJKCompatibilityIdeographs}|
+	\p{InCJKCompatibilityIdeographsSupplement}|
+	\p{InCJKRadicalsSupplement}|
+	\p{InCJKSymbolsAndPunctuation}|
+	\p{InCJKUnifiedIdeographsExtensionA}|
+	\p{InCJKUnifiedIdeographs}|
+	\p{InCJKUnifiedIdeographsExtensionB}|
+	\p{InEnclosedCJKLettersAndMonths}
+    )+)/'~{'.$gb->encode($1, $chk).'~}'/egx;
+
+    return $str;
+}
+
+1;
+__END__