1 package Encode::CN::HZ;
7 $VERSION = do { my @r = ( q$Revision: 2.4 $ =~ /\d+/g ); sprintf "%d." . "%02d" x $#r, @r };
9 use Encode qw(:fallbacks);
11 use base qw(Encode::Encoding);
12 __PACKAGE__->Define('hz');
14 # HZ is a combination of ASCII and escaped GB, so we implement it
15 # with the GB2312(raw) encoding here. Cf. RFCs 1842 & 1843.
17 # not ported for EBCDIC. Which should be used, "~" or "\x7E"?
22 my ( $obj, $str, $chk ) = @_;
24 my $GB = Encode::find_encoding('gb2312-raw');
26 my $in_ascii = 1; # default mode is ASCII.
28 while ( length $str ) {
29 if ($in_ascii) { # ASCII mode
30 if ( $str =~ s/^([\x00-\x7D\x7F]+)// ) { # no '~' => ASCII
33 # EBCDIC should need ascii2native, but not ported.
35 elsif ( $str =~ s/^\x7E\x7E// ) { # escaped tilde
38 elsif ( $str =~ s/^\x7E\cJ// ) { # '\cJ' == LF in ASCII
41 elsif ( $str =~ s/^\x7E\x7B// ) { # '~{'
42 $in_ascii = 0; # to GB
44 else { # encounters an invalid escape, \x80 or greater
48 else { # GB mode; the byte ranges are as in RFC 1843.
49 no warnings 'uninitialized';
50 if ( $str =~ s/^((?:[\x21-\x77][\x21-\x7E])+)// ) {
51 $ret .= $GB->decode( $1, $chk );
53 elsif ( $str =~ s/^\x7E\x7D// ) { # '~}'
61 $_[1] = '' if $chk; # needs_lines guarantees no partial character
66 my ( $obj, undef, $src, $pos, $trm, $chk ) = @_;
67 my ( $rdst, $rsrc, $rpos ) = \@_[ 1 .. 3 ];
69 my $GB = Encode::find_encoding('gb2312-raw');
71 my $in_ascii = 1; # default mode is ASCII.
73 my $ini_pos = pos($$rsrc);
75 substr( $src, 0, $pos ) = '';
77 my $ini_len = bytes::length($src);
79 # $trm is the first of the pair '~~', then 2nd tilde is to be removed.
80 # XXX: Is better C<$src =~ s/^\x7E// or die if ...>?
81 $src =~ s/^\x7E// if $trm eq "\x7E";
83 while ( length $src ) {
85 if ($in_ascii) { # ASCII mode
86 if ( $src =~ s/^([\x00-\x7D\x7F])// ) { # no '~' => ASCII
89 elsif ( $src =~ s/^\x7E\x7E// ) { # escaped tilde
92 elsif ( $src =~ s/^\x7E\cJ// ) { # '\cJ' == LF in ASCII
95 elsif ( $src =~ s/^\x7E\x7B// ) { # '~{'
96 $in_ascii = 0; # to GB
99 else { # encounters an invalid escape, \x80 or greater
103 else { # GB mode; the byte ranges are as in RFC 1843.
104 if ( $src =~ s/^((?:[\x21-\x77][\x21-\x7F])+)// ) {
105 $now = $GB->decode( $1, $chk );
107 elsif ( $src =~ s/^\x7E\x7D// ) { # '~}'
116 next if !defined $now;
120 if ( $now eq $trm ) {
122 $$rpos = $ini_pos + $pos + $ini_len - bytes::length($src);
123 pos($$rsrc) = $ini_pos;
129 $$rpos = $ini_pos + $pos + $ini_len - bytes::length($src);
130 pos($$rsrc) = $ini_pos;
131 return ''; # terminator not found
135 my ( $obj, $str, $chk ) = @_;
137 my $GB = Encode::find_encoding('gb2312-raw');
139 my $in_ascii = 1; # default mode is ASCII.
141 no warnings 'utf8'; # $str may be malformed UTF8 at the end of a chunk.
143 while ( length $str ) {
144 if ( $str =~ s/^([[:ascii:]]+)// ) {
146 $tmp =~ s/~/~~/g; # escapes tildes
148 $ret .= "\x7E\x7D"; # '~}'
151 $ret .= pack 'a*', $tmp; # remove UTF8 flag.
153 elsif ( $str =~ s/(.)// ) {
155 my $tmp = $GB->encode( $s, $chk );
156 last if !defined $tmp;
157 if ( length $tmp == 2 ) { # maybe a valid GB char (XXX)
159 $ret .= "\x7E\x7B"; # '~{'
164 elsif ( length $tmp ) { # maybe FALLBACK in ASCII (XXX)
166 $ret .= "\x7E\x7D"; # '~}'
172 else { # if $str is malformed UTF8 *and* if length $str != 0.
176 $_[1] = $str if $chk;
178 # The state at the end of the chunk is discarded, even if in GB mode.
179 # That results in the combination of GB-OUT and GB-IN, i.e. "~}~{".
180 # Parhaps it is harmless, but further investigations may be required...
183 $ret .= "\x7E\x7D"; # '~}'
194 Encode::CN::HZ -- internally used by Encode::CN