package MIME::WordDecoder;


=head1 NAME

MIME::WordDecoder - decode RFC-1522 encoded words to a local representation


=head1 SYNOPSIS

See L<MIME::Words> for the basics of encoded words.
See L<"DESCRIPTION"> for how this class works.

    use MIME::WordDecoder;
     
    
    ### Get the default word-decoder (used by unmime()):
    $wd = default MIME::WordDecoder;
      
    ### Get a word-decoder which maps to ISO-8859-1 (Latin1):
    $wd = supported MIME::WordDecoder "ISO-8859-1";
    
       
    ### Decode a MIME string (e.g., into Latin1) via the default decoder:
    $str = $wd->decode('To: =?ISO-8859-1?Q?Keld_J=F8rn_Simonsen?= <keld>');
      
    ### Decode a string using the default decoder, non-OO style:
    $str = unmime('To: =?ISO-8859-1?Q?Keld_J=F8rn_Simonsen?= <keld>');
 

=head1 DESCRIPTION

A MIME::WordDecoder consists, fundamentally, of a hash which maps
a character set name (US-ASCII, ISO-8859-1, etc.) to a subroutine which 
knows how to take bytes in that character set and turn them into 
the target string representation.  Ideally, this target representation 
would be Unicode, but we don't want to overspecify the translation 
that takes place: if you want to convert MIME strings directly to Big5, 
that's your own decision.

The subroutine will be invoked with two arguments: DATA (the data in
the given character set), and CHARSET (the upcased character set name).

For example:

    ### Keep 7-bit characters as-is, convert 8-bit characters to '#':
    sub keep7bit {  
	local $_ = shift;
	tr/\x00-\x7F/#/c;
	$_;
    }

Here's a decoder which uses that:

   ### Construct a decoder:
   $wd = MIME::WordDecoder->new({'US-ASCII'   => "KEEP",   ### sub { $_[0] }
                                 'ISO-8859-1' => \&keep7bit,
                                 'ISO-8859-2' => \&keep7bit,
                                 'Big5'       => "WARN",
                                 '*'          => "DIE"});
         
   ### Convert some MIME text to a pure ASCII string...   
   $ascii = $wd->decode('To: =?ISO-8859-1?Q?Keld_J=F8rn_Simonsen?= <keld>');
      
   ### ...which will now hold: "To: Keld J#rn Simonsen <keld>"



=head1 PUBLIC INTERFACE

=over

=cut

use strict;
use Carp qw( carp croak );
use MIME::Words qw(decode_mimewords);
use Exporter;
use vars qw(@ISA @EXPORT);

@ISA = qw(Exporter);
@EXPORT = qw( unmime );



#------------------------------
#
# Globals
# 
#------------------------------

### Decoders.
my %DecoderFor = ();

### Standard handlers.
my %Handler = 
(
 KEEP   => sub {$_[0]},
 IGNORE => sub {''},
 WARN   => sub { carp "ignoring text in character set `$_[1]'\n" },
 DIE    => sub { croak "can't handle text in character set `$_[1]'\n" },
 );

### Global default decoder.  We init it below.
my $Default;


#------------------------------

=item default [DECODER]

I<Class method.>
Get/set the default DECODER object.

=cut

sub default {
    my $class = shift;
    if (@_) {
	$Default = shift;
    }
    $Default;
}

#------------------------------

=item supported CHARSET, [DECODER]

I<Class method.>
If just CHARSET is given, returns a decoder object which maps
data into that character set (the character set is forced to
all-uppercase).  

    $wd = supported MIME::WordDecoder "ISO-8859-1";

If DECODER is given, installs such an object:

    MIME::WordDecoder->supported("ISO-8859-1" => 
				 (new MIME::WordDecoder::ISO_8859 "1"));

You should not override this method.

=cut

sub supported {
    my ($class, $charset, $decoder) = @_;
    $DecoderFor{uc($charset)} = $decoder if (@_ > 2);
    $DecoderFor{uc($charset)};
}

#------------------------------

=item new [\@HANDLERS]

I<Class method, constructor.>
If \@HANDLERS is given, then @HANDLERS is passed to handler()
to initiallize the internal map.

=cut

sub new {
    my ($class, $h) = @_;
    my $self = bless { MWD_Map=>{} }, $class;   

    ### Init the map:
    $self->handler(@$h);
    
    ### Add fallbacks:
    $self->{MWD_Map}{'*'}   ||= $Handler{WARN};
    $self->{MWD_Map}{'raw'} ||= $self->{MWD_Map}{'US-ASCII'};
    $self;
}

#------------------------------

=item handler CHARSET=>\&SUBREF, ...

I<Instance method.>
Set the handler SUBREF for a given CHARSET, for as many pairs
as you care to supply.

When performing the translation of a MIME-encoded string, a
given SUBREF will be invoked when translating a block of text
in character set CHARSET.  The subroutine will be invoked with 
the following arguments:

    DATA    - the data in the given character set.
    CHARSET - the upcased character set name, which may prove useful
              if you are using the same SUBREF for multiple CHARSETs.
    DECODER - the decoder itself, if it contains configuration information
              that your handler function needs.

For example:

    $wd = new MIME::WordDecoder;
    $wd->handler('US-ASCII'   => "KEEP");
    $wd->handler('ISO-8859-1' => \&handle_latin1,
		 'ISO-8859-2' => \&handle_latin1,
		 '*'          => "DIE");

Notice that, much as with %SIG, the SUBREF can also be taken from
a set of special keywords:

   KEEP     Pass data through unchanged.
   IGNORE   Ignore data in this character set, without warning.
   WARN     Ignore data in this character set, with warning.
   DIE      Fatal exception with "can't handle character set" message.

The subroutine for the special CHARSET of 'raw' is used for raw
(non-MIME-encoded) text, which is supposed to be US-ASCII.  
The handler for 'raw' defaults to whatever was specified for 'US-ASCII'
at the time of construction.

The subroutine for the special CHARSET of '*' is used for any 
unrecognized character set.  The default action for '*' is WARN.

=cut

sub handler {
    my $self = shift;
    
    ### Copy the hash, and edit it:
    while (@_) {
	my $c   = shift; 
	my $sub = shift;
	$self->{MWD_Map}{$c} = $self->real_handler($sub);
    }
    $self;
}

#------------------------------

=item decode STRING

I<Instance method.>
Decode a STRING which might contain MIME-encoded components into a
local representation (e.g., UTF-8, etc.).

=cut

sub decode {
    my ($self, $str) = @_;
    defined($str) or return undef;
    join('', map {
	### Get the data and (upcased) charset:
	my $data    = $_->[0];
	my $charset = (defined($_->[1]) ? uc($_->[1]) : 'raw');
	$charset =~ s/\*\w+\Z//;   ### RFC2184 language suffix

	### Get the handler; guess if never seen before:
	defined($self->{MWD_Map}{$charset}) or
	    $self->{MWD_Map}{$charset} = 
		($self->real_handler($self->guess_handler($charset)) || 0);
	my $subr = $self->{MWD_Map}{$charset} || $self->{MWD_Map}{'*'}; 

	### Map this chunk:
	&$subr($data, $charset, $self);
    } decode_mimewords($str));
}

#------------------------------
#
# guess_handler CHARSET    
#
# Instance method.  
# An unrecognized charset has been seen.  Guess a handler subref 
# for the given charset, returning false if there is none.
# Successful mappings will be cached in the main map.
#
sub guess_handler {
    undef;
}

#------------------------------
#
# real_handler HANDLER
#
# Instance method.  
# Translate the given handler, which might be a subref or a string.
#
sub real_handler {
    my ($self, $sub) = @_;
    (!$sub) or 
	(ref($sub) eq 'CODE') or 
	    $sub = ($Handler{$sub} || croak "bad named handler: $sub\n");
    $sub;
}

#------------------------------

=item unmime STRING

I<Function, exported.>
Decode the given STRING using the default() decoder.
See L<default()|/default>.

=cut

sub unmime($) {
    my $str = shift;
    $Default->decode($str);
}


=back

=cut





=head1 SUBCLASSES

=over

=cut

#------------------------------------------------------------
#------------------------------------------------------------

=item MIME::WordDecoder::ISO_8859

A simple decoder which keeps US-ASCII and the 7-bit characters
of ISO-8859 character sets and UTF8, and also keeps 8-bit
characters from the indicated character set.

    ### Construct:
    $wd = new MIME::WordDecoder::ISO_8859 2;    ### ISO-8859-2
       
    ### What to translate unknown characters to (can also use empty):
    ### Default is "?".
    $wd->unknown("?");
    
    ### Collapse runs of unknown characters to a single unknown()?
    ### Default is false.
    $wd->collapse(1);


According to B<http://czyborra.com/charsets/iso8859.html> 
(ca. November 2000):

ISO 8859 is a full series of 10 (and soon even more) standardized
multilingual single-byte coded (8bit) graphic character sets for
writing in alphabetic languages:

    1. Latin1 (West European) 
    2. Latin2 (East European) 
    3. Latin3 (South European) 
    4. Latin4 (North European) 
    5. Cyrillic 
    6. Arabic 
    7. Greek 
    8. Hebrew 
    9. Latin5 (Turkish) 
   10. Latin6 (Nordic) 

The ISO 8859 charsets are not even remotely as complete as the truly
great Unicode but they have been around and usable for quite a while
(first registered Internet charsets for use with MIME) and have
already offered a major improvement over the plain 7bit US-ASCII.

Characters 0 to 127 are always identical with US-ASCII and the
positions 128 to 159 hold some less used control characters: the
so-called C1 set from ISO 6429.

=cut

package MIME::WordDecoder::ISO_8859;

use strict;
use vars qw(@ISA);
@ISA = qw( MIME::WordDecoder );


#------------------------------
#
# HANDLERS
#
#------------------------------

### Keep 7bit characters.
### Turn all else to the special \x00.
sub h_keep7bit {  
    local $_    = $_[0];
#   my $unknown = $_[2]->{MWDI_Unknown};

    s{[\x80-\xFF]}{\x00}g;
    $_;
}

### Note: should use Unicode::String, converting/manipulating 
### everything into full Unicode form.

### Keep 7bit UTF8 characters (ASCII).
### Keep ISO-8859-1 if this decoder is for Latin-1.
### Turn all else to the special \x00.
sub h_utf8 {  
    local $_    = $_[0];
#   my $unknown = $_[2]->{MWDI_Unknown};
    my $latin1 = ($_[2]->{MWDI_Num} == 1);
    print STDERR "UTF8 in:  <$_>\n"; 

    my $tgt = '';
    while (m{\G(
          ([\x00-\x7F])                | # 0xxxxxxx
	  ([\xC0-\xDF] [\x80-\xBF])    | # 110yyyyy 10xxxxxx
	  ([\xE0-\xEF] [\x80-\xBF]{2}) | # 1110zzzz 10yyyyyy 10xxxxxx
	  ([\xF0-\xF7] [\x80-\xBF]{3}) | # 11110uuu 10uuzzzz 10yyyyyy 10xxxxxx
	  .                              # error; synch
	  )}gcsx and ($1 ne '')) {   

	if    (defined($2))            { $tgt .= $2 }
	elsif (defined($3) && $latin1) { $tgt .= "\x00" }
        else                           { $tgt .= "\x00" }
    }

    print STDERR "UTF8 out: <$tgt>\n"; 
    $tgt;
}

### Keep characters which are 7bit in UTF8 (ASCII).
### Keep ISO-8859-1 if this decoder is for Latin-1.
### Turn all else to the special \x00.
sub h_utf16 {  
    local $_    = $_[0];
#   my $unknown = $_[2]->{MWDI_Unknown};
    my $latin1 = ($_[2]->{MWDI_Num} == 1);
    print STDERR "UTF16 in:  <$_>\n"; 

    my $tgt = '';
    while (m{\G(
		(  \x00  ([\x00-\x7F])) |  # 00000000 0xxxxxxx
		(  \x00  ([\x80-\xFF])) |  # 00000000 1xxxxxxx
		( [^\x00] [\x00-\xFF])  |  # etc
		)
	     }gcsx and ($1 ne '')) {

	if    (defined($2))            { $tgt .= $3 }
	elsif (defined($4) && $latin1) { $tgt .= $5 }
        else                           { $tgt .= "\x00" }
    }

    print STDERR "UTF16 out: <$tgt>\n"; 
    $tgt;
}


#------------------------------
#
# PUBLIC INTERFACE
#
#------------------------------

#------------------------------
#
# new NUMBER
#
sub new {
    my ($class, $num) = @_;

    my $self = $class->SUPER::new();
    $self->handler('raw'      => 'KEEP',
		   'US-ASCII' => 'KEEP');

    $self->{MWDI_Num} = $num;
    $self->{MWDI_Unknown} = "?";
    $self->{MWDI_Collapse} = 0;
    $self;
}

#------------------------------
#
# guess_handler CHARSET
#
sub guess_handler {
    my ($self, $charset) = @_;
    return 'KEEP'              if (($charset =~ /^ISO[-_]?8859[-_](\d+)$/) && 
				   ($1 eq $self->{MWDI_Num}));
    return \&h_keep7bit        if ($charset =~ /^ISO[-_]?8859/);
    return \&h_utf8            if ($charset =~ /^UTF[-_]?8$/);
    return \&h_utf16           if ($charset =~ /^UTF[-_]?16$/);
    undef;
}

#------------------------------
#
# unknown [REPLACEMENT]
#
sub unknown {
    my $self = shift;
    $self->{MWDI_Unknown} = shift if @_;
    $self->{MWDI_Unknown};
}

#------------------------------
#
# collapse [YESNO]
#
sub collapse {
    my $self = shift;
    $self->{MWDI_Collapse} = shift if @_;
    $self->{MWDI_Collapse};
}

#------------------------------
#
# decode STRING
#
sub decode {
    my $self = shift;

    ### Do inherited action:
    my $basic = $self->SUPER::decode(@_);
    defined($basic) or return undef;

    ### Translate/consolidate illegal characters:
    $basic =~ tr{\x00}{\x00}c     if $self->{MWDI_Collapse};
    $basic =~ s{\x00}{$self->{MWDI_Unknown}}g;
    $basic;
}

#------------------------------------------------------------
#------------------------------------------------------------

=item MIME::WordDecoder::US_ASCII

A subclass of the ISO-8859-1 decoder which discards 8-bit characters.  
You're probably better off using ISO-8859-1.

=cut

package MIME::WordDecoder::US_ASCII;

use strict;
use vars qw(@ISA);
@ISA = qw( MIME::WordDecoder::ISO_8859 );

sub new {
    my ($class) = @_;
    return $class->SUPER::new("1");
}

sub decode {
    my $self = shift;

    ### Do inherited action:
    my $basic = $self->SUPER::decode(@_);
    defined($basic) or return undef;

    ### Translate/consolidate 8-bit characters:
    $basic =~ tr{\x80-\xFF}{}c     if $self->{MWDI_Collapse};
    $basic =~ s{[\x80-\xFF]}{$self->{MWDI_Unknown}}g;
    $basic;
}

=back

=cut

#------------------------------------------------------------
#------------------------------------------------------------

package MIME::WordDecoder;

### Now we can init the default handler.
$Default = (MIME::WordDecoder::ISO_8859->new('1'));

### Add US-ASCII handler:
$DecoderFor{"US-ASCII"} = MIME::WordDecoder::US_ASCII->new;

### Add ISO-8859-{1..15} handlers:
for (1..15) { 
    $DecoderFor{"ISO-8859-$_"} = MIME::WordDecoder::ISO_8859->new($_);
}

{ 
  package main; no strict; local $^W = 0;
  my @x = <::DATA>;
  eval join('',<::DATA>) || die $@ unless caller();
}
1;           # end the module
__END__


=head1 AUTHOR

Eryq (F<eryq@zeegee.com>), ZeeGee Software Inc (F<http://www.zeegee.com>).
David F. Skoll (dfs@roaringpenguin.com) http://www.roaringpenguin.com


=head1 VERSION

$Revision: 1.2 $ $Date: 2004/09/07 15:13:54 $

=cut


BEGIN { unshift @INC, ".", "./etc", "./lib" };
import MIME::WordDecoder;

### Decode a MIME string (e.g., into Latin1) via the default decoder:
my $charset = $ARGV[0] || 'ISO-8859-1';
my $wd = MIME::WordDecoder->supported($charset) || die "unsupported charset: $charset\n";

$wd->unknown('#');
my @encs = (
	    'ASCII:  =?US-ASCII?Q?Keith_Moore?= <moore@cs.utk.edu>',
	    'Latin1: =?ISO-8859-1?Q?Keld_J=F8rn_Simonsen?= <keld@dkuug.dk>',
	    'Latin1: =?ISO-8859-1?Q?Andr=E9_?= Pirard <PIRARD@vm1.ulg.ac.be>',
	    'Latin1: =?ISO-8859-1?Q?Andr=E9_?=Pirard <PIRARD@vm1.ulg.ac.be>',
	    ' UTF-8: =?UTF-8?Q?Andr=E9_?=Pirard <PIRARD@vm1.ulg.ac.be>',
	    'UTF-16: =?UTF-16?Q?=00A=00n=00d=00r=00=E9?= Pirard <PIRARD@vm1.ulg.ac.be>',
	    ('=?ISO-8859-1?B?SWYgeW91IGNhbiByZWFkIHRoaXMgeW8=?='.
	     '=?ISO-8859-2?B?dSB1bmRlcnN0YW5kIHRoZSBleGFtcGxlLg==?='.
	     '=?US-ASCII?Q?.._cool!?='));
$str = $wd->decode(join "\n", @encs);
print "$str\n";
1;