#!/usr/local/bin/jperl -- -*-Perl-*-
#
# nkf -e /cdrom/cdrom0/mai94.txt | jperl mai2sgml.pl > mai94.sgml  (113,106,483 byte)
# nkf -e /cdrom/cdrom0/mai95.txt | jperl mai2sgml.pl > mai95.sgml  (118,720,559 byte)
#

use I18N::Japanese;


$ad{"01"} = "１面";
$ad{"02"} = "２面";
$ad{"03"} = "３面";
$ad{"04"} = "解説";
$ad{"05"} = "社説";
$ad{"07"} = "国際";
$ad{"08"} = "経済";
$ad{"10"} = "特集";
$ad{"12"} = "総合";
$ad{"13"} = "家庭";
$ad{"14"} = "文化";
$ad{"15"} = "読書";
$ad{"16"} = "科学";
$ad{"18"} = "芸能";
$ad{"35"} = "スポーツ";
$ad{"41"} = "社会";

$tag{"C0"} = "DOCNO";
$tag{"AD"} = "SECTION";
$tag{"AE"} = "AE";
$tag{"S1"} = "WORDS";
$tag{"T1"} = "HEADLINE";
$tag{"T2"} = "TEXT";

sub zen2han($) {
    $_[0] =~ tr/　！”＃＄％＆’（）＊＋，−．／０-９：；＜＝＞？＠Ａ-Ｚ［￥］＾―　ａ-ｚ｛｜｝￣　/ !-~/;
    $_[0];
}

sub transfer($$) {
    my ( $key, $context ) = @_;
    my $data;

    if (  $key eq 'AF' ) {
	$data = zen2han( $context );
    } elsif (  $key eq 'C0' ) {
	$data = zen2han( $context );
	$year = substr($data, 0, 2 );
    } elsif ( $key eq 'AE' ) {
	$data  = ( $context eq 'Ｙ' ) ? '有' : '無' ;
    } elsif ( $key eq 'S1' ) {
	my $size;
	( $size ) = /.*（全(.*)文字）/;
	$data = zen2han( $size );
    } elsif ( $key eq 'AD' ) {
	$data = $ad{zen2han($context)}
    } else {
	$data = $context;
    }

    $data;
}

sub output {
    my $key;

    print "<DOC>\n";

    foreach $key ( 'C0', 'AD', 'AE', 'S1', 'T1' ) {
	print "<", $tag{$key}, ">", $keyword{$key}->[0], "</", $tag{$key}, ">\n";
    }
    foreach $key ( 'T2' ) {
	print "<",$tag{$key},">\n", join("\n",@{$keyword{$key}}), "\n</",$tag{$key},">\n";
    }
    print "</DOC>\n";
}

$first = 1;

while (<>) {
    chomp;
    ( $tag, $context ) = /＼(.*)＼(.*)/;
    $key = zen2han( $tag );
    $data = transfer( $key, $context );
    if ( $key eq "ID" ) {
	if ( $first == 1 ) {
	    $first = 0;
	} elsif ( $first == 0 ) {
	    output;
	    undef %keyword;
	    $first = -1;
	} else {
	    print "\n";
	    output;
	    undef %keyword;
	}
    }
    $keyword{$key} = [] unless $keyword{$key};
    push @{$keyword{$key}}, $data;
}

output;