pre

#!/usr/bin/perl

use strict;
use warnings;

my $path = `pwd`;
chop $path;

# usage: <lang> <charset>

die "usage: $0 <lang> <charset>\n" unless @ARGV == 2;

my $lang = $ARGV[0];
if ($lang !~ /^[a-z]{2,3}(|_[A-Z]{2})(|-[a-z0-9]+)$/) {
  die "invalid language string: $lang\n";
  # FIXME: Make more informative
}

my @included_charsets = qw(cp1250 cp1251 cp1252 cp1253 cp1254 cp1255
                           cp1256 cp1257 cp1258 iso-8859-10
                           iso-8859-11 iso-8859-13 iso-8859-14
                           iso-8859-15 iso-8859-16 iso-8859-1
                           iso-8859-2 iso-8859-3 iso-8859-4 iso-8859-5
                           iso-8859-6 iso-8859-7 iso-8859-8 iso-8859-9
                           koi8-r);
my $charset = $ARGV[1];

my $charset_included = 0;
foreach (@included_charsets) {
  $charset_included = 1 if ($_ eq $charset);
}
if (!$charset_included) {
  -e "maps/$charset.txt" or die "Unable to find $charset.txt.\n";
  # FIXME: Make more informative 
  system ("./mkchardata maps/$charset.txt");
}

my $encoding;
if ($charset_included) {
  $encoding = $charset;
} else {
  local $/ = undef;
  open F, "$path/maps/$charset.cset" or die;
  local $_ = <F>;
  my ($cs) = /^= (.+)$/m;
  if (grep {$_ eq $cs} @included_charsets) {
    $encoding = $charset;
  } else {
    $encoding = 'utf-8';
  }
}

mkdir $lang or die "Unable to mkdir $lang";
chdir $lang or die;

open F, ">info";
print F << "---";
name_english Unknown
name_native Unknown
lang $lang
---

if (!$charset_included) {
  print F "data-file $charset.cset\n";
  print F "data-file $charset.cmap\n";
  system ("cp $path/maps/$charset.cset $path/maps/$charset.cmap .");
  system ("mkdir misc");
  system ("cp $path/maps/$charset.txt misc/");
}

print F << "---";
author:
  name Unknown
  email unknown at someplace com
copyright Unknown
#url Unknown
version 0.01-0
#source-version 0.01
accurate unknown
complete unknown
#alias $lang unknown
dict:
  name $lang
  add $lang
---

close F;

open F, ">$lang.dat";
print F "name $lang\n";
print F "charset $charset\n";
print F "#" unless $encoding eq 'utf-8';
print F "data-encoding utf-8\n";
print F "#special ' -*-\n";

open F, ">Copyright";
print F "Copyright (c) YEAR by SO AND SO under the WHAT.\n";
close F;

open F, ">$lang.wl";
close F;

system "ln -s $path/proc .";