#!/usr/bin/perl -s

use Guesser;
use strict;

our ($id,$h);

$id ||= "_corpus_name_";
my $dir = shift || "./";

if ($h) {
 die("usage mkbitextra -id=name File.paths
      File.blocks
      File.pairs\n");
}

if ($dir =~ m!([^/]+)\.pairs$!) {
  $id = $1;
} else {
  unless ($dir =~ /\.blocks$/){
    unless ($dir =~ /\.paths$/){
      printf STDERR ("\n****** Calculo da lista de ficheiros\n");
      system("find $dir > $id.paths");
    }

    printf STDERR ("****** Calculo de blocos\n");
    Guesser::list2blocks("$id.paths");
  }

  printf STDERR ("****** Blocos para pares PT/EN\n");
  Guesser::blocks2pairs("portuguese","english","$id.blocks");

  printf STDERR ("****** Verificar correspondencias\n");
  system("time bitextcheck -lang1=pt -lang2=en -f $id.blocks.pairs > $id.pairs");
}

printf STDERR ("****** Converter HTML para PML\n");
system("html2pml -name=$id -listofpairs -noimg  $id.pairs") == 0 or die $!;

printf STDERR ("****** Alinhar frases em CWB\n");
system("xmlalign2cqp $id.A.out $id.B.out") == 0 or die $!;

printf STDERR ("****** CWB para TMX\n");
my $corpusname = id2corpusname("$id.A.out");
system("cqpalign2tmx $corpusname > $id.tmx") == 0 or die $!;

printf STDERR ("****** Conversão TMX para input do twente aligner\n");
# system("tmxsplit -twente  $id.tmx ");

printf STDERR ("****** Alinhamento à palavra (usando Natools)\n");
# system("nat-these $id.tmx-Ling1 $id.tmx-Ling2");

sub id2corpusname{
 my $n=shift;
 for ($n){
   s/\./-/g;
   s/[A-Z]/lc($&)/ge;
 }
 $n;
}