#!/usr/bin/perl -ws
use strict; use warnings;
use Text::Perfide::WordBags;
use Text::Perfide::BookPairs;
use utf8::all;

our($bpairs,$rv,$av,$warn,$debug,$v,$nr,$same,$dv,$recalc,$normbf);
$nr     //= 3;
$rv     //= 0.2;
$av     //= 0.4;
$dv     //= 0.9;

my $options = {};
$options->{warn}   = 1 if defined($warn);
$options->{debug}  = 1 if defined($debug);
$options->{recalc} = 1 if defined($recalc);
$options->{v}      = 1 if defined($v);

$options->{nr} = $nr;
$options->{rv} = $rv;
$options->{av} = $av;
$options->{dv} = $dv;

if(@ARGV<1)     { usage(); }
elsif(@ARGV==1 and defined($same)){
                    print STDERR "File '$ARGV[0]' contains list of books to process.\n";
                    my $list = open_list($ARGV[0]);
                    calcbagfiles($list,$options);
                    calc_dupvers($list,$options);
                    rmbagfiles($list) unless defined($normbf);
}
elsif(@ARGV==2) {   print STDERR "Files '$ARGV[0]' and '$ARGV[1]' are lists of books to pair.\n";
                    my ($list_file1,$list_file2) = @ARGV;
                    my ($list1,$list2) = double_open_lists($list_file1,$list_file2);
                    foreach my $file1 (@$list1){
                        listpairs([$file1,@$list2],$options);
                    }
                    rmbagfiles($list1) unless defined($normbf);
                    rmbagfiles($list2) unless defined($normbf);
}
else            {   print STDERR "Finding pairs for book '$ARGV[0]'.\n";
                    calcbagfiles(\@ARGV,$options);
                    listpairs(\@ARGV,$options);
                    rmbagfiles(\@ARGV) unless defined($normbf);
}


sub listpairs{
    if      ($same)     {   calc_dupvers(@_);       }
    elsif   ($bpairs)   {   calc_bpairs(@_);        }
    else                {   calc_default(@_);       }
}

sub usage {
    print STDERR "Usage:\n\tpairbooks [options] book candidates*\n";
    print STDERR "\tpairbooks [options] list1 list2\n";
    print STDERR "\t\t(pair books from list1 with books from list2)\n"; 
    print STDERR "\tpairbooks -same [options] book_list\n";
    print STDERR "\t\t(find similar versions in books from book_list)\n";
}

sub open_list {
    open LIST,'<',$_[0];
    my @list = <LIST>;
    chomp @list;
    return \@list;
}

sub double_open_lists {
    my ($list_file1,$list_file2) = @_;
    open LIST1,'<',$list_file1 or die "Could not open '$list_file1'";
    open LIST2,'<',$list_file2 or die "Could not open '$list_file2'";
    my @list1 = <LIST1>;
    my @list2 = <LIST2>;
    chomp @list1;
    chomp @list2;
    return (\@list1,\@list2);
}

__END__

=head1 NAME

pairbooks - for a given book, finds its most probable pairs in a collection

=head1 SYNOPSIS

 pairbooks [options] book candidates*

 pairbooks [options] book_list1 book_list2

=head1 DESCRIPTION


=head1 Options

 -nr=3      Returns the 3 most similar candidates

 -bpairs    Output results in .bpairs format (1 pair of books per
            line, separated with a \t

 -rv=LOW    Reject value - book pairs with pairability value 
            lower that LOW will be automatically rejected.
            Default is 0.2

 -av=HIGH   Accept value - book pairs with pairability value 
            equal or above HIGH will be automatically approved.
            Default is 0.4

 -dv=VAL    Duplicates value - book pairs with pairability value
            equal or above VAL will be considered duplicates
            (the same book in the same language).
            Default is 0.9. Use with -same.

 -warn      Comments pairs with pairability value under HIGH
            (see -av).
            Rejected pairs will have a leading '# X', while
            dubious pairs will have a leading '# ?'.

 -same      Instead of finding pairs, tries to find candidates
            to be *the same book in the same language* (see -dv)

            
 -debug     Prints debug information

 -recalc    Calculate file.bag even if it exists already.
 
 -normbf    Do not remove .bag files at the end

 -v

=head1 AUTHOR

Andre Santos, andrefs@cpan.org

J.Joao Almeida, jj@di.uminho.pt

=head1 SEE ALSO

perl(1).

=cut