#!/usr/bin/perl -s
use strict;
use warnings;
use Data::Dumper;
$Data::Dumper::Indent=0;
$Data::Dumper::Terse=1;

use Text::Perfide::BookCleaner;
use File::Basename;

our ($b,$p1,$p2,$p3,$p4,$p5,$simplify,$normpar,$j,$v,$c,$minhf,
     $pipe,$latin1,$enc,$o,$dir,$dry,$outlist,$nrom_off);
my ($v1,$v2,$v3,$v4,$v5);
my $commit =1 if $c || $pipe;

nrom_off() if $nrom_off; ## turns off Roman Number titles

if (defined($j)){
	if($j =~ '1') 	{ $p1=1			} else { $p1=0;       }
	if($j =~ '2') 	{ $p2=1			} else { $p2=0;       }
	if($j =~ '3') 	{ $p3=1			} else { $p3=0;       }
	if($j =~ '4') 	{ $p4=1			} else { $p4=0;       }
	if($j =~ '5') 	{ $p5=1			} else { $p5=0;       }
	if($j =~ 'c') 	{ $commit=1		} else { $commit=0;   }
	if($j =~ 'p') 	{ $pipe=1;$commit=1	} else { $pipe=0; }
	#if($j =~ 's') 	{ $simplify=1	} else { $simplify=0; }
	#if($j =~ 'np')	{ $normpar=1	} else { $normpar=0;  }
}

if (defined($v)){
	if($v =~ '1') 	{ $v1=1			} else { $v1=0;       }
	if($v =~ '2') 	{ $v2=1			} else { $v2=0;       }
	if($v =~ '3') 	{ $v3=1			} else { $v3=0;       }
	if($v =~ '4') 	{ $v4=1			} else { $v4=0;       }
	if($v =~ '5') 	{ $v5=1			} else { $v5=0;       }
}

$p1       //= 1;
$p2       //= 1;
$p3       //= 1;
$p4       //= 0;
$p5       //= 1;
$commit   //= 0;
$simplify //= 1; 
$normpar  //= 1;
$pipe     //= 0;
$enc      //="";
$minhf    //= 5;
$enc = "CP1252" if $latin1;
my $htreshold = $minhf;
my $ftreshold = $minhf;

my $cp1252_simplification = $simplify;
my $utf_simplification    = $simplify;

my @books;
if(@ARGV == 1 and $ARGV[0] =~ /dbooks$/){
	print STDERR "Argument '$ARGV[0]' contains list of books to clean.\n";
	open BOOKS,'<',$ARGV[0];
	@books = <BOOKS>;
	chomp @books;
}
else{
	@books = @ARGV;
}
files(@books);

sub files{
  my $di1 ={};
  my $di2 ={};
  my $di3 ={};
  my $di4 ={};
  my $di5 ={};
  my $pbtab={};
  my $txt;
  my $outl;
  if(defined($outlist)){ open $outl, '>', $outlist or die "Could not open file '$outlist'"; }
  __print("\n===");
  for my $f (@_) {
	next if $f =~ /^#/;
	if(defined($dry)){print "$f\n"; next;}
    __print( " $f:\n"); 
    $txt = gettxt($f);
	$dir //= dirname($f);
	$f = basename($f);
	my $out = $o;
	$out //= "$dir/$f.bc_out";
    ($di1,$txt,$pbtab) = pages($txt)      if($p1) ; printdi($di1); puttxt("$dir/$f.ou1",$txt) if $v1;
    ($di2,$txt) = sections($txt)   if($p2) ; printdi($di2); puttxt("$dir/$f.ou2",$txt) if $v2;
    ($di3,$txt) = paragraphs($txt) if($p3) ; printdi($di3); puttxt("$dir/$f.ou3",$txt) if $v3;
    ($di4,$txt) = footnotes($txt)  if($p4) ; printdi($di4); puttxt("$dir/$f.ou4",$txt) if $v4;
    ($di5,$txt) = chars($txt)      if($p5) ; printdi($di5); puttxt("$dir/$f.ou5",$txt) if $v5;
    $txt = commit($txt)               if($commit)  ;
	print $outl "$out\n" if defined($outl);
	puttxt($out,$txt);
  }
}

sub puttxt { my ($file,$txt)=@_;
	if($pipe){ writefile($txt) }
	else     { 
		open(my $fd,'>',$file) or die "Could not open file '$file' for writing!";
		writefile($txt,$fd);
		close $fd;
	}
}

sub __print{
 if ($pipe) { print STDERR @_ }
 else       { print @_ }
}

sub printdi{ my $di=shift;
##    print "DEBUG:", Dumper($di);
    if(ref($di) eq "HASH"){
      for(keys %$di){
        if(ref($di->{$_})){ __print("$_=",Dumper($di->{$_}),";\n");  }
        else              { __print("$_=$di->{$_};\n") if $di->{$_}; }
      }
    }
    else { __print("??=$di\n")}
}

__END__

=head1 NAME

bookcleaner - prepare books for alignment and other operations

=head1 SYNOPSIS

 bookcleaner [options] file*

 bookcleaner [options] file.dbooks

=head1 DESCRIPTION

Prepare a textual book (or a list of books in a file with the
extension "dbooks", with one book path per line) for future 
align operations. The following steps are done:

=head2 Step1 -- pages, headers footers

Step1 -- pages, headers footers   (-p1=0 to skip this step)

=head2 Step2 -- sections 

Step2 -- sections                 (-p2=0 to skip this step)

=head2 Step3 -- paragraphs 

Step3 -- paragraphs               (-p3=0 to skip this step)

=head2 Step4 -- footnotes 

Step4 -- footnotes                (Deactivated by default. -p4=1 to perform this step. )

=head2 Step5 -- char level cleaning     

Step5 -- char level cleaning     (-p5=0 to skip this step)

=head2 Commit

Commit

=head1 Options

 -c  Commit at the end (removes several debug marks (_pb, etc) before creating output file

 -j=1c    Just do step 1 and commit
 -j=...p  Just ... and send output to STDOUT

 -simplify     to do several char level simplifications:
     translate some CP1252 chars to unicode
     translate several dashes, quotes and double quotes  to ascii
     defaul=1 
     use -simplify=0 to avoid simplification

 -v=34  Create temporary output files of the step 3 (file.ou3) and 4 (file-ou4)

 -minhf=3 removes headers or footers if they appear more than 3 
      times (def:5)

 -pipe   send output to STDOUT

 -latin1

 -o=FILE send output to FILE (default is original file with extension bc_out)

 -dry 	Dry run (DEBUG option, makes bookcleaner  do  nothing and just output 
 		the names of the files received	as input

 -dir=DIR create all output files under DIR/

 -nrom_off     turns off Roman Number titles

=head1 AUTHOR

Andre Santos

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

=head1 SEE ALSO

perl(1).

Text::Perfide::BookCleaner(3pm)

Ontology sections.the

=cut