#!/usr/bin/perl

####################################
#
# censor2txt -f html-file -o text-file
#
# Reads Censor output as html file until "Masked Sequence" and writes to text file as tab-delimited
# Authors: C. Grunau, D. Roquis, 28 mars 2011
#####################################

use diagnostics;
use strict;
use Carp;
use FileHandle;
use File::Path;
use File::Basename;
use Getopt::Std;

my $FILE       = FileHandle->new();
my $start      = "<tr align=\"CENTER\"><td>smr_";
my $end        = "<\/tbody>";
my $end2       = "<h2><tt>Masked Sequence";
my $result     = "";
my $result2    = "";
my $text       = "";
my $extraction = "";
my $header     = "Name\tFrom\tTo\tLink\tName\tFrom\tTo\tClass\tDir\tSim\tPos\tScore\n";
my $footer     = "";

use vars qw( $opt_f $opt_o);
# Command line parsing 
#
getopts('f:o:');
#
# Read input file and split into fasta sequences on the fly
#
#replace Mac and PC CRLF by UNIX LF
system ("perl -pi -e 's/\r/\n/g' $opt_f");
open($FILE,$opt_f) || croak(sprintf("Cannot open file \"%s\"",$opt_f));
while ( <$FILE> ) {
    chomp;
    $text = $text.$_;
       }
close($FILE) || croak(sprintf("Cannot close open file \"%s\"",$opt_f));


#Copy text from the beginning of the file till the "Masked Sequence" section
    while ($text =~ /$start(.*?)$end2/g) { $result = $start.$1.$end2;}

#Copy all the table rows in the prevoiously selected text
    while ($result =~ /$start(.*?)$end/g) { $result2 = $start.$1.$end; $extraction = $extraction.$result2;}

#Lines 75 to 86 removes all HTML tags in order to save the output as a .txt file
#The next two lines remove the tags around hyperlinks
$extraction =~ s/<a href="/\t/g;
$extraction =~ s/;format=EMBL">/\t/g;

#Replace the end of a row in a table by a  newline caracter
$extraction =~ s/\<\/td><\/tr><\/tbody><tr align="CENTER"><td>/\n/g;

#Remove all the remaining HTML tags and replace them by tabulations
$extraction =~ s/\<[^\<]+\>/\t/gs;

$extraction =~ s/[\t ]+/\t/gs; # Replace duplicate whitespace mid-string with 1 space
$extraction =~ s/[\t ]*$//gms; # Remove ending spaces/tabs
$extraction =~ s/^[\t ]*//gms; # Remove starting spaces/tabs

#make sure file is empty, overwrites existing file without mercy
open($FILE,">".$opt_o) || croak(sprintf("Cannot open file \"%s\"",$opt_o));
         print $FILE "";
close ($FILE);

#Add the header and the footer
$extraction = $header.$extraction.$footer;

open($FILE,">".$opt_o) || croak(sprintf("Cannot open file \"%s\"",$opt_o));
         print $FILE $extraction;
close ($FILE);