#!/usr/bin/perl

# Module to OCR Facebook e-mail imaging.
#
# Usage: ocr2.pl <image dir> <search domain>

use Shell qw(gocr);
use String::Approx  'amatch';
use strict;

if($ARGV[0]  "" || $ARGV[1] == "")	{
	print "Usage: ocr2.pl <images> <search domain>\n";
	die(0);
}

my $dirname = @ARGV[0];
my $file;

my @text;
my @matched;
my @hardmatched;
my $i = 0;
my $j = 0;

my $tmp;
my @tmp;

my $domain = @ARGV[1];

opendir(DIR, $dirname) or die "can't opendir $dirname: $!";

while ($file = readdir(DIR)) {
    @text[$i] = lc(gocr(' -i ' . $dirname . $file . ' -m 4 -m 2 -a 98 -p /Users/joe/Documents/SecWork/tmpdb/ -C "1234567890abcdefghijklmnopqvwxyz@._-"'));
    $i++;
}
closedir(DIR);

foreach(@text)	{
	print "Found: $_\n";

	$tmp = substr($_, rindex($_, "@") + 1);
	chomp($tmp);

	if($tmp eq $domain)	{
		@hardmatched[$j] = $_;
		$j++;
	}
}

print "Hard Matches:\n";

print @hardmatched;

$j = 0;

foreach (@text)	{
	@tmp = substr($_, rindex($_, "@") + 1);
	if(amatch($domain, @tmp))	{
		@matched[$j] = $_;
		$j++;
	}
}

print "Fuzzy Matches:\n";

print @matched;