use 5.010;
use strict;
use warnings;

use Path::Tiny;

my $data = path($ARGV[0])->slurp_raw;
my $outcome = get_nr_of_cos_objects();
find_obj_blocks($outcome);


# Returns the number of COS items found in an uncompressed PDF file
# We're looking for a line that starts with "/Size", followed by a
# number that indicates the number of objects found.
# Note that we have to subtract 1 from the result found since the first
# item in the XREF table is referring to a non-obj block (root)
sub get_nr_of_cos_objects {
    my $regex = qr/(\/Size \d*)/;
    my $result = $1 if ($data =~ $regex) or -1;
    $result = $1 if $result =~ /(\d+)/;
    $result -= 1;

    # $1 is the result of the above regex execution.
    say "Number of COS: " . $result . ".";
    
    return $result;
}


# Searches for the occurences of patterns like "^x 0 obj" where x is
# a number from 1 to nr of objects passed as a parameter.
sub find_obj_blocks {
    my @counter = (1..$_[0]);
    my $result;
    
    for (@counter) {
        if ($data =~ qr/^\Q$_\E 0 obj/mp) {
            $result = $_;
            printf("Begin of Block [%5d] found at position [$-[0]]\n", $_);
            printf("End   of Block [%5d] found at position [$+[0]]\n\n", $_);
        }
    }
}