Merge pull request #6 from estrabd/RT-113716-signature_as_char_string…

…-possible-not-correct-at-all

Replaced unsigned int unpack template to used signed (per puzzle.h).

.gitignore

@@ -0,0 +1,9 @@
+Makefile.old
+Makefile
+MYMETA.json
+MYMETA.yml
+Libpuzzle.bs
+Libpuzzle.c
+Libpuzzle.o
+blib/
+pm_to_blib

MANIFEST

@@ -20,3 +20,6 @@ t/pics/luxmarket_tshirt01_sheum.jpg
 typemap
 Libpuzzle.xs
 CONTRIBUTING
+bin/p5-puzzle-diff
+bin/p5-puzzle-sig
+bin/p5-puzzle-ngrams

Makefile.PL

@@ -1,13 +1,6 @@
 use 5.014004;
 use ExtUtils::MakeMaker;
 
-# Check that the local system has libpuzzle installed using
-# Devel::CheckLib.
-
-use lib 'inc';
-use Devel::CheckLib;
-check_lib_or_exit (lib => 'puzzle', header => 'puzzle.h');
-
 # See lib/ExtUtils/MakeMaker.pm for details of how to influence
 # the contents of the Makefile that is written.
 my $repo = 'https://github.com/estrabd/Image-Libpuzzle';

bin/p5-puzzle-diff

@@ -0,0 +1,52 @@
+#!/usr/bin/env perl
+
+use strict;
+use warnings;
+
+use Image::Libpuzzle ();
+
+my $p1 = Image::Libpuzzle->new;
+my $p2 = Image::Libpuzzle->new;
+
+my $sig1 = $p1->fill_cvec_from_file($ARGV[0]);
+my $sig2 = $p2->fill_cvec_from_file($ARGV[1]);
+
+print $p1->vector_normalized_distance($p2),qq{\n};
+
+__END__
+=pod
+
+=head1 NAME
+
+p5-puzzle-diff - output default distance between two images
+
+=head1 SYNOPSIS
+
+  $ p5-puzzle-diff path/to/image1 path/to/image2
+
+=head1 DESCRIPTION
+
+This script is an implementation of the puzzle-diff(8) utility provide with libpuzzle(3) using
+the bindings provided for by L<Image::Libpuzzle>.
+
+The output is the normalized distance between two images.
+
+=head1 BUGS AND LIMITATIONS
+
+Currently this script accepts no options other than the names of the image files to compare.
+
+Please report them via L<https://github.com/estrabd/p5-puzzle-xs/issues>.
+
+=head1 AUTHOR
+
+B. Estrade <estrabd@gmail.com>
+
+=head1 COPYRIGHT AND LICENSE
+
+Copyright (C) 2015 by B. Estrade
+
+This library is free software; you can redistribute it and/or modify it under
+the same terms as Perl itself, either Perl version 5.14.4 or, at your option,
+any later version of Perl 5 you may have available.
+
+=cut

bin/p5-puzzle-ngrams

@@ -0,0 +1,62 @@
+#!/usr/bin/env perl
+
+use strict;
+use warnings;
+
+use Image::Libpuzzle ();
+
+my $ngram_size = $Image::Libpuzzle::DEFAULT_NGRAM_SIZE;
+
+if ($ARGV[1] and int $ARGV[1] > 0) {
+  $ngram_size = int $ARGV[1];
+}
+if ($ARGV[1] and int $ARGV[1] <= 0) {
+  warn qq{optional argument for ngram size must be >= 1; default is $Image::Libpuzzle::DEFAULT_NGRAM_SIZE};
+  exit 1;
+}
+
+my $p1 = Image::Libpuzzle->new;
+
+my $sig1 = $p1->fill_cvec_from_file($ARGV[0]);
+
+my $ngrams_ref = $p1->signature_as_char_ngrams($ngram_size);
+
+foreach my $n (@$ngrams_ref) {
+  print qq{$n\n};
+}
+
+__END__
+=pod
+
+=head1 NAME
+
+p5-puzzle-ngrams - produce list of ngrams from an image
+
+=head1 SYNOPSIS
+
+  $ p5-puzzle-sig path/to/image [ngram_size]
+
+=head1 DESCRIPTION
+
+This script outputs a ngrams of the character representation of an image as provide for in libpuzzle(3) using
+the bindings provided for by L<Image::Libpuzzle>.
+
+=head1 BUGS AND LIMITATIONS
+
+Currently this script accepts no options other than the name of the file to sign.
+
+Please report them via L<https://github.com/estrabd/p5-puzzle-xs/issues>.
+
+=head1 AUTHOR
+
+B. Estrade <estrabd@gmail.com>
+
+=head1 COPYRIGHT AND LICENSE
+
+Copyright (C) 2015 by B. Estrade
+
+This library is free software; you can redistribute it and/or modify it under
+the same terms as Perl itself, either Perl version 5.14.4 or, at your option,
+any later version of Perl 5 you may have available.
+
+=cut

bin/p5-puzzle-sig

@@ -0,0 +1,50 @@
+#!/usr/bin/env perl
+
+use strict;
+use warnings;
+
+use Image::Libpuzzle ();
+
+my $p1 = Image::Libpuzzle->new;
+
+my $sig1 = $p1->fill_cvec_from_file($ARGV[0]);
+
+print $p1->signature_as_char_string, qq{\n};
+
+__END__
+=pod
+
+=head1 NAME
+
+p5-puzzle-sig - output character representation of image signature
+
+=head1 SYNOPSIS
+
+  $ p5-puzzle-sig path/to/image
+
+=head1 DESCRIPTION
+
+This script outputs a character representation of an image as provide for in libpuzzle(3) using
+the bindings provided for by L<Image::Libpuzzle>.
+
+The output is the character representation of the signature, in the range of [0 .. 4]. 
+
+=head1 BUGS AND LIMITATIONS
+
+Currently this script accepts no options other than the name of the file to sign.
+
+Please report them via L<https://github.com/estrabd/p5-puzzle-xs/issues>.
+
+=head1 AUTHOR
+
+B. Estrade <estrabd@gmail.com>
+
+=head1 COPYRIGHT AND LICENSE
+
+Copyright (C) 2015 by B. Estrade
+
+This library is free software; you can redistribute it and/or modify it under
+the same terms as Perl itself, either Perl version 5.14.4 or, at your option,
+any later version of Perl 5 you may have available.
+
+=cut

lib/Image/Libpuzzle.pm

@@ -17,13 +17,15 @@ our $PUZZLE_CVEC_SIMILARITY_HIGH_THRESHOLD  = Image::Libpuzzle->PUZZLE_CVEC_SIMI
 our $PUZZLE_CVEC_SIMILARITY_LOW_THRESHOLD   = Image::Libpuzzle->PUZZLE_CVEC_SIMILARITY_LOW_THRESHOLD;
 our $PUZZLE_CVEC_SIMILARITY_LOWER_THRESHOLD = Image::Libpuzzle->PUZZLE_CVEC_SIMILARITY_LOWER_THRESHOLD;
 
-# uses unpack as bin to char and $self accessor to get signature directly from the internal cvec
+# uses unpack as bin to a signed char, then add 2 to remove negative number; provides range of 0..4;
+#  and $self accessor to get signature directly from the internal cvec
 sub signature_as_char_string {
   my $self = shift;
-  my @sig  = unpack("C*", $self->get_signature());
+  # take as 
+  my @sig  = unpack("c*", $self->get_signature());
   my $sig  = q{};
   foreach my $i (@sig) {
-    $sig .= sprintf("%02d", $i);
+    $sig .= sprintf("%d", $i+2);
   }
   return $sig;
 }
@@ -384,9 +386,10 @@ considering two images as similar. Used by C<is_most_similar>.
 =head2 C<signature_as_char_string()>
 
 Returns a stringified version of the signature. The string is generated by
-unpack'ing into an array of ASCII characters (C*). Before the array of character
-codes is joined into a string, they are padded. For example, 1 turns into 001;
-25 turns into 025; 211 remains the same.
+unpack'ing into an array of ASCII of signed characters (c*). Before the array of character
+codes is joined into a string, 2 is added. This means the range of each character in the signature
+is [0..4]. The net result is that any image, no matter the size, has a stringified signature
+that is of the same length.
 
 Assumes C<fill_cvec_from_file> has been called on a valid image already.
 
@@ -408,7 +411,7 @@ Assumes C<fill_cvec_from_file> has been called on a valid image already.
 
 Returns a stringified version of the signature. The string is generated by
 unpack'ing into an array of hexidecimal digits (H*). Before the array of character
-codes is joined into a string. They are not padded like the C* based strings that
+codes is joined into a string. They are modified like the c* based strings that
 are generated with C<signature_as_char_ngrams>. As a result, the signatures are
 not as long and therefore create fewer ngrams with C<signature_as_hex_ngrams>.