phashion 1.0.0 → 1.0.1

data/CHANGES.md

@@ -0,0 +1,16 @@
+History
+=========
+
+1.0.1
+-------
+
+* Remove RubyInline, use standard Ruby extension infrastructure.
+* Update duplicate threshold constant based on wider image testing.
+* Make duplicate threshold variable so users can tune it based on their dataset.
+* Add Phashion::Image#fingerprint method which exposes an Image's 64-bit hash.
+
+
+1.0.0
+-------
+
+Initial release.

data/README.md

@@ -9,7 +9,7 @@ Installation
 
 First you need to install pHash.  pHash requires three libraries: CImg, ffmpeg and libjpeg.  My system already came with libjpeg on it so I didn't have to do anything for it.  YMMV.
 
-Install CImg.h by downloading the latest version from cimg.sf.net and placing the CImg.h header file in /usr/local/include.
+Install CImg.h by downloading the latest version from http://cimg.sf.net and placing the CImg.h header file in /usr/local/include.
 
 If you are working with audio or video, you will need to install ffmpeg:
 

data/Rakefile

@@ -10,8 +10,8 @@ begin
     gem.email = "mperham@gmail.com"
     gem.homepage = "http://github.com/mperham/phashion"
     gem.authors = ["Mike Perham"]
-    gem.add_dependency 'RubyInline'
-    gem.version = '1.0.0'
+    gem.add_development_dependency 'rake-compiler', '>= 0.7.0'
+    gem.version = '1.0.1'
     # gem is a Gem::Specification... see http://www.rubygems.org/read/chapter/20 for additional settings
   end
   Jeweler::GemcutterTasks.new
@@ -50,3 +50,9 @@ Rake::RDocTask.new do |rdoc|
   rdoc.rdoc_files.include('README*')
   rdoc.rdoc_files.include('lib/**/*.rb')
 end
+
+
+gem 'rake-compiler', '>= 0.7.0'
+require "rake/extensiontask"
+
+Rake::ExtensionTask.new("phashion_ext")

data/TODO.md

@@ -0,0 +1,6 @@
+TODO
+=======
+
+- Add video and audio APIs (someone else will have to do this, I have no need for it).
+- Add broader support for the entire pHash API, include other types of hashes.  Currently
+  the image support hardcodes use of the DCT hash.

data/ext/phashion_ext/extconf.rb

@@ -0,0 +1,13 @@
+require 'mkmf'
+
+$CFLAGS << " -x c++ #{ENV["CFLAGS"]}"
+$LIBS << " -lpHash #{ENV["LIBS"]}"
+
+# TODO: need to figure this stuff out
+# dir_config 'pHash'
+# if !have_library('pHash', 'ph_dct_imagehash')
+#   puts "Unable to find pHash library, please use 'gem install phashion -- --with-pHash-dir=/phash/install/root'"
+#   exit 1
+# end
+
+create_makefile('phashion_ext')

data/ext/phashion_ext/phashion_ext.c

@@ -0,0 +1,35 @@
+#include "ruby.h"
+#include "pHash.h"
+
+static VALUE image_hash_for(VALUE self, VALUE _filename) {
+    char * filename = StringValuePtr(_filename);
+    ulong64 hash;
+    if (-1 == ph_dct_imagehash(filename, hash)) {
+      rb_raise(rb_eRuntimeError, "Unknown pHash error");
+    }
+    return ULL2NUM(hash);
+}
+
+
+static VALUE hamming_distance(VALUE self, VALUE a, VALUE b) {
+    int result = 0;
+    result = ph_hamming_distance(NUM2ULL(a), NUM2ULL(b));
+    if (-1 == result) {
+      rb_raise(rb_eRuntimeError, "Unknown pHash error");
+    }
+    return INT2NUM(result);
+}
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+  void Init_phashion_ext() {
+    VALUE c = rb_cObject;
+    c = rb_const_get(c, rb_intern("Phashion"));
+
+    rb_define_singleton_method(c, "hamming_distance", (VALUE(*)(ANYARGS))hamming_distance, 2);
+    rb_define_singleton_method(c, "image_hash_for", (VALUE(*)(ANYARGS))image_hash_for, 1);
+  }
+#ifdef __cplusplus
+}
+#endif

data/lib/phashion.rb

@@ -1,5 +1,4 @@
 require 'rubygems'
-require 'inline'
 
 ##
 # Provides a clean and simple API to detect duplicate image files using
@@ -9,11 +8,13 @@ require 'inline'
 # int ph_dct_imagehash(const char *file, ulong64 &hash);
 # int ph_hamming_distance(ulong64 hasha, ulong64 hashb);
 
-class Phashion
-  VERSION = '1.0.0'
+module Phashion
+  VERSION = '1.0.1'
   
   class Image
-    DUPE_THRESHOLD = 26
+    SETTINGS = {
+      :dupe_threshold => 15
+    }
     
     attr_reader :filename
     def initialize(filename)
@@ -21,83 +22,14 @@ class Phashion
     end
     
     def duplicate?(other)
-      Phashion.hamming_distance(hash_code, other.send(:hash_code)) < DUPE_THRESHOLD
+      Phashion.hamming_distance(fingerprint, other.fingerprint) < SETTINGS[:dupe_threshold]
     end
-
-    private
     
-    def hash_code
+    def fingerprint
       @hash ||= Phashion.image_hash_for(@filename)
     end
   end
   
-  def self.image_hash_for(filename)
-  end
-
-  def self.hamming_distance(hashA, hashB)
-  end
-
-  inline do |builder|
-    if test ?d, "/opt/local" then
-      builder.add_compile_flags "-I/opt/local/include"
-      builder.add_link_flags "-L/opt/local/lib"
-    end
-
-    builder.add_compile_flags '-x c++', '-lstdc++'
-    builder.add_link_flags "-lpHash"
-    builder.include '"pHash.h"'
-
-    builder.c_singleton <<-"END"
-      VALUE image_hash_for(const char *filename) {
-        ulong64 hash;
-        if (-1 == ph_dct_imagehash(filename, hash)) {
-          rb_raise(rb_eRuntimeError, "Unknown pHash error");
-        }
-        return ULL2NUM(hash);
-      }
-    END
-
-    builder.c_singleton <<-"END"
-      VALUE hamming_distance(VALUE a, VALUE b) {
-        int result = 0;
-        result = ph_hamming_distance(NUM2ULL(a), NUM2ULL(b));
-        if (-1 == result) {
-          rb_raise(rb_eRuntimeError, "Unknown pHash error");
-        }
-        return INT2NUM(result);
-      }
-    END
-
-  end
 end
 
-if __FILE__ == $0
-  
-  def memory
-    `ps -o vsz,rss -p #{$$}`.strip
-  end
-  
-  def assert_duplicate(a, b)
-    raise ArgumentError, "#{a.filename} not dupe of #{b.filename}" unless a.duplicate?(b)
-  end
-
-  def assert_not_duplicate(a, b)
-    raise ArgumentError, "#{a.filename} dupe of #{b.filename}" if a.duplicate?(b)
-  end
-
-  FILES = %w(86x86-0a1e.jpeg 86x86-83d6.jpeg 86x86-a855.jpeg avatar.jpg)
-
-  images = FILES.map {|f| PHash::Image.new("#{File.dirname(__FILE__) + '/../test/'}#{f}")}
-  # GC.start
-  # puts memory
-  assert_duplicate images[0], images[1]
-  assert_duplicate images[1], images[2]
-  assert_duplicate images[0], images[2]
-
-  assert_not_duplicate images[0], images[3]
-  assert_not_duplicate images[1], images[3]
-  assert_not_duplicate images[2], images[3]
-  # GC.start
-  # puts memory
-  
-end
+require 'phashion_ext'

metadata

@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
   segments: 
   - 1
   - 0
-  - 0
-  version: 1.0.0
+  - 1
+  version: 1.0.1
 platform: ruby
 authors: 
 - Mike Perham
@@ -18,7 +18,7 @@ date: 2010-05-20 00:00:00 -05:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
-  name: RubyInline
+  name: rake-compiler
   prerelease: false
   requirement: &id001 !ruby/object:Gem::Requirement 
     requirements: 
@@ -26,24 +26,30 @@ dependencies:
       - !ruby/object:Gem::Version 
         segments: 
         - 0
-        version: "0"
-  type: :runtime
+        - 7
+        - 0
+        version: 0.7.0
+  type: :development
   version_requirements: *id001
 description: Simple wrapper around the pHash library
 email: mperham@gmail.com
 executables: []
 
-extensions: []
-
+extensions: 
+- ext/phashion_ext/extconf.rb
 extra_rdoc_files: 
 - LICENSE
 - README.md
 files: 
 - .document
 - .gitignore
+- CHANGES.md
 - LICENSE
 - README.md
 - Rakefile
+- TODO.md
+- ext/phashion_ext/extconf.rb
+- ext/phashion_ext/phashion_ext.c
 - lib/phashion.rb
 - test/86x86-0a1e.jpeg
 - test/86x86-83d6.jpeg