ent 0.0.2 → 0.1.0

data/History.txt

@@ -0,0 +1,6 @@
+== 0.1.0 / 2011-03-21
+  * Official initial release 
+
+== 0.0.2 / 2011-03-10
+  * Initial version published.
+

data/Rakefile

@@ -36,7 +36,7 @@ CLEAN.include("coverage/")
 
 require 'rspec/core'
 require 'rspec/core/rake_task'
-RSpec::Core::RakeTask.new(:spec) do |spec|
+RSpec::Core::RakeTask.new(:spec => :compile ) do |spec|
   spec.pattern = FileList['spec/**/*_spec.rb']
 end
 

data/VERSION

@@ -1 +1 @@
-0.0.2
+0.1.0

data/ext/ent_native/ent_native.c

@@ -1,8 +1,8 @@
 #include "ruby.h"
 
-#include "randtest.h"
-#include "iso8859.h"
+#include <stdio.h>
 
+#include "randtest.h"
 
 #ifndef DBL2NUM
 #define DBL2NUM(dbl)  rb_float_new(dbl)
@@ -78,21 +78,33 @@ VALUE rb_rt_read_string(VALUE self, VALUE rb_buf) {
     rb_raise(rb_eIOError, "data cannot be added after finalizing");
 
   rt_add(ctx, RSTRING_PTR(rb_buf), RSTRING_LEN(rb_buf));
+
   return LONG2NUM(RSTRING_LEN(rb_buf));
 }
 
 
 VALUE rb_rt_read_file(VALUE self, VALUE rb_filename) {
   rt_ctx *ctx;
+  FILE *fp;
+  int oc;
+  size_t fsz = 0;
 
   Check_Type(rb_filename, T_STRING);
-
   Data_Get_Struct(self, rt_ctx, ctx);
   if(ctx->ended)
     rb_raise(rb_eIOError, "data cannot be added after finalizing");
 
-  // todo ...
-  return Qnil;
+  if ((fp = fopen(RSTRING_PTR(rb_filename), "rb")) == NULL)
+    rb_sys_fail(0);
+
+  while((oc=fgetc(fp)) != EOF) {
+    unsigned char ocb = (unsigned char) oc;
+    rt_add(ctx, &ocb, 1);
+    fsz++;
+  }
+  fclose(fp);
+
+  return LONG2NUM(fsz);
 }
 
 
@@ -109,6 +121,7 @@ VALUE rb_rt_entropy(VALUE self) {
 
 VALUE rb_rt_mean(VALUE self) {
   rt_ctx *ctx;
+
   Data_Get_Struct(self, rt_ctx, ctx);
 
   if(ctx->ended)
@@ -128,15 +141,25 @@ VALUE rb_rt_chisquare(VALUE self) {
     return Qnil;
 }
 
-
-VALUE rb_rt_datasum(VALUE self) {
+VALUE rb_rt_chisquare_probability(VALUE self) {
   rt_ctx *ctx;
+  double ret, chip;
+
   Data_Get_Struct(self, rt_ctx, ctx);
 
-  if(ctx->ended)
-    return DBL2NUM(ctx->r_datasum);
-  else
+  if(! ctx->ended)
     return Qnil;
+
+  chip = pochisq(ctx->r_chisq, (ctx->binary ? 1 : 255));
+
+  if (chip < 0.0001)
+    ret = 0.009;
+  else if (chip > 0.9999)
+    ret = 99.991;
+  else
+    ret = chip * 100;
+
+  return DBL2NUM(ret);
 }
 
 
@@ -193,15 +216,25 @@ VALUE rb_rt_chisquare_force(VALUE self) {
   return DBL2NUM(ctx->r_chisq);
 }
 
-
-VALUE rb_rt_datasum_force(VALUE self) {
+VALUE rb_rt_chisquare_probability_force(VALUE self) {
   rt_ctx *ctx;
+  double ret, chip;
+
   Data_Get_Struct(self, rt_ctx, ctx);
 
   if(! ctx->ended)
     rt_end(ctx);
 
-  return DBL2NUM(ctx->r_datasum);
+  chip = pochisq(ctx->r_chisq, (ctx->binary ? 1 : 255));
+
+  if (chip < 0.0001)
+    ret = 0.009;
+  else if (chip > 0.9999)
+    ret = 99.991;
+  else
+    ret = chip * 100;
+
+  return DBL2NUM(ret);
 }
 
 
@@ -239,7 +272,6 @@ VALUE rb_rt_result(VALUE self) {
   rb_hash_aset(ret, ID2SYM(rb_intern("entropy")), DBL2NUM(ctx->r_ent));
   rb_hash_aset(ret, ID2SYM(rb_intern("mean")), DBL2NUM(ctx->r_mean));
   rb_hash_aset(ret, ID2SYM(rb_intern("chisquare")), DBL2NUM(ctx->r_chisq));
-  rb_hash_aset(ret, ID2SYM(rb_intern("datasum")), DBL2NUM(ctx->r_datasum));
   rb_hash_aset(ret, ID2SYM(rb_intern("montepi")), DBL2NUM(ctx->r_montepicalc));
   rb_hash_aset(ret, ID2SYM(rb_intern("scc")), DBL2NUM(ctx->r_scc));
   return ret;
@@ -256,20 +288,20 @@ void Init_ent_native() {
   rb_define_method(c_random_test, "binary?", rb_rt_get_binmode, 0);
 
   rb_define_method(c_random_test, "read", rb_rt_read_string, 1);
-//  rb_define_method(c_random_test, "read_string", rb_rt_read_string, 1);
-//  rb_define_method(c_random_test, "read_file", rb_rt_read_file, 1);
+  rb_define_method(c_random_test, "read_string", rb_rt_read_string, 1);
+  rb_define_method(c_random_test, "read_file", rb_rt_read_file, 1);
   rb_define_method(c_random_test, "finalize", rb_rt_final, 0);
   rb_define_method(c_random_test, "result", rb_rt_result, 0);
   rb_define_method(c_random_test, "entropy", rb_rt_entropy, 0);
   rb_define_method(c_random_test, "mean", rb_rt_mean, 0);
   rb_define_method(c_random_test, "chisquare", rb_rt_chisquare, 0);
-  rb_define_method(c_random_test, "datasum", rb_rt_datasum, 0);
+  rb_define_method(c_random_test, "chisquare_probability", rb_rt_chisquare_probability, 0);
   rb_define_method(c_random_test, "montepi", rb_rt_montepi, 0);
   rb_define_method(c_random_test, "scc", rb_rt_scc, 0);
   rb_define_method(c_random_test, "entropy!", rb_rt_entropy_force, 0);
   rb_define_method(c_random_test, "mean!", rb_rt_mean_force, 0);
   rb_define_method(c_random_test, "chisquare!", rb_rt_chisquare_force, 0);
-  rb_define_method(c_random_test, "datasum!", rb_rt_datasum_force, 0);
+  rb_define_method(c_random_test, "chisquare_probability!", rb_rt_chisquare_probability_force, 0);
   rb_define_method(c_random_test, "montepi!", rb_rt_montepi_force, 0);
   rb_define_method(c_random_test, "scc!", rb_rt_scc_force, 0);
  }

data/spec/ent_spec.rb

@@ -1,7 +1,5 @@
 require File.expand_path(File.dirname(__FILE__) + '/spec_helper')
 
-describe "Ent" do
-  it "fails" do
-    fail "hey buddy, you should probably rename this file and start specing for real"
-  end
+describe Ent do
+
 end

data/spec/random_test_spec.rb

@@ -0,0 +1,109 @@
+require File.expand_path(File.dirname(__FILE__) + '/spec_helper')
+
+describe Ent::RandomTest do
+  context "byte entropy" do
+    before :each do
+      @rt = Ent::RandomTest.new
+    end
+
+    it "should not have binary testing enabled" do
+      @rt.should_not be_binary
+    end
+
+    it "should initialize correctly" do
+      @rt.should be_kind_of Ent::RandomTest
+      @rt.entropy.should be_nil
+      @rt.mean.should be_nil
+      @rt.chisquare.should be_nil
+      @rt.scc.should be_nil
+    end
+
+    it "should default with no randomness" do
+      @rt.should be_kind_of Ent::RandomTest
+      @rt.entropy.should be_nil
+      @rt.mean.should be_nil
+      @rt.chisquare.should be_nil
+      @rt.chisquare_probability.should be_nil
+      @rt.scc.should be_nil
+
+      @rt.finalize
+
+      @rt.entropy.should == 0
+      @rt.chisquare_probability.should > 99.99
+      @rt.mean.should be_nan
+      @rt.chisquare.should be_nan
+    end
+
+    it "should read a string and calculate its entropy" do
+      @rt.read_string("foo").should == 3
+      @rt.finalize
+      (0.918295 .. 0.918297).should be_include(@rt.entropy)
+      (423.66 .. 423.67).should be_include(@rt.chisquare)
+      (108.0000 .. 108.0001).should be_include(@rt.mean)
+      @rt.chisquare_probability.should > 99.99
+    end
+
+    it "should read a file and calculate its entropy" do
+      @rt.read_file(sample_file("foo_test")).should == 3
+      @rt.finalize
+      (0.918295 .. 0.918297).should be_include(@rt.entropy)
+      (423.66 .. 423.67).should be_include(@rt.chisquare)
+      (108.0000 .. 108.0001).should be_include(@rt.mean)
+      @rt.chisquare_probability.should > 99.99
+    end
+
+  end
+
+  context "binary entropy" do
+    before :each do
+      @rt = Ent::RandomTest.new(true)
+    end
+
+
+    it "should have binary testing enabled" do
+      @rt.should be_binary
+    end
+
+    it "should initialize correctly" do
+      @rt.should be_kind_of Ent::RandomTest
+      @rt.entropy.should be_nil
+      @rt.mean.should be_nil
+      @rt.chisquare.should be_nil
+      @rt.scc.should be_nil
+    end
+
+    it "should default with no randomness" do
+      @rt.should be_kind_of Ent::RandomTest
+      @rt.entropy.should be_nil
+      @rt.mean.should be_nil
+      @rt.chisquare.should be_nil
+      @rt.scc.should be_nil
+
+      @rt.finalize
+
+      @rt.entropy.should == 0
+      @rt.chisquare_probability.should < 0.01
+      @rt.mean.should be_nan
+      @rt.chisquare.should be_nan
+    end
+
+    it "should read a string and calculate its entropy" do
+      @rt.read_string("foo").should == 3
+      @rt.finalize
+      (0.918295 .. 0.918297).should be_include(@rt.entropy)
+      (2.66 .. 2.67).should be_include(@rt.chisquare)
+      (0.6666 .. 0.6667).should be_include(@rt.mean)
+      @rt.chisquare_probability.should < 0.01
+    end
+
+    it "should read a file and calculate its entropy" do
+      @rt.read_file(sample_file("foo_test")).should == 3
+      @rt.finalize
+      (0.918295 .. 0.918297).should be_include(@rt.entropy)
+      (2.66 .. 2.67).should be_include(@rt.chisquare)
+      (0.6666 .. 0.6667).should be_include(@rt.mean)
+      @rt.chisquare_probability.should < 0.01
+    end
+  end
+end
+

data/spec/samples/foo_test

@@ -0,0 +1 @@
+foo

data/spec/spec_helper.rb

@@ -1,6 +1,9 @@
-$LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
-$LOAD_PATH.unshift(File.dirname(__FILE__))
+require 'pathname'
 require 'rspec'
+SPEC_DIR = Pathname.new(__FILE__).dirname
+$LOAD_PATH.unshift(SPEC_DIR.join('..', 'lib').to_s)
+$LOAD_PATH.unshift(SPEC_DIR.to_s)
+
 require 'ent'
 
 # Requires supporting files with custom matchers and macros, etc,
@@ -10,3 +13,12 @@ Dir["#{File.dirname(__FILE__)}/support/**/*.rb"].each {|f| require f}
 RSpec.configure do |config|
   
 end
+
+def sample_path(fname)
+  Pathname.new(__FILE__).dirname.join("samples", fname)
+end
+
+def sample_file(fname)
+  sample_path(fname).to_s
+end
+

metadata

@@ -4,9 +4,9 @@ version: !ruby/object:Gem::Version
   prerelease: false
   segments: 
   - 0
+  - 1
   - 0
-  - 2
-  version: 0.0.2
+  version: 0.1.0
 platform: ruby
 authors: 
 - Eric Monti
@@ -14,13 +14,13 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2011-03-10 00:00:00 -06:00
+date: 2011-03-21 00:00:00 -05:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
   name: rspec
-  prerelease: false
   requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
     requirements: 
     - - ~>
       - !ruby/object:Gem::Version 
@@ -30,11 +30,12 @@ dependencies:
         - 0
         version: 2.3.0
   type: :development
+  prerelease: false
   version_requirements: *id001
 - !ruby/object:Gem::Dependency 
   name: bundler
-  prerelease: false
   requirement: &id002 !ruby/object:Gem::Requirement 
+    none: false
     requirements: 
     - - ~>
       - !ruby/object:Gem::Version 
@@ -44,11 +45,12 @@ dependencies:
         - 0
         version: 1.0.0
   type: :development
+  prerelease: false
   version_requirements: *id002
 - !ruby/object:Gem::Dependency 
   name: jeweler
-  prerelease: false
   requirement: &id003 !ruby/object:Gem::Requirement 
+    none: false
     requirements: 
     - - ~>
       - !ruby/object:Gem::Version 
@@ -58,11 +60,12 @@ dependencies:
         - 2
         version: 1.5.2
   type: :development
+  prerelease: false
   version_requirements: *id003
 - !ruby/object:Gem::Dependency 
   name: rcov
-  prerelease: false
   requirement: &id004 !ruby/object:Gem::Requirement 
+    none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
@@ -70,11 +73,12 @@ dependencies:
         - 0
         version: "0"
   type: :development
+  prerelease: false
   version_requirements: *id004
 - !ruby/object:Gem::Dependency 
   name: rake-compiler
-  prerelease: false
   requirement: &id005 !ruby/object:Gem::Requirement 
+    none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
@@ -82,6 +86,7 @@ dependencies:
         - 0
         version: "0"
   type: :development
+  prerelease: false
   version_requirements: *id005
 description: Calculate the entropy of data
 email: esmonti@gmail.com
@@ -94,12 +99,12 @@ extra_rdoc_files:
 - README.rdoc
 - ext/ent_native/chisq.c
 - ext/ent_native/ent_native.c
-- ext/ent_native/iso8859.c
 - ext/ent_native/randtest.c
 files: 
 - .document
 - .rspec
 - Gemfile
+- History.txt
 - LICENSE.txt
 - README.rdoc
 - Rakefile
@@ -107,12 +112,12 @@ files:
 - ext/ent_native/chisq.c
 - ext/ent_native/ent_native.c
 - ext/ent_native/extconf.rb
-- ext/ent_native/iso8859.c
-- ext/ent_native/iso8859.h
 - ext/ent_native/randtest.c
 - ext/ent_native/randtest.h
 - lib/ent.rb
 - spec/ent_spec.rb
+- spec/random_test_spec.rb
+- spec/samples/foo_test
 - spec/spec_helper.rb
 has_rdoc: true
 homepage: http://github.com/emonti/ent
@@ -124,6 +129,7 @@ rdoc_options: []
 require_paths: 
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
@@ -131,6 +137,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
       - 0
       version: "0"
 required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
@@ -140,10 +147,11 @@ required_rubygems_version: !ruby/object:Gem::Requirement
 requirements: []
 
 rubyforge_project: 
-rubygems_version: 1.3.6
+rubygems_version: 1.3.7
 signing_key: 
 specification_version: 3
 summary: Calculate the entropy of data
 test_files: 
 - spec/ent_spec.rb
+- spec/random_test_spec.rb
 - spec/spec_helper.rb

data/ext/ent_native/iso8859.c

@@ -1,19 +0,0 @@
-
-/* ISO 8859/1 Latin-1 alphabetic and upper and lower case bit vector tables. */
-
-/* LINTLIBRARY */
-
-unsigned char isoalpha[32] = {
-    0,0,0,0,0,0,0,0,127,255,255,224,127,255,255,224,0,0,0,0,0,0,0,0,255,255,
-    254,255,255,255,254,255
-};
-
-unsigned char isoupper[32] = {
-    0,0,0,0,0,0,0,0,127,255,255,224,0,0,0,0,0,0,0,0,0,0,0,0,255,255,254,254,
-    0,0,0,0
-};
-
-unsigned char isolower[32] = {
-    0,0,0,0,0,0,0,0,0,0,0,0,127,255,255,224,0,0,0,0,0,0,0,0,0,0,0,1,255,255,
-    254,255
-};

data/ext/ent_native/iso8859.h

@@ -1,17 +0,0 @@
-
-/* ISO 8859/1 Latin-1 "ctype" macro replacements. */
-
-extern unsigned char isoalpha[32], isoupper[32], isolower[32];
-
-#define isISOspace(x)	((isascii(((unsigned char) (x))) && isspace(((unsigned char) (x)))) || ((x) == 0xA0))
-#define isISOalpha(x)	((isoalpha[(((unsigned char) (x))) / 8] & (0x80 >> ((((unsigned char) (x))) % 8))) != 0)
-#define isISOupper(x)	((isoupper[(((unsigned char) (x))) / 8] & (0x80 >> ((((unsigned char) (x))) % 8))) != 0)
-#define isISOlower(x)	((isolower[(((unsigned char) (x))) / 8] & (0x80 >> ((((unsigned char) (x))) % 8))) != 0)
-#define isISOprint(x)   ((((x) >= ' ') && ((x) <= '~')) || ((x) >= 0xA0))
-#define toISOupper(x)   (isISOlower(x) ? (isascii(((unsigned char) (x))) ?  \
-                            toupper(x) : (((((unsigned char) (x)) != 0xDF) && \
-                            (((unsigned char) (x)) != 0xFF)) ? \
-			    (((unsigned char) (x)) - 0x20) : (x))) : (x))
-#define toISOlower(x)   (isISOupper(x) ? (isascii(((unsigned char) (x))) ?  \
-                            tolower(x) : (((unsigned char) (x)) + 0x20)) \
-			    : (x))