nilsimsa 1.0.1 → 1.1.1

data/Gemfile

@@ -0,0 +1,14 @@
+source :rubygems
+
+group :development do
+  if RUBY_VERSION =~ /^1.9/
+    gem "ruby-debug19", :require => "ruby-debug"
+  else
+    gem 'ruby-debug'
+  end
+end
+
+group :test do
+  gem "rspec"
+  gem "rake"
+end

data/{README → README.md}

@@ -1,5 +1,7 @@
 nilsimsa
 --------
+[![Build Status](https://secure.travis-ci.org/jwilkins/nilsimsa.png)](http://travis-ci.org/jwilkins/nilsimsa)
+
 Nilsimsa is a distance based hash, which is the opposite of more familiar
 hashes like MD5.  Instead of small changes making a large difference in
 the resulting hash (to avoid collisions), distance based hashes cause

data/Rakefile

@@ -0,0 +1,19 @@
+require 'bundler'
+Bundler.require
+require 'rspec/core/rake_task'
+
+RSpec::Core::RakeTask.new(:spec)
+
+desc 'Default: run specs'
+task :default => :spec
+
+task :cleanbuild do
+  `rm ext/*.o ext/Makefile`
+  if RUBY_VERSION =~ /^1.9/
+    `cd ext && ruby extconf.rb && make`
+  elsif RUBY_VERSION =~ /^1.8/
+    `cd ext && ruby extconf.rb && make`
+  else
+    puts "Ruby version #{RUBY_VERSION}? Can't help you.. "
+  end
+end

data/bin/nilsimsa

@@ -1,5 +1,4 @@
 #!/usr/bin/env ruby
-require 'rubygems'
 require 'nilsimsa'
 
 if ARGV.size > 0 then

data/ext/extconf.rb

@@ -1,4 +1,11 @@
 #!/usr/bin/env ruby
 require 'mkmf'
+require 'rbconfig'
+
+CONFIG["DLEXT"] = "bundle"
+CONFIG["LDSHARED"] = "$(CC) -shared"
+CONFIG["CCDLFLAGS"] = " -fPIC"
+        
+have_header('ruby.h') or missing('ruby.h')
 
 create_makefile( 'nilsimsa_native' )

data/ext/nilsimsa.c

@@ -49,6 +49,15 @@
 
 #define tran3(a,b,c,n) (((tran[((a)+(n))&255]^tran[(b)]*((n)+(n)+1))+tran[(c)^tran[n]])&255)
 
+#ifdef HAVE_RUBY_IO_H
+#ifndef RSTRING_PTR
+#define RSTRING_PTR(s) (RSTRING(s))
+#endif
+#ifndef RSTRING_LEN
+#define RSTRING_LEN(s) (RSTRING(s))
+#endif
+#else
+#endif
 
 struct nsrecord {
   int acc[256];  /* counts each trigram's hash */
@@ -256,10 +265,13 @@ rbns_update(VALUE self, VALUE data) {
   char *chdata;
   long chdata_len;
   r = get_nsr( self );
+  VALUE str;
 
   Check_Type( data, T_STRING );
-  chdata = rb_str2cstr( data, &chdata_len );
-  nsr_update( r, chdata, chdata_len );
+
+  str = StringValue(data);
+  //chdata = rb_str2cstr( data, &chdata_len );
+  nsr_update( r, (RSTRING_PTR(str)), (RSTRING_LEN(str)) );
   return data;
 }
 
@@ -269,18 +281,25 @@ rbns_nilsimsa(VALUE self, VALUE other) {
   char *d1;
   char *d2;
 
-  d1 = rb_str2cstr( rb_funcall( self, rb_intern( "digest" ), 0 ), &len );
-  if (len < NSR_CODE_LEN) {
+  VALUE str1;
+  VALUE str2;
+
+  //d1 = rb_str2cstr( rb_funcall( self, rb_intern( "digest" ), 0 ), &len );
+  str1 =  rb_funcall( self, rb_intern( "digest" ), 0 );
+  str1 = StringValue(str1);
+  if(RSTRING_LEN(str1) < NSR_CODE_LEN) {
     return Qnil;
   }
 
   Check_Type( other, T_STRING );
-  d2 = rb_str2cstr( other, &len );
-  if (len < NSR_CODE_LEN) {
+  //d2 = rb_str2cstr( other, &len );
+  str2 = StringValue( other);
+  if (RSTRING_LEN(str2) < NSR_CODE_LEN) {
     return Qnil;
   }
 
-  return INT2NUM( nilsimsa( d1, d2 ) );
+  //return INT2NUM( nilsimsa( d1, d2 ) );
+  return INT2NUM( nilsimsa( RSTRING_PTR(str1), RSTRING_PTR(str2)) );
 }
 
 VALUE

data/{nilsimsa.rb → lib/nilsimsa.rb}

@@ -5,7 +5,6 @@
 # inspired by Digest::Nilsimsa-0.06 from Perl CPAN and
 # the original C nilsimsa-0.2.4 implementation by cmeclax
 # http://ixazon.dynip.com/~cmeclax/nilsimsa.html
-
 class Nilsimsa
 
   TRAN =
@@ -45,16 +44,20 @@ class Nilsimsa
   "\x04\x05\x05\x06\x05\x06\x06\x07\x05\x06\x06\x07\x06\x07\x07\x08"
 
   def initialize(*data)
-    @threshold=0; @count=0
-    @acc =Array::new(256,0)
+    @threshold=0;
+    @count=0
+    @acc = Array::new(256,0)
     @lastch0=@lastch1=@lastch2=@lastch3= -1
 
     data.each do |d| update(d) end  if data && (data.size>0)
   end
 
-  def tran3(a,b,c,n)
+  def tran3_orig(a,b,c,n)
     (((TRAN[(a+n)&255]^TRAN[b]*(n+n+1))+TRAN[(c)^TRAN[n]])&255)
   end
+  def tran3(a,b,c,n)
+    ((((TRAN[(a+n)&255].ord)^(TRAN[b].ord)*(n+n+1))+(TRAN[(c)^(TRAN[n].ord)]).ord)&255)
+  end
 
   def update(data)
     data.each_byte do |ch|
@@ -83,19 +86,25 @@ class Nilsimsa
   def digest
     @total=0;
     case @count
-      when 0..2:
-      when 3   : @total +=1
-      when 4   : @total +=4
-      else     
-        @total +=(8*@count)-28    
+      when 0..2 then ;
+      when 3   then @total +=1
+      when 4   then @total +=4
+      else @total +=(8*@count)-28
     end
     @threshold=@total/256	
 
-    @code=String::new(
-           "\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00" <<
-           "\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00")
+    @code="\x00"*32
     (0..255).each do |i|
-      @code[i>>3]+=( ((@acc[i]>@threshold)?(1):(0))<<(i&7) )
+      offset = i>>3
+      cur_val = @code[offset].ord
+      @code[offset] = (cur_val + ( ((@acc[i].ord>@threshold)?(1):(0))<<(i&7) )).chr
+#      cv = @code[i>>3].ord
+#      if @acc[i] > @threshold
+#      #@code[i>>3]+=( (((@acc[i])>@threshold)?(1):(0))<<(i&7) )
+#        @code[cv] = (@code[cv].ord + (1 <<(i&7))).chr
+#      else
+#        @code[cv] = (@code[cv].ord + (0 <<(i&7))).chr
+#      end
     end
 
     @code[0..31].reverse
@@ -126,52 +135,16 @@ class Nilsimsa
   def nilsimsa(otherdigest)
     bits=0; myd=digest
     (0..31).each do |i|
-      bits += POPC[255&myd[i]^otherdigest[i]]
+      bits += POPC[255&myd[i].ord^otherdigest[i].ord].ord
     end
     (128-bits)
   end
-  
 end
 
-def selftest
-  n1 = Nilsimsa::new;
-  n1.update("abcdefgh")
-  puts "abcdefgh:  #{n1.hexdigest=='14c8118000000000030800000004042004189020001308014088003280000078'}"
-  n2 = Nilsimsa::new("abcd","efgh")
-  puts "abcd efgh: #{n2.hexdigest=='14c8118000000000030800000004042004189020001308014088003280000078'}"
-  puts "digest:    #{n1 == n2.digest}"
-  n1.update("ijk")
-  puts "ijk:       #{n1.hexdigest=='14c811840010000c0328200108040630041890200217582d4098103280000078'}"
-  puts "nilsimsa:  #{n1.nilsimsa(n2.digest)==109}"
-  puts
+begin                               # load C core - if available
+  #require "#{File.join(File.dirname(__FILE__), '..', 'ext', 'nilsimsa_native')}"
+rescue LoadError => e
+  # ignore lack of native module
 end
 
-if __FILE__ == $0 then
-  if ARGV.size>0 then
-    begin                               # load C core - if available
-      require 'nilsimsa_native'
-    rescue LoadError => e
-      # ignore lack of native module
-    end
 
-    ARGV.each do |filename|
-      if FileTest::exists?(filename) then
-        n = Nilsimsa::new
-        n.file(filename)
-        puts n.hexdigest+" #{filename}"
-      else
-        puts "error: can't find '#{filename}'"
-      end
-    end
-  else
-    puts 'Running selftest using native ruby version'
-    selftest
-    begin                               # load C core - if available
-      require './nilsimsa_native'
-      puts 'Running selftest using compiled nilsimsa in current dir'
-      selftest
-    rescue LoadError => e
-      puts "Couldnt run selftest with compiled nilsimsa"
-    end
-  end
-end

data/nilsimsa.gemspec

@@ -0,0 +1,20 @@
+SPEC = Gem::Specification.new do |spec|
+  # Descriptive and source information for this gem.
+  spec.name = "nilsimsa"
+  spec.version = "1.1.1"
+  spec.summary = "Computes Nilsimsa values.  Nilsimsa is a distance based hash"
+  spec.author = "Jonathan Wilkins"
+  spec.email = "jwilkins[at]nospam[dot]bitland[dot]net"
+  spec.has_rdoc = true
+  spec.extra_rdoc_files = ["README.md"]
+
+  spec.files = %w(Gemfile Rakefile README.md nilsimsa.gemspec
+                  lib/nilsimsa.rb
+                  bin/nilsimsa
+                  examples/simple.rb
+                  ext/extconf.rb ext/nilsimsa.c)
+  spec.executables = ['nilsimsa']
+
+  # optional native component
+  spec.extensions = ['ext/extconf.rb']
+end

metadata

@@ -1,55 +1,56 @@
---- !ruby/object:Gem::Specification 
-rubygems_version: 0.9.4
-specification_version: 1
+--- !ruby/object:Gem::Specification
 name: nilsimsa
-version: !ruby/object:Gem::Version 
-  version: 1.0.1
-date: 2007-11-15 00:00:00 -08:00
-summary: Computes Nilsimsa values.  Nilsimsa is a distance based hash
-require_paths: 
-- .
-email: jwilkins[at]nospam[dot]bitland[dot]net
-homepage: 
-rubyforge_project: 
-description: 
-autorequire: nilsimsa.rb
-default_executable: 
-bindir: bin
-has_rdoc: true
-required_ruby_version: !ruby/object:Gem::Version::Requirement 
-  requirements: 
-  - - ">"
-    - !ruby/object:Gem::Version 
-      version: 0.0.0
-  version: 
+version: !ruby/object:Gem::Version
+  version: 1.1.1
+  prerelease: 
 platform: ruby
-signing_key: 
-cert_chain: 
-post_install_message: 
-authors: 
+authors:
 - Jonathan Wilkins
-files: 
-- README
-- examples
-- ext
-- nilsimsa.rb
-- bin
-- gemspec.rb
-- examples/simple.rb
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2012-06-21 00:00:00.000000000Z
+dependencies: []
+description: 
+email: jwilkins[at]nospam[dot]bitland[dot]net
+executables:
+- nilsimsa
+extensions:
+- ext/extconf.rb
+extra_rdoc_files:
+- README.md
+files:
+- Gemfile
+- Rakefile
+- README.md
+- nilsimsa.gemspec
+- lib/nilsimsa.rb
 - bin/nilsimsa
-- ext/nilsimsa.c
+- examples/simple.rb
 - ext/extconf.rb
-test_files: []
-
+- ext/nilsimsa.c
+homepage: 
+licenses: []
+post_install_message: 
 rdoc_options: []
-
-extra_rdoc_files: 
-- README
-executables: 
-- nilsimsa
-extensions: 
-- ext/extconf.rb
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
 requirements: []
-
-dependencies: []
-
+rubyforge_project: 
+rubygems_version: 1.8.10
+signing_key: 
+specification_version: 3
+summary: Computes Nilsimsa values.  Nilsimsa is a distance based hash
+test_files: []

data/gemspec.rb

@@ -1,33 +0,0 @@
-require 'rake'
-require 'mkmf'
-
-SPEC = Gem::Specification.new do |spec|
-  # Descriptive and source information for this gem.
-  spec.name = "nilsimsa"
-  spec.version = "1.0.1"
-  spec.summary = "Computes Nilsimsa values.  Nilsimsa is a distance based hash"
-  spec.author = "Jonathan Wilkins"
-  spec.email = "jwilkins[at]nospam[dot]bitland[dot]net"
-  spec.has_rdoc = true
-  spec.extra_rdoc_files = ["README"]
-  spec.require_path = "."
-  spec.autorequire = "nilsimsa.rb"
-
-  unfiltered_files = FileList['*', 'examples/*', 'bin/*', 'ext/*']
-  spec.files = unfiltered_files.delete_if do |filename|
-    filename.include?(".gem") || filename.include?("Makefile") ||
-    filename.include?(".so")  || filename.include?(".o")
-  end
-  spec.executables = ['nilsimsa']
-
-  # optional native component
-  if cc_command
-    spec.extensions << 'ext/extconf.rb'
-  end
-
-  puts "Building gem w/ "
-  spec.files.each do |f|
-    puts "- #{f}"
-  end
-
-end