langusta 0.1.0 → 0.1.1

data/Rakefile

@@ -16,7 +16,7 @@ Jeweler::Tasks.new do |gem|
   gem.homepage = "http://github.com/jasiek/langusta"
   gem.license = "Apache 2.0"
   gem.summary = %Q{Language detection library based on http://code.google.com/p/language-detection/.}
-  gem.description = %Q{Uses naive bayesian filter.}
+  gem.description = %Q{Highly accurate language detection library, uses naive bayesian filter.}
   gem.email = "jan.szumiec@gmail.com"
   gem.authors = ["Jan Szumiec"]
 end

data/VERSION

@@ -1 +1 @@
-0.1.0
+0.1.1

data/langusta.gemspec

@@ -5,13 +5,13 @@
 
 Gem::Specification.new do |s|
   s.name = %q{langusta}
-  s.version = "0.1.0"
+  s.version = "0.1.1"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["Jan Szumiec"]
-  s.date = %q{2011-04-08}
+  s.date = %q{2011-04-10}
   s.default_executable = %q{langusta}
-  s.description = %q{Uses naive bayesian filter.}
+  s.description = %q{Highly accurate language detection library, uses naive bayesian filter.}
   s.email = %q{jan.szumiec@gmail.com}
   s.executables = ["langusta"]
   s.extra_rdoc_files = [

data/lib/langusta.rb

@@ -12,7 +12,7 @@ require 'oniguruma'
 require 'yajl'
 
 module Langusta
-  VERSION = '0.1.0'
+  VERSION = '0.1.1'
 
   autoload :RegexHelper, 'langusta/regex_helper'
   autoload :UCS2String, 'langusta/ucs2_string'
@@ -32,5 +32,9 @@ module Langusta
   PROFILES_PATH = File.join(ABSOLUTE_PATH, 'profiles')
   UPPERCASE_BIN = File.join(ABSOLUTE_PATH, 'data/uppercase.bin')
   MESSAGES_PROPERTIES = File.join(ABSOLUTE_PATH, 'data/messages.properties')
+
+  class DuplicateProfilesError < StandardError; end
+  class NoProfilesLoadedError < StandardError; end
+  class NoFeaturesInTextError < StandardError; end
 end
 

data/lib/langusta/detector.rb

@@ -45,7 +45,7 @@ module Langusta
     def detect_block
       cleaning_text()
       ngrams = extract_ngrams()
-      raise "no features in text" if ngrams.empty?
+      raise NoFeaturesInTextError if ngrams.empty?
       @langprob = Array.new(@lang_list.length, 0.0)
 
       @n_trial.times do
@@ -68,6 +68,7 @@ module Langusta
       end
     end
 
+    # TODO: this looks like it's not referenced anywhere.
     def set_prior_map(prior_map)
       @prior_map = Array.new[@lang_list.length]
       sump = 0.0
@@ -99,7 +100,6 @@ module Langusta
       maxp
     end
 
-    private
     def cleaning_text
       non_latin_count = latin_count = 0
       @text.each_char do |c|

data/lib/langusta/detector_factory.rb

@@ -1,6 +1,4 @@
 module Langusta
-  class LangDetectException < StandardError; end
-
   class DetectorFactory
     attr_reader :word_lang_prob_map, :lang_list
 
@@ -14,7 +12,7 @@ module Langusta
     # @param [Fixnum] index at which the language profile is to be added.
     # @param [Fixnum] counts how many language profiles are to be added to this factory in total.
     def add_profile(profile, index, langsize)
-      raise LangDetectException.new("duplicate the same language profile") if @lang_list.include?(profile.name)
+      raise DuplicateProfilesError.new(profile.name) if @lang_list.include?(profile.name)
       @lang_list << profile.name
       profile.freq.keys.each do |word|
         if not @word_lang_prob_map.has_key?(word)
@@ -39,7 +37,7 @@ module Langusta
 
     private
     def create_detector
-      raise LangDetectException.new("need to load profiles") if @lang_list.length == 0
+      raise NoProfilesLoadedError if @lang_list.empty?
       detector = Detector.new(self)
     end
   end

data/test/test_detector.rb

@@ -49,4 +49,12 @@ class DetectorTest < Test::Unit::TestCase
     detector.append(UCS2String.new("\x30\x42\x30\x42\x30\x42\x30\x42\x00a"))
     assert_equal("jp", detector.detect())
   end
+
+  def test_exceptions
+    detector = @factory.create()
+    detector.append(UCS2String.new(''))
+    assert_raises(NoFeaturesInTextError) do
+      detector.detect()
+    end
+  end
 end

data/test/test_detector_factory.rb

@@ -6,11 +6,23 @@ class DetectorFactoryTest < Test::Unit::TestCase
     factory = DetectorFactory.new
 
     factory.add_profile(profile, 0, 1)
-    assert_raises(LangDetectException) do
-      factory.add_profile(profile, 1, 1)
-    end
     
     detector = factory.create(0.123)
     assert_equal(0.123, detector.alpha)
   end
+
+  def test_exceptions
+    profile = LangProfile.new
+    factory = DetectorFactory.new
+
+    assert_raises(NoProfilesLoadedError) do
+      factory.create()
+    end
+
+    factory.add_profile(profile, 0, 2)
+
+    assert_raises(DuplicateProfilesError) do
+      factory.add_profile(profile, 1, 2)
+    end
+  end
 end

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: langusta
 version: !ruby/object:Gem::Version 
-  hash: 27
+  hash: 25
   prerelease: 
   segments: 
   - 0
   - 1
-  - 0
-  version: 0.1.0
+  - 1
+  version: 0.1.1
 platform: ruby
 authors: 
 - Jan Szumiec
@@ -15,7 +15,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2011-04-08 00:00:00 +02:00
+date: 2011-04-10 00:00:00 +02:00
 default_executable: langusta
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -124,7 +124,7 @@ dependencies:
   name: ruby-debug
   version_requirements: *id007
   prerelease: false
-description: Uses naive bayesian filter.
+description: Highly accurate language detection library, uses naive bayesian filter.
 email: jan.szumiec@gmail.com
 executables: 
 - langusta