farsi_processor 0.2.0 → 0.2.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
-SHA1:
-  metadata.gz: 14caf0cd116b2df5f3c3dec3266733e2356d8a3b
-  data.tar.gz: 231eb73eda6fd8d035ab607f81128a1179eaf756
+SHA256:
+  metadata.gz: dd54aed2bba8590ca05133cf27b13230d81827686e884509259ab3a97e4eae2b
+  data.tar.gz: 1f321d0f1062af86e72973025cd923509545271e980190cb53cd8bb7a377b178
 SHA512:
-  metadata.gz: 5d50d1e956b6a582c37de9a62fa122f652a0ae2b036ef7303df142d4ae4317feb8961986b7f62c0e2603e682c487b3cfe9cd5fe6a5b936d6fe92bc108198a582
-  data.tar.gz: a7dd822e3a43d8ea0c3855fed403aa7b3972185e0bfe4cd0b8e9ed7ba4c6d9b3a9849ce4cf2d3457da3a97275d9cfd02d2c6c6049b783ab6081b6f942f09c14a
+  metadata.gz: 38b5af6c8d3403a0c5b4f4ab75393cfcb2c130ac8259f639f21daf85eca2b844126c06b2e9183b05c1ee0975b5a085f9f3d1517570b55ce40eef94466bde2de8
+  data.tar.gz: 964bce4fb3443c21065b6ab7e794d830fbb7be18501ef1646e710c246803630cf49c3f8edcdc72954bdf63da18ed410ed7c099ecad1d1b8d05c52fedab6a6612

data/lib/farsi_processor.rb

@@ -1,8 +1,11 @@
 require 'farsi_processor/version'
-require_relative 'farsi_normalizer'
-require_relative 'farsi_stemmer'
+require_relative 'normalizer'
+require_relative 'stemmer'
 
 class FarsiProcessor
+  include Normalizer
+  include Stemmer
+
   def self.process(word, options = {})
     new(word, options).process
   end
@@ -15,11 +18,19 @@ class FarsiProcessor
     new(word, options).stem
   end
 
-  attr_reader :word, :options
+  attr_reader :word, :options, :excepts, :onlys
 
   def initialize(word, options = {})
     @word = word
     @options = options
+
+    @onlys = []
+    @excepts = []
+    if options[:only]
+      @onlys = options[:only]
+    elsif options[:except]
+      @excepts = options[:except]
+    end
   end
 
   def process
@@ -27,11 +38,15 @@ class FarsiProcessor
     stem
   end
 
-  def normalize
-    @word = FarsiNormalizer.process(word, options)
-  end
+  private
 
-  def stem
-    @word = FarsiStemmer.process(word, options)
+  def filter_rules(group)
+    if excepts.any?
+      group.reject { |k, _v| excepts.include?(k) }
+    elsif onlys.any?
+      group.select { |k, _v| onlys.include?(k) }
+    else
+      group
+    end
   end
 end

data/lib/farsi_processor/version.rb

@@ -1,3 +1,3 @@
 class FarsiProcessor
-  VERSION = "0.2.0"
+  VERSION = "0.2.1"
 end

data/lib/{farsi_normalizer.rb → normalizer.rb}

@@ -1,6 +1,4 @@
-require 'farsi_processor/version'
-
-class FarsiNormalizer
+module Normalizer
   ARABIC_KAF = "\u0643".freeze # ك
   FARSI_KEHEH = "\u06a9".freeze # ک
 
@@ -45,25 +43,7 @@ class FarsiNormalizer
     SUKUN
   ].freeze
 
-  def self.process(word, options = {})
-    new(word, options).process
-  end
-
-  attr_reader :word, :excepts, :onlys
-
-  def initialize(word, options = {})
-    @word = word.dup
-
-    @onlys = []
-    @excepts = []
-    if options[:only]
-      @onlys = options[:only]
-    elsif options[:except]
-      @excepts = options[:except]
-    end
-  end
-
-  def process
+  def normalize
     map_charachters
     remove_diacritics
     word
@@ -71,27 +51,17 @@ class FarsiNormalizer
 
   private
 
-  def filter_rules(group)
-    if excepts.any?
-      group.reject { |k, _v| excepts.include?(k) }
-    elsif onlys.any?
-      group.select { |k, _v| onlys.include?(k) }
-    else
-      group
-    end
-  end
-
   def map_charachters
     rules = filter_rules(CHARACTERS_MAPPINGS)
     return if rules.empty?
 
-    word.gsub!(/[#{rules.keys.join}]/, rules)
+    @word = word.gsub(/[#{rules.keys.join}]/, rules)
   end
 
   def remove_diacritics
     rules = filter_rules(DIACRITICS)
     return if rules.empty?
 
-    word.gsub!(/[#{rules.join}]/, '')
+    @word = word.gsub(/[#{rules.join}]/, '')
   end
 end

data/lib/{farsi_stemmer.rb → stemmer.rb}

@@ -1,6 +1,4 @@
-require 'farsi_processor/version'
-
-class FarsiStemmer
+module Stemmer
   ALEF = "\u0627".freeze # ا
   YEH = "\u06cc".freeze # ی
   HEH = "\u0647".freeze # ه
@@ -27,41 +25,13 @@ class FarsiStemmer
     ALEF + YEH
   ] + PLURAL_FORMS
 
-  def self.process(word, options = {})
-    new(word, options).process
-  end
-
-  attr_reader :word, :excepts, :onlys
-
-  def initialize(word, options = {})
-    @word = word.dup
-
-    @onlys = []
-    @excepts = []
-    if options[:only]
-      @onlys = options[:only]
-    elsif options[:except]
-      @excepts = options[:except]
-    end
-  end
-
-  def process
+  def stem
     stem_suffix
-    word.strip
+    @word = word.strip
   end
 
   private
 
-  def filter_rules(group)
-    if excepts.any?
-      group.reject { |k, _v| excepts.include?(k) }
-    elsif onlys.any?
-      group.select { |k, _v| onlys.include?(k) }
-    else
-      group
-    end
-  end
-
   def stem_suffix
     filter_rules(SUFFIXES).each do |suffix|
       if word.end_with?(suffix)

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: farsi_processor
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.2.1
 platform: ruby
 authors:
 - mark jad
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2017-12-27 00:00:00.000000000 Z
+date: 2018-01-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -72,10 +72,10 @@ files:
 - bin/console
 - bin/setup
 - farsi_processor.gemspec
-- lib/farsi_normalizer.rb
 - lib/farsi_processor.rb
 - lib/farsi_processor/version.rb
-- lib/farsi_stemmer.rb
+- lib/normalizer.rb
+- lib/stemmer.rb
 homepage: https://github.com/mshka/farsi_processor
 licenses:
 - MIT
@@ -96,7 +96,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.5.2
+rubygems_version: 2.7.4
 signing_key: 
 specification_version: 4
 summary: farsi_processor is a Ruby gem to process (stem and normalize) persian/farsi