persian 0.1.0 → 0.2.1

data/lib/persian/num_text.rb

@@ -0,0 +1,53 @@
+# -*- coding: UTF-8 -*-
+
+# Persian module
+module Persian
+  # Persian NumText class
+  # Work with number in words
+  class NumText < Number
+    def self.num_to_char(num, inner = false)
+      num = Persian::Number.number(num, lang: 'en', return: 'int') if num.is_a? String
+
+      words = ''
+
+      if ((num / 100) >= 0) && ((num / 100) < 10)
+
+        if num < 10
+          if num.zero? && inner
+            words += ''
+          elsif inner
+            words += ' و '
+            words += ONES[num]
+          else
+            words += ONES[num]
+          end
+
+        elsif num < 20
+          words += TEENS[num - 10]
+        elsif num < 100
+          words += ' و ' if inner
+
+          words += DECIMAL[num / 10] + num_to_char(num % 10, true)
+        else
+          words += ' و ' if inner
+
+          words += HUNDREDS[num / 100] + num_to_char(num % 100, true)
+        end
+      else
+        words += ' و ' if inner
+
+        if (num.to_s.length % 3).nonzero?
+          current_split = num.to_s[0..(num.to_s.length % 3 - 1).to_i]
+          more_split = num.to_s.sub(current_split, '')
+          words += num_to_char(current_split) + ' ' + LONGSCALE[(num.to_s.length / 3)] + num_to_char(more_split, true)
+        else
+          current_split = num.to_s[0..2]
+          more_split = num.to_s.sub(current_split, '')
+          words += num_to_char(current_split) + ' ' + LONGSCALE[(num.to_s.length / 3) - 1] + num_to_char(more_split, true)
+        end
+      end
+
+      words
+    end
+  end
+end

data/lib/persian/number.rb

@@ -1,32 +1,81 @@
 # -*- coding: UTF-8 -*-
 
-class Persian
+# Persian module
+module Persian
+  # Persian Number class
+  # Init an instance with pass a number
+  # Basic operators works with persian string of numbers
+  class Number
+    def initialize(num)
+      @value = Persian::Number.number(num)
+    end
 
-  def self.number num, opts = {:lang => "fa", :return => "string" }
+    def +(other)
+      Number.to_persian(Number.to_i(@value) + Number.to_i(other))
+    end
 
-    if num.is_a? Numeric
-      num = num.to_s
+    def -(other)
+      Number.to_persian(Number.to_i(@value) - Number.to_i(other))
     end
 
-    if opts[:lang] == "fa" || opts[:lang] == nil
-      nums = @english_persian_numbers.merge(@arabic_persian_numbers)
-    elsif opts[:lang] == "en"
-      nums = @persian_english_numbers.merge(@arabic_english_numbers)
-    elsif opts[:lang] == "ar"
-      nums = @persian_arabic_numbers.merge(@english_arabic_numbers)
+    def *(other)
+      Number.to_persian(Number.to_i(@value) * Number.to_i(other))
     end
 
-    nums.each {|k, v|
-      num.gsub!(k, v)
-    }
+    def /(other)
+      Number.to_persian(Number.to_i(@value) / Number.to_i(other))
+    end
+
+    def self.number(num, opts = { lang: 'fa', return: 'string' })
+      opts[:lang] = 'fa' if opts[:lang].nil?
+      opts[:return] = 'string' if opts[:return].nil?
+      num = num.to_s if num.is_a? Numeric
+
+      if opts[:lang] == 'fa' || opts[:lang].nil?
+        nums = EN_FA_NUM.merge(AR_FA_NUM)
+      elsif opts[:lang] == 'en'
+        nums = FA_EN_NUM.merge(AR_EN_NUM)
+      elsif opts[:lang] == 'ar'
+        nums = FA_AR_NUM.merge(EN_AR_NUM)
+      end
 
-    if opts[:return] == "string"
-      return num
-    elsif opts[:return] == "int"
-      return num.to_i
-    else
-      return num
+      nums.each { |k, v| num.gsub!(k, v) }
+
+      if opts[:return]
+        case opts[:return]
+        when 'string'
+          num
+        when 'int'
+          num.to_i
+        end
+      else
+        num
+      end
     end
-  end
 
+    def self.number_with_colon(num)
+      num = number(num)
+      num.reverse.gsub(/(\S{3})(?=\S)/, '\\1,').reverse
+    end
+
+    def self.to_persian(num)
+      number(num, lang: 'fa')
+    end
+
+    def self.to_english(num)
+      number(num, lang: 'en')
+    end
+
+    def self.to_arabic(num)
+      number(num, lang: 'ar')
+    end
+
+    def self.to_i(num)
+      number(num, lang: 'en', return: 'int')
+    end
+
+    def self.random(params = nil)
+      number(rand(params))
+    end
+  end
 end

data/lib/persian/text/keyboard.rb

@@ -0,0 +1,22 @@
+# -*- coding: UTF-8 -*-
+
+# Persian module
+module Persian
+  # Persian Text class
+  # Digest Persian texts
+  class Text
+    # Replace english characters with it's key persian value on standard persian keyboard
+    # For now just support QWERTY keyboard
+    def self.english_to_persian_char(text)
+      EN_FA_KEYBOARD_CHAR.each { |k, v| text.gsub!(k, v) }
+      text
+    end
+
+    # Replace standard persian keyboard characters with it's key persian value on english keyboard
+    # For now just support QWERTY keyboard
+    def self.persian_to_english_char(text)
+      EN_FA_KEYBOARD_CHAR.each { |v, k| text.gsub!(k, v) }
+      text
+    end
+  end
+end

data/lib/persian/text/text.rb

@@ -0,0 +1,197 @@
+# -*- coding: UTF-8 -*-
+
+# Persian module
+module Persian
+  # Persian Text class
+  # Digest Persian texts
+  class Text
+    # Replace Arabic characters with Persian characters.
+    def self.character(text)
+      AR_FA_CHAR.each { |k, v| text.gsub!(k, v) }
+      text
+    end
+
+    # Remove extra spaces in text
+    def self.remove_extra_spaces(text)
+      text = text.split.join(' ')
+      text = text.split('‌').join('‌')
+      text
+    end
+
+    # Remove Arabic harecats from text
+    def self.remove_harekats(text)
+      HAREKATS.each { |v| text.gsub!(v, '') }
+      text
+    end
+
+    # Remove All barckets
+    def self.remove_brackets(text)
+      BRACKETS.each { |v| text.gsub!(v, '') }
+      text
+    end
+
+    # Remove Persian signs
+    def self.remove_signs(text, with = '')
+      SIGNS.each { |v| text.gsub!(v, with) }
+      text
+    end
+
+    # Replace general brackets with one type brackets
+    # Default: 	0xAB & 	0xBB
+    def self.general_brackets(text, left = '«', right = '»')
+      text.gsub!(/"(.*?)"/, left + '\1' + right)
+      text.gsub!(/\[(.*?)\]/, left + '\1' + right)
+      text.gsub!(/\{(.*?)\}/, left + '\1' + right)
+      text.gsub!(/\((.*?)\)/, left + '\1' + right)
+      text
+    end
+
+    # Add '‌ی' after names that end with ه, ا, و
+    def self.fix_y_after_vowel(text)
+      text += '‌ی' if END_VOWEL.include? text[-1]
+      text
+    end
+
+    # Replace Space with Zero-width none-joiner after می and نمی
+    def self.replace_zwnj_mi(text)
+      mi = 'می'
+      nmi = 'نمی'
+      text.gsub!(/(^|\s)(#{mi}|#{nmi})\s(\S+)/, '\1\2‌\3')
+      text
+    end
+
+    # Resplace ست with \sاست if lastest character before \s is ا
+    def self.ast(text)
+      a = 'ا'
+      ast = 'است'
+      st = 'ست'
+
+      text.gsub!(/(#{a})\s(#{ast})/, '\1' + st)
+      text
+    end
+
+    # Remove keshide from text
+    def self.keshide(text)
+      text.gsub!(/ـ+/, '')
+      text
+    end
+
+    # Use ی instead of ئ if next char is ی
+    # Example پائیز => پاییز
+    def self.replace_e_y(text)
+      e = 'ئ'
+      y = 'ی'
+      text.gsub!(/#{e}(#{y})/, '\1\1')
+      text
+    end
+
+    def self.three_dots(text)
+      text.gsub!(/\.{3,}/, '…')
+      text
+    end
+
+    def self.suffix(text)
+      tar = 'تر'
+      ee = 'ی'
+      n = 'ن'
+      ha = 'ها'
+      ye = 'ی'
+      text.gsub!(/\s+(#{tar}(#{ee}(#{n})?)?)|(#{ha}(#{ye})?)\s+/, '‌\1')
+      text
+    end
+
+    def self.remove_extra_question_mark(text)
+      mark = '؟'
+      text.gsub!(/(#{mark}){2,}/, '\1')
+      text
+    end
+
+    def self.add_zwnj(text, point)
+      text = text.scan(/^.{#{point}}|.+/).join('‌')
+      text
+    end
+
+    def self.remove_question_exclamation(text)
+      question = '؟'
+      exclamation = '!'
+      text.gsub!(/(#{question})+(#{exclamation})+/, '\1\2')
+      text
+    end
+
+    def self.remove_stopwords(text)
+      stopwords = ['و', 'در', 'به', 'این', 'با', 'از', 'که', 'است', 'را']
+      words = text.scan(/\S+/)
+      keywords = words.select { |word| !stopwords.include?(word) }
+      keywords.join(' ')
+    end
+
+    def self.remove_space_noghtevirgool(text)
+      noghtevirgool = '؛'
+      text.gsub!(/\s+(#{noghtevirgool})/, '\1')
+      text
+    end
+
+    def self.remove_signs_after_noghtevirgool(text)
+      signs = '[\.،؛:!؟\-…]'
+      noghtevirgool = '؛'
+      text.gsub!(/(#{noghtevirgool})[#{signs}]+/, '\1')
+      text
+    end
+
+    def self.space_after_noghtevirgool(text)
+      noghtevirgool = '؛'
+      text.gsub!(/(#{noghtevirgool})(\S)/, '\1 \2')
+      text
+    end
+
+    def self.remove_noghtevirgool_para_end(text)
+      noghtevirgool = '؛'
+      text.gsub!(/#{noghtevirgool}(\n|$)/, '.\1')
+      text
+    end
+
+    def self.remove_noghtevirgool_baz_start(text)
+      noghtevirgool = '؛'
+
+      regex = /([\(\[«])[ ‌]*[#{noghtevirgool}]/
+      text.gsub!(regex, '\1')
+      text
+    end
+
+    def self.remove_space_before_virgool(text)
+      virgool = '،'
+
+      text.gsub!(/\s+(#{virgool})/, '\1')
+      text
+    end
+
+    def self.remove_signs_after_virgool(text)
+      pattern = /(،)([ ‌]+)?([،؛:!؟\-][\.،؛:!؟\-]*|\.(?!\.))/
+
+      text.gsub!(pattern, '\1\2')
+      text
+    end
+
+    def self.space_after_virgool(text)
+      virgool = '،'
+
+      text.gsub!(/(#{virgool})(\S)/, '\1 \2')
+      text
+    end
+
+    def self.rm_char(text, char)
+      text.gsub!(/(#{char})/, '')
+      text
+    end
+
+    def self.rm_virgool_in_end(text)
+      text.gsub!(/(،)([ ‌\n]+)?$/, '.\2')
+      text
+    end
+
+    def self.space_after_dot(text)
+      text.gsub!(/(\.)(\S)/, '\1 \2')
+      text
+    end
+  end
+end

data/lib/persian/tokenizer.rb

@@ -0,0 +1,42 @@
+# -*- coding: UTF-8 -*-
+
+# Persian module
+module Persian
+  # Persian tokenize class
+  class Tokenizer
+    # Basic persian word tokenizer
+    # Return an array of words
+    def self.tokenize(text)
+      symbols = ['!', '﷼', ':', '؛', '؟', '،', '-', '.']
+      pair_pre = ['(', '{', '«', '<', '[']
+      pair_post = [')', '}', '»', '>', ']']
+      prepost = ["'", '"']
+
+      # Split text with space characters
+      splits = text.split(/\s/)
+
+      return [''] if splits.empty?
+
+      options = symbols + pair_pre + pair_post + prepost
+
+      pattern = /[^#{Regexp.escape(options.join)}]+/
+      tokens = []
+
+      splits.each do |split|
+        first, middle, last = split.partition(pattern)
+        tokens << first.split unless first.empty?
+        tokens << middle unless middle.empty?
+        tokens << last.split unless last.empty?
+      end
+
+      tokens.flatten
+    end
+
+    # Split paragraphs
+    # Return an array of paragraphs
+    def self.split_paragraphs(text)
+      text = text.split("\n").reject(&:empty?)
+      text
+    end
+  end
+end

data/lib/persian/unicode.rb

@@ -0,0 +1,42 @@
+# -*- coding: UTF-8 -*-
+
+# Persian module
+module Persian
+  # Persian Unicode class
+  class Unicode
+    def self.codepoint_to_char(char)
+      return [char].pack('U') if char.is_a? Fixnum
+      [char.hex].pack('U')
+    end
+
+    # Return text between RIGHT-TO-LETF EMBEDDING(U+202B) and Pop Directional Format(U+202C)
+    def self.rle(text)
+      lre_tag = 0x202B
+      pop_tag = 0x202C
+
+      codepoint_to_char(lre_tag) + text + codepoint_to_char(pop_tag)
+    end
+
+    # Return text between LETF-TO-RIGHT EMBEDDING(U+202A) and Pop Directional Format(U+202C)
+    def self.lre(text)
+      rle_tag = 0x202A
+      pop_tag = 0x202C
+
+      codepoint_to_char(rle_tag) + text + codepoint_to_char(pop_tag)
+    end
+
+    def self.rlo(text)
+      lro_tag = 0x202E
+      pop_tag = 0x202C
+
+      codepoint_to_char(lro_tag) + text + codepoint_to_char(pop_tag)
+    end
+
+    def self.lro(text)
+      rlo_tag = 0x202D
+      pop_tag = 0x202C
+
+      codepoint_to_char(rlo_tag) + text + codepoint_to_char(pop_tag)
+    end
+  end
+end