Class: Rababa::Encoders::ArabicEncoderWithStartSymbol

Inherits:
TextEncoder
  • Object
show all
Defined in:
lib/rababa/encoders.rb

Instance Attribute Summary

Attributes inherited from TextEncoder

#input_id_to_symbol, #input_pad_id, #start_symbol_id, #target_id_to_symbol, #utarget_id_to_symbol

Instance Method Summary collapse

Methods inherited from TextEncoder

#clean, #input_to_sequence

Methods included from Harakats

#basic_cleaners, #collapse_whitespace, #extract_haraqat, #extract_stack, #remove_diacritics, #valid_arabic_cleaners

Constructor Details

#initialize(cleaner = "basic_cleaners", reverse_input: bool = false, reverse_target: bool = false) ⇒ ArabicEncoderWithStartSymbol

Returns a new instance of ArabicEncoderWithStartSymbol.



89
90
91
92
93
94
95
96
97
98
99
100
101
# File 'lib/rababa/encoders.rb', line 89

def initialize(cleaner="basic_cleaners",
               reverse_input: bool = false,
               reverse_target: bool = false)

    input_chars = "بض.غىهظخة؟:طس،؛فندؤلوئآك-يذاصشحزءمأجإ ترقعث".chars
    target_charts = Rababa::ArabicConstants::ALL_POSSIBLE_HARAQAT.keys

    super(input_chars, target_charts,
          cleaner=cleaner,
          reverse_input=reverse_input)

    @start_symbol_id = @target_symbol_to_id["s"]
end