Class: Bm25::Validator

Inherits:
Object
  • Object
show all
Defined in:
lib/bm25/validator.rb

Class Method Summary collapse

Class Method Details

.is_onechar?(word) ⇒ Boolean

Return True if the word is one character

Parameters:

  • word (String)

Returns:

  • (Boolean) —

    return True if the word is one character



24
25
26
 
# File 'lib/bm25/validator.rb', line 24

def is_onechar?(word)
  return word.size == 1
end

.is_stopword?(word) ⇒ Boolean

Return True if the word is a stopword

Parameters:

  • word (String)

Returns:

  • (Boolean) —

    return True if the word is a stopword



8
9
10
11
12
13
14
15
16
17
18
19
20
 
# File 'lib/bm25/validator.rb', line 8

def is_stopword? (word)
  match = false
  stopword_path = File.join( File.dirname(__FILE__), 'stopword.txt' )
  File.open(stopword_path, "r") do |f|
    f.each_line do |t|
      if t.chomp === word
        match = true
        break
      end
    end
  end
  return match
end

.validate_word(word_obj) ⇒ Boolean

Return True if the mecab obj is noun

Parameters:

  • mecab (Object) —

    object

Returns:

  • (Boolean) —

    return True if the mecab obj is noun



30
31
32
33
34
35
36
37
38
39
40
41
42
 
# File 'lib/bm25/validator.rb', line 30

def validate_word(word_obj)
  n = word_obj
  if (n.is_bos? || n.is_eos?) ||
      n.feature.scan(/名詞/).length === 0 ||
      n.surface.match(/[\/\d]/) ||
      self.is_stopword?(n.surface) ||
      self.is_onechar?(n.surface)
      return true
  end
  
  return false

end