Project

General

Profile

Download (1.53 KB)

Feature #38446 » 38446-v2.patch

Go MAEDA, 2024-04-30 10:14

View differences:

lib/redmine/search.rb
135 135 
      def tokens
136 136 
        # extract tokens from the question
137 137 
        # eg. hello "bye bye" => ["hello", "bye bye"]
138 
        tokens = @question.scan(%r{(([[:space:]]|^)"[^"]+"([[:space:]]|$)|[[:^space:]]+)}).collect {|m| m.first.gsub(%r{(^[[:space:]]*"[[:space:]]*|[[:space:]]*"[[:space:]]*$)}, '')}
138 
        tokens = @question.scan(/"[^"]+"|[^\p{Zs}]+/).map do |token|
139 
          # Remove quotes from quoted tokens, strip surrounding whitespace
140 
          # e.g. "\" foo bar \"" => "foo bar"
141 
          token.gsub(/\A"\p{Zs}*|\p{Zs}*"\Z/, '')
142 
        end
139 143 
        # tokens must be at least 2 characters long
140 144 
        # but for Chinese characters (Chinese HANZI/Japanese KANJI), tokens can be one character
141 145 
        # no more than 5 tokens to search for
test/unit/lib/redmine/search_test.rb
30 30 
    value = "全角\u3000スペース"
31 31 
    assert_equal %w[全角 スペース], Redmine::Search::Tokenizer.new(value).tokens
32 32 
  end
33 

  		




  
  34
  
    
  def test_tokenize_should_support_multiple_phrases

  		




  
  35
  
    
    value = '"phrase one" "phrase two"'

  		




  
  36
  
    
    assert_equal ["phrase one", "phrase two"], Redmine::Search::Tokenizer.new(value).tokens

  		




  
  37
  
    
  end

  		




  33
  38
  
    
end

  		















  (2-2/2)