Class: Runestone::WebSearch

Inherits:
Object
  • Object
show all
Defined in:
lib/runestone/web_search.rb

Defined Under Namespace

Classes: And, Match, Or, PartialMatch, Phrase, Token

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(values) ⇒ WebSearch

Returns a new instance of WebSearch.



78
79
80
 
# File 'lib/runestone/web_search.rb', line 78

def initialize(values)
  @values = values
end

Instance Attribute Details

#values ⇒ Object

Returns the value of attribute values.



24
25
26
 
# File 'lib/runestone/web_search.rb', line 24

def values
  @values
end

Class Method Details

.parse(query, prefix: :last) ⇒ Object

prefix options: :all, :last, :none (default: :last)



27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
 
# File 'lib/runestone/web_search.rb', line 27

def self.parse(query, prefix: :last)
  prefix ||= :last
  Runestone.normalize!(query)

  q = []
  stack = []
  knot = false
  tokens = query.gsub(/\"\s+\"/, '""').split(' ')
  tokens.each_with_index do |token, i|
    token.gsub!(/\(|\)|:|\||!|\&|\*/, '')
    if token.start_with?('-')
      knot = true
      token.delete_prefix!('-')
    else
      knot = false
    end

    next if token.empty? || token == '""' || %w(' ").include?(token)
  
    if token.start_with?('"') && token.end_with?('"')
      token.delete_prefix!('"')
      token.delete_suffix!('"')
    
      q << Phrase.new([token], negative: knot)
    elsif token.start_with?('"')
      token.delete_prefix!('"')
      stack.push(:phrase)
      q << Phrase.new([Token.new(token)], negative: knot)
    elsif token.end_with?('"')
      token.delete_suffix!('"')
      q.last.values << Token.new(token)
      stack.pop
    else
      token = Token.new(token, negative: knot)
      if !knot && prefix == :last && tokens.size - 1 == i
        token.prefix = true
      elsif !knot && prefix == :all
        token.prefix = true
      end
    
      if stack.last == :phrase
        q.last.values << token
      else
        q << token
      end
    end
  end
  
  new(q)
end

Instance Method Details

#synonymize ⇒ Object 



96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
 
# File 'lib/runestone/web_search.rb', line 96

def synonymize
  parts = []
  @values.each do |token|
    if token.is_a?(Phrase) || token.negative
      parts << token
    else
      parts << [] if parts.empty? || parts.last.is_a?(Phrase) || (!parts.last.is_a?(Array) && parts.last.negative)
      parts.last << token
    end
  end

  parts.map! do |part|
    if !part.is_a?(Phrase) && (part.is_a?(Array) || !part.negative)
      synonymize_part(part)
    else
      part
    end
  end

  Runestone::WebSearch.new(parts)
end

#synonymize_part(part) ⇒ Object 



118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
 
# File 'lib/runestone/web_search.rb', line 118

def synonymize_part(part)
  pending_matches = []
  matches = []
  
  part.each_with_index do |token, i|
    
    pending_matches.select! do |match|
      if match.end_index + 1 == i && match.substitution[token.value]
        match.substitution[token.value].map do |nm|
          if nm.is_a?(Hash)
            match.end_index = i
            match.alts = nm
            true
          else
            matches << Match.new(match.start_index..i, Phrase.new(Array(nm), distance: 1))
            false
          end
        end
      else
        false
      end
    end

    if match = Runestone.synonyms[token.value]
      match.each do |m|
        if m.is_a?(Hash)
          pending_matches << PartialMatch.new(i, i, m)
        else
          matches << Match.new(i, Phrase.new(m.split(/\s+/), distance: 1))
        end
      end
    end
    
  end

  matches.select! do |match|
    if match.index.is_a?(Integer)
      case part[match.index]
      when Or
        part[match.index].values << match.substitution
      else
        part[match.index] = Or.new([part[match.index], match.substitution])
      end

      false
    else
      true
    end
  end

  groups = matches.inject([]) do |memo, match|
    if memo.empty?
      memo << [match]
    elsif i = memo.index { |k| k.none? { |j| j.index.overlaps?(match.index) } }
      memo[i] << match
    else
      memo << [match]
    end
    memo
  end

  if groups.empty?
    And.new(part)
  else
    orrs = Or.new([])
    groups.each do |g|
      p = []
      p << And.new(part[0..g.first.index.begin-1]) if g.first.index.begin > 0
      g.each do |m|
        p << Or.new([And.new(part[m.index]), m.substitution])
      end
      p << And.new(part[g.last.index.end+1..-1]) if g.last.index.end < part.size
      orrs.values << And.new(p)
    end
    orrs
  end
end

#to_s(use_synonyms: true, allow_typos: true) ⇒ Object 



196
197
198
 
# File 'lib/runestone/web_search.rb', line 196

def to_s(use_synonyms: true, allow_typos: true)
  self.values.join(' & ')
end

#typos ⇒ Object 



82
83
84
85
86
87
88
89
90
91
92
93
94
 
# File 'lib/runestone/web_search.rb', line 82

def typos
  tokens = @values.select{|t| t.is_a?(Token) && !t.negative }
  sw = Runestone::Corpus.similar_words(*tokens.map(&:value))
  q = @values.map do |t|
    if t.is_a?(Token) && sw.has_key?(t.value)
      Token.new(t.value, prefix: t.prefix, negative: t.negative, alts: sw[t.value])
    else
      t
    end
  end
  
  Runestone::WebSearch.new(q)
end