Module: Factbase::IndexedTerm

Defined in:
lib/factbase/indexed/indexed_term.rb

Overview

Term with an index.

Author

Yegor Bugayenko ([email protected])

Copyright

Copyright © 2024-2025 Yegor Bugayenko

License

MIT

Instance Method Summary collapse

Instance Method Details

#predict(maps, params) ⇒ Array<Hash>|nil

Reduces the provided list of facts (maps) to a smaller array, if it’s possible.

NIL must be returned if indexing is prohibited in this case.

Parameters:

  • maps (Array<Hash>) —

    Array of facts

  • params (Hash) —

    Key/value params to use

Returns:

  • (Array<Hash>|nil) —

    Returns a new array, or NIL if the original array must be used



22
23
24
25
26
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
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
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
 
# File 'lib/factbase/indexed/indexed_term.rb', line 22

def predict(maps, params)
  key = [maps.object_id, @operands.first, @op]
  case @op
  when :one
    if @idx[key].nil?
      @idx[key] = []
      prop = @operands.first.to_s
      maps.to_a.each do |m|
        @idx[key].append(m) if !m[prop].nil? && m[prop].size == 1
      end
    end
    (maps & []) | @idx[key]
  when :exists
    if @idx[key].nil?
      @idx[key] = []
      prop = @operands.first.to_s
      maps.to_a.each do |m|
        @idx[key].append(m) unless m[prop].nil?
      end
    end
    (maps & []) | @idx[key]
  when :absent
    if @idx[key].nil?
      @idx[key] = []
      prop = @operands.first.to_s
      maps.to_a.each do |m|
        @idx[key].append(m) if m[prop].nil?
      end
    end
    (maps & []) | @idx[key]
  when :eq
    if @operands.first.is_a?(Symbol) && _scalar?(@operands[1])
      if @idx[key].nil?
        @idx[key] = {}
        prop = @operands.first.to_s
        maps.to_a.each do |m|
          m[prop]&.each do |v|
            @idx[key][v] = [] if @idx[key][v].nil?
            @idx[key][v].append(m)
          end
        end
      end
      vv =
        if @operands[1].is_a?(Symbol)
          params[@operands[1].to_s] || []
        else
          [@operands[1]]
        end
      if vv.empty?
        (maps & [])
      else
        j = vv.map { |v| @idx[key][v] || [] }.reduce(&:|)
        (maps & []) | j
      end
    end
  when :gt
    if @operands.first.is_a?(Symbol) && _scalar?(@operands[1])
      prop = @operands.first.to_s
      cache_key = [maps.object_id, @operands.first, :sorted]
      if @idx[cache_key].nil?
        @idx[cache_key] = []
        maps.to_a.each do |m|
          values = m[prop]
          next if values.nil?
          values.each do |v|
            @idx[cache_key] << [v, m]
          end
        end
        @idx[cache_key].sort_by! { |pair| pair[0] }
      end
      threshold = @operands[1].is_a?(Symbol) ? params[@operands[1].to_s]&.first : @operands[1]
      return nil if threshold.nil?
      i = @idx[cache_key].bsearch_index { |pair| pair[0] > threshold } || @idx[cache_key].size
      result = @idx[cache_key][i..].map { |pair| pair[1] }.uniq
      (maps & []) | result
    end
  when :lt
    if @operands.first.is_a?(Symbol) && _scalar?(@operands[1])
      prop = @operands.first.to_s
      cache_key = [maps.object_id, @operands.first, :sorted]
      if @idx[cache_key].nil?
        @idx[cache_key] = []
        maps.to_a.each do |m|
          values = m[prop]
          next if values.nil?
          values.each do |v|
            @idx[cache_key] << [v, m]
          end
        end
        @idx[cache_key].sort_by! { |pair| pair[0] }
      end
      threshold = @operands[1].is_a?(Symbol) ? params[@operands[1].to_s]&.first : @operands[1]
      return nil if threshold.nil?
      i = @idx[cache_key].bsearch_index { |pair| pair[0] >= threshold } || @idx[cache_key].size
      result = @idx[cache_key][0...i].map { |pair| pair[1] }.uniq
      (maps & []) | result
    end
  when :and
    r = nil
    if @operands.all? { |o| o.op == :eq } && @operands.size > 1 \
      && @operands.all? { |o| o.operands.first.is_a?(Symbol) && _scalar?(o.operands[1]) }
      props = @operands.map { |o| o.operands.first }.sort
      key = [maps.object_id, props, :multi_and_eq]
      if @idx[key].nil?
        @idx[key] = {}
        maps.to_a.each do |m|
          _all_tuples(m, props).each do |t|
            @idx[key][t] = [] if @idx[key][t].nil?
            @idx[key][t].append(m)
          end
        end
      end
      tuples = Enumerator.product(
        *@operands.sort_by { |o| o.operands.first }.map do |o|
          if o.operands[1].is_a?(Symbol)
            params[o.operands[1].to_s] || []
          else
            [o.operands[1]]
          end
        end
      )
      j = tuples.map { |t| @idx[key][t] || [] }.reduce(&:|)
      r = (maps & []) | j
    else
      @operands.each do |o|
        n = o.predict(maps, params)
        break if n.nil?
        if r.nil?
          r = n
        elsif n.size < r.size * 8 # to skip some obvious matchings
          r &= n.to_a
        end
        break if r.size < maps.size / 32 # it's already small enough
        break if r.size < 128 # it's obviously already small enough
      end
    end
    r
  when :or
    r = nil
    @operands.each do |o|
      n = o.predict(maps, params)
      if n.nil?
        r = nil
        break
      end
      r = maps & [] if r.nil?
      r |= n.to_a
      return maps if r.size > maps.size / 4 # it's big enough already
    end
    r
  when :not
    if @idx[key].nil?
      yes = @operands.first.predict(maps, params)
      if yes.nil?
        @idx[key] = { r: nil }
      else
        yes = yes.to_a.to_set
        @idx[key] = { r: maps.to_a.reject { |m| yes.include?(m) } }
      end
    end
    r = @idx[key][:r]
    if r.nil?
      nil
    else
      (maps & []) | r
    end
  end
end