Class: Dentaku::Parser

Inherits:
Object
  • Object
show all
Defined in:
lib/dentaku/parser.rb

Constant Summary collapse

AST_OPERATIONS = 
{
  add:      AST::Addition,
  subtract: AST::Subtraction,
  multiply: AST::Multiplication,
  divide:   AST::Division,
  pow:      AST::Exponentiation,
  negate:   AST::Negation,
  mod:      AST::Modulo,
  bitor:    AST::BitwiseOr,
  bitand:   AST::BitwiseAnd,

  lt:       AST::LessThan,
  gt:       AST::GreaterThan,
  le:       AST::LessThanOrEqual,
  ge:       AST::GreaterThanOrEqual,
  ne:       AST::NotEqual,
  eq:       AST::Equal,

  and:      AST::And,
  or:       AST::Or,
}.freeze

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(tokens, options = {}) ⇒ Parser

Returns a new instance of Parser.



29
30
31
32
33
34
35
36
 
# File 'lib/dentaku/parser.rb', line 29

def initialize(tokens, options = {})
  @input             = tokens.dup
  @output            = []
  @operations        = options.fetch(:operations, [])
  @arities           = options.fetch(:arities, [])
  @function_registry = options.fetch(:function_registry, nil)
  @case_sensitive    = options.fetch(:case_sensitive, false)
end

Instance Attribute Details

#arities ⇒ Object (readonly)

Returns the value of attribute arities.



27
28
29
 
# File 'lib/dentaku/parser.rb', line 27

def arities
  @arities
end

#case_sensitive ⇒ Object (readonly)

Returns the value of attribute case_sensitive.



27
28
29
 
# File 'lib/dentaku/parser.rb', line 27

def case_sensitive
  @case_sensitive
end

#input ⇒ Object (readonly)

Returns the value of attribute input.



27
28
29
 
# File 'lib/dentaku/parser.rb', line 27

def input
  @input
end

#operations ⇒ Object (readonly)

Returns the value of attribute operations.



27
28
29
 
# File 'lib/dentaku/parser.rb', line 27

def operations
  @operations
end

#output ⇒ Object (readonly)

Returns the value of attribute output.



27
28
29
 
# File 'lib/dentaku/parser.rb', line 27

def output
  @output
end

Instance Method Details

#consume(count = 2) ⇒ Object 



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
 
# File 'lib/dentaku/parser.rb', line 38

def consume(count = 2)
  operator = operations.pop
  operator.peek(output)

  args_size = operator.arity || count
  min_size = operator.arity || operator.min_param_count || count
  max_size = operator.arity || operator.max_param_count || count

  if output.length < min_size
    fail! :too_few_operands, operator: operator, expect: min_size, actual: output.length
  end

  if output.length > max_size && operations.empty?
    fail! :too_many_operands, operator: operator, expect: max_size, actual: output.length
  end

  fail! :invalid_statement if output.size < args_size
  args = Array.new(args_size) { output.pop }.reverse

  output.push operator.new(*args)
rescue ::ArgumentError => e
  raise Dentaku::ArgumentError, e.message
rescue NodeError => e
  fail! :node_invalid, operator: operator, child: e.child, expect: e.expect, actual: e.actual
end

#function(token) ⇒ Object 



300
301
302
 
# File 'lib/dentaku/parser.rb', line 300

def function(token)
  function_registry.get(token.value)
end

#function_registry ⇒ Object 



304
305
306
 
# File 'lib/dentaku/parser.rb', line 304

def function_registry
  @function_registry ||= Dentaku::AST::FunctionRegistry.new
end

#operation(token) ⇒ Object 



296
297
298
 
# File 'lib/dentaku/parser.rb', line 296

def operation(token)
  AST_OPERATIONS.fetch(token.value)
end

#parse ⇒ Object 



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
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
 
# File 'lib/dentaku/parser.rb', line 64

def parse
  return AST::Nil.new if input.empty?

  while token = input.shift
    case token.category
    when :datetime
      output.push AST::DateTime.new(token)

    when :numeric
      output.push AST::Numeric.new(token)

    when :logical
      output.push AST::Logical.new(token)

    when :string
      output.push AST::String.new(token)

    when :identifier
      output.push AST::Identifier.new(token, case_sensitive: case_sensitive)

    when :operator, :comparator, :combinator
      op_class = operation(token)

      if op_class.right_associative?
        while operations.last && operations.last < AST::Operation && op_class.precedence < operations.last.precedence
          consume
        end

        operations.push op_class
      else
        while operations.last && operations.last < AST::Operation && op_class.precedence <= operations.last.precedence
          consume
        end

        operations.push op_class
      end

    when :null
      output.push AST::Nil.new

    when :function
      func = function(token)
      if func.nil?
        fail! :undefined_function, function_name: token.value
      end

      arities.push 0
      operations.push func

    when :case
      case_index = operations.index { |o| o == AST::Case } || -1
      token_index = case_index + 1

      case token.value
      when :open
        # special handling for case nesting: strip out inner case
        # statements and parse their AST segments recursively
        if operations.include?(AST::Case)
          open_cases = 0
          case_end_index = nil

          input.each_with_index do |input_token, index|
            if input_token.category == :case
              if input_token.value == :open
                open_cases += 1
              end

              if input_token.value == :close
                if open_cases > 0
                  open_cases -= 1
                else
                  case_end_index = index
                  break
                end
              end
            end
          end
          inner_case_inputs = input.slice!(0..case_end_index)
          subparser = Parser.new(
            inner_case_inputs,
            operations: [AST::Case],
            arities: [0],
            function_registry: @function_registry
          )
          subparser.parse
          output.concat(subparser.output)
        else
          operations.push AST::Case
          arities.push(0)
        end
      when :close
        if operations[token_index] == AST::CaseThen
          while operations.last != AST::Case
            consume
          end

          operations.push(AST::CaseConditional)
          consume(2)
          arities[-1] += 1
        elsif operations[token_index] == AST::CaseElse
          while operations.last != AST::Case
            consume
          end

          arities[-1] += 1
        end

        unless operations.count >= 1 && operations.last == AST::Case
          fail! :unprocessed_token, token_name: token.value
        end
        consume(arities.pop.succ)
      when :when
        if operations[token_index] == AST::CaseThen
          while ![AST::CaseWhen, AST::Case].include?(operations.last)
            consume
          end
          operations.push(AST::CaseConditional)
          consume(2)
          arities[-1] += 1
        elsif operations.last == AST::Case
          operations.push(AST::CaseSwitchVariable)
          consume
        end

        operations.push(AST::CaseWhen)
      when :then
        if operations[token_index] == AST::CaseWhen
          while ![AST::CaseThen, AST::Case].include?(operations.last)
            consume
          end
        end
        operations.push(AST::CaseThen)
      when :else
        if operations[token_index] == AST::CaseThen
          while operations.last != AST::Case
            consume
          end

          operations.push(AST::CaseConditional)
          consume(2)
          arities[-1] += 1
        end

        operations.push(AST::CaseElse)
      else
        fail! :unknown_case_token, token_name: token.value
      end

    when :access
      case token.value
      when :lbracket
        operations.push AST::Access
      when :rbracket
        while operations.any? && operations.last != AST::Access
          consume
        end

        unless operations.last == AST::Access
          fail! :unbalanced_bracket, token: token
        end
        consume
      end

    when :array
      case token.value
      when :array_start
        operations.push AST::Array
        arities.push 0
      when :array_end
        while operations.any? && operations.last != AST::Array
          consume
        end

        unless operations.last == AST::Array
          fail! :unbalanced_bracket, token: token
        end

        consume(arities.pop.succ)
      end

    when :grouping
      case token.value
      when :open
        if input.first && input.first.value == :close
          input.shift
          arities.pop
          consume(0)
        else
          operations.push AST::Grouping
        end

      when :close
        while operations.any? && operations.last != AST::Grouping
          consume
        end

        lparen = operations.pop
        unless lparen == AST::Grouping
          fail! :unbalanced_parenthesis, token
        end

        if operations.last && operations.last < AST::Function
          consume(arities.pop.succ)
        end

      when :comma
        fail! :invalid_statement if arities.empty?
        arities[-1] += 1
        while operations.any? && operations.last != AST::Grouping && operations.last != AST::Array
          consume
        end

      else
        fail! :unknown_grouping_token, token_name: token.value
      end

    else
      fail! :not_implemented_token_category, token_category: token.category
    end
  end

  while operations.any?
    consume
  end

  unless output.count == 1
    fail! :invalid_statement
  end

  output.first
end