Class: Flexr::Regexp::SourceLexer

Inherits:
Lexer
  • Object
show all
Defined in:
lib/flexr/regexp/tokenizer.rb,
lib/flexr/regexp/tokenizer.flexr.rb

Overview

This specification is intentionally small: it is a dogfood target for the source lexer path, not the parser's implementation.

Constant Summary

Constants included from DSL

DSL::DSL_METHODS

Instance Attribute Summary

Attributes included from DSL

#__flexr_compiled, #__flexr_config, #__flexr_generated_actions, #__flexr_rules, #__flexr_states

Attributes included from Flexr::Runtime

#buffer, #error_mode, #filename, #max_lookahead_size, #max_token_size, #on_error, #steps

Instance Method Summary collapse

Methods included from DSL

#__flexr_add_generated_eof, #__flexr_add_generated_rule, #__flexr_bind_generated_action, #__flexr_generated?, #__flexr_mark_generated!, #__flexr_reset!, #__flexr_set_compiled!, #__flexr_spec, #accel, #all_states, #backend, #compile!, #dfa, #emits, #encoding, #inherited, #on_eof, #option, #rule, #state, #token_kind

Methods included from Flexr::Runtime

#begin_state, #beginning_of_line?, #binary_input, #byte_pos, #consume_step!, #defer_token_size_check!, #each_token, #echo, #emit, #ensure_lookahead_size!, #error!, #initialize, #input, #last_location, #less, #lineno, #more, #more_text_start, #next_token, #pop, #push, #racc_next_token, #reject, #scan_steps_guarded?, #skip, #state, #text, #text_bytesize, #token_limit_required?, #tokens, #utf8_boundary?, #utf8_input?, #valid_utf8_at?

Instance Method Details

#__flexr_generated_accelerate(region, position, token_rule, token_start) ⇒ Object 



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
 
# File 'lib/flexr/regexp/tokenizer.rb', line 171

def __flexr_generated_accelerate(region, position, token_rule, token_start)
  mode = @accel_mode
  cursor = position
  matched = false
  loop do
    if cursor >= buffer.bytesize
      return matched ? cursor : nil if buffer.eof_loaded?
      return matched ? cursor : nil unless buffer.ensure_available?(cursor + 1)
    end
    segment = buffer.source
    segment_position = cursor - buffer.base_offset
    if segment.encoding == ::Encoding::UTF_8 && region.utf8_regexp
      regexp = region.utf8_regexp
    else
      segment = segment.b unless segment.encoding == ::Encoding::BINARY
      regexp = region.regexp
    end
    length = begin
      if %i[strscan auto].include?(mode) && defined?(::StringScanner)
        @__flexr_generated_accel_scanner ||= ::StringScanner.new("".b)
        @__flexr_generated_accel_scanner.string = segment
        @__flexr_generated_accel_scanner.pos = segment_position
        @__flexr_generated_accel_scanner.skip(regexp)
      else
        match = regexp.match(segment, segment_position)
        match && match.begin(0) == segment_position ? match.end(0) - segment_position : nil
      end
    rescue ArgumentError
      nil
    end
    __flexr_generated_record_auto_accel_miss(region) if mode == :auto && length.to_i < 8
    return matched ? cursor : nil unless length&.positive?
    matched = true
    cursor += length
    consume_step!(length)
    defer_token_size_check!(cursor - (more_text_start || token_start), rule: token_rule)
    return cursor if cursor < buffer.bytesize || buffer.eof_loaded?
  end
end

#__flexr_generated_acceleration_region(dfa, state) ⇒ Object 



158
159
160
161
162
163
164
165
166
167
168
169
170
 
# File 'lib/flexr/regexp/tokenizer.rb', line 158

def __flexr_generated_acceleration_region(dfa, state)
  @__flexr_generated_accel_regions ||= {}
  regions = (@__flexr_generated_accel_regions[dfa] ||= Flexr::Automaton::Accel.extract(dfa).to_h { |region| [region.state, region] })
  region = regions[state]
  return unless region
  return if @__flexr_generated_disabled_auto_accel&.key?(region)
  return if utf8_input? && region.bytes.any? { |byte| byte >= 128 }
  return if dfa.accepts[state].any? do |acceptance|
    rule = self.class.__flexr_rules.fetch(acceptance.rule_index)
    acceptance.bol_only || acceptance.end_anchor || rule.trailing
  end
  region
end

#__flexr_generated_acceptance(dfa, state, start_position, cursor, best) ⇒ Object 



234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
 
# File 'lib/flexr/regexp/tokenizer.rb', line 234

def __flexr_generated_acceptance(dfa, state, start_position, cursor, best)
  dfa.accepts[state].each do |acceptance|
    next if acceptance.bol_only && !beginning_of_line?
    next if acceptance.end_anchor && !(buffer.eof?(cursor) || buffer.getbyte(cursor) == 0x0a)
    rule = self.class.__flexr_rules.fetch(acceptance.rule_index)
    next if best && cursor < best.total_end_pos
    next if best && cursor == best.total_end_pos &&
      rule.index > best.rule.index
    best ||= (@__flexr_generated_match ||= Flexr::Runtime::Match.new)
    best.rule = rule
    best.start_pos = start_position
    best.end_pos = cursor
    best.total_end_pos = cursor
  end
  best
end

#__flexr_generated_execute(rule) ⇒ Object 



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
 
# File 'lib/flexr/regexp/tokenizer.rb', line 250

def __flexr_generated_execute(rule)
  case rule.index
  when 0
  when 1
    instance_exec(&self.class.__flexr_generated_actions.fetch(rule.index))
  when 2
    instance_exec(&self.class.__flexr_generated_actions.fetch(rule.index))
  when 3
    instance_exec(&self.class.__flexr_generated_actions.fetch(rule.index))
  when 4
    instance_exec(&self.class.__flexr_generated_actions.fetch(rule.index))
  when 5
    instance_exec(&self.class.__flexr_generated_actions.fetch(rule.index))
  when 6
    instance_exec(&self.class.__flexr_generated_actions.fetch(rule.index))
  when 7
    instance_exec(&self.class.__flexr_generated_actions.fetch(rule.index))
  when 8
    instance_exec(&self.class.__flexr_generated_actions.fetch(rule.index))
  when 9
    instance_exec(&self.class.__flexr_generated_actions.fetch(rule.index))
  when 10
    instance_exec(&self.class.__flexr_generated_actions.fetch(rule.index))
  else
    instance_exec(&rule.action)
  end
end

#__flexr_generated_fast_path?(dfa) ⇒ Boolean

Returns:

  • (Boolean) 


219
220
221
222
223
224
225
 
# File 'lib/flexr/regexp/tokenizer.rb', line 219

def __flexr_generated_fast_path?(dfa)
  @__flexr_generated_fast_paths ||= {}
  return @__flexr_generated_fast_paths[dfa] if @__flexr_generated_fast_paths.key?(dfa)
  rules = dfa.rule_ids.map { |rule_index| self.class.__flexr_rules.fetch(rule_index) }
  @__flexr_generated_fast_paths[dfa] = self.class.__flexr_config.backend != :firstmatch &&
    self.class.__flexr_config.options[:allow_empty_match] != true && rules.none?(&:trailing)
end

#__flexr_generated_record_auto_accel_miss(region) ⇒ Object 



210
211
212
213
214
215
216
217
218
 
# File 'lib/flexr/regexp/tokenizer.rb', line 210

def __flexr_generated_record_auto_accel_miss(region)
  @__flexr_generated_auto_accel_misses ||= Hash.new(0)
  @__flexr_generated_auto_accel_misses[region] += 1
  @__flexr_generated_auto_accel_miss_total = @__flexr_generated_auto_accel_miss_total.to_i + 1
  @__flexr_generated_auto_accel_off = true if @__flexr_generated_auto_accel_miss_total >= 9
  return if @__flexr_generated_auto_accel_misses[region] < 3
  @__flexr_generated_disabled_auto_accel ||= {}
  @__flexr_generated_disabled_auto_accel[region] = true
end

#__flexr_generated_scan_simple(dfa, position) ⇒ Object 



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
 
# File 'lib/flexr/regexp/tokenizer.rb', line 81

def __flexr_generated_scan_simple(dfa, position)
  current = dfa.start
  direct = dfa.direct
  direct_nxt = direct&.fetch(:nxt)
  direct_classes = direct&.fetch(:classes)
  ec = dfa.ec
  accepts = dfa.accepts
  rules = self.class.__flexr_rules
  source = buffer.stable_source
  guarded_steps = scan_steps_guarded?
  token_start = more_text_start || position
  token_limit = max_token_size
  check_token_limit = token_limit_required?
  cursor = position
  best = nil
  accel_mode = @accel_mode
  accelerate = accel_mode != :none && !(accel_mode == :auto && @__flexr_generated_auto_accel_off)
  if source && !guarded_steps && !check_token_limit && !accelerate
    while cursor < source.bytesize
      byte = source.getbyte(cursor)
      current = if direct_nxt
        class_id = ec[byte]
        value = direct_nxt[(current * direct_classes) + class_id]
        value >= 0 ? value : nil
      else
        dfa.transition(current, byte)
      end
      break unless current
      cursor += 1
      acceptance = accepts[current].first
      next unless acceptance
      rule = rules.fetch(acceptance.rule_index)
      best ||= (@__flexr_generated_match ||= Flexr::Runtime::Match.new)
      best.rule = rule
      best.start_pos = position
      best.end_pos = cursor
      best.total_end_pos = cursor
    end
    return best
  end
  while cursor < (source ? source.bytesize : buffer.bytesize) || buffer.ensure_available?(cursor + 1)
    if accelerate
      region = __flexr_generated_acceleration_region(dfa, current)
      acceptance = accepts[current].first
      accelerated_end = __flexr_generated_accelerate(region, cursor, acceptance&.rule_index, position) if region
      if accelerated_end && accelerated_end > cursor
        cursor = accelerated_end
        best = __flexr_generated_acceptance(dfa, current, position, cursor, best)
        next
      end
    end
    byte = source ? source.getbyte(cursor) : buffer.getbyte(cursor)
    consume_step! if guarded_steps
    current = if direct_nxt
      class_id = ec[byte]
      value = direct_nxt[(current * direct_classes) + class_id]
      value >= 0 ? value : nil
    else
      dfa.transition(current, byte)
    end
    break unless current
    cursor += 1
    acceptance = accepts[current].first
    if check_token_limit
      size = cursor - token_start
      defer_token_size_check!(size, rule: acceptance&.rule_index) if size > token_limit
    end
    next unless acceptance
    rule = rules.fetch(acceptance.rule_index)
    best ||= (@__flexr_generated_match ||= Flexr::Runtime::Match.new)
    best.rule = rule
    best.start_pos = position
    best.end_pos = cursor
    best.total_end_pos = cursor
  end
  best
end

#__flexr_generated_simple_fast_path?(dfa) ⇒ Boolean

Returns:

  • (Boolean) 


226
227
228
229
230
231
232
233
 
# File 'lib/flexr/regexp/tokenizer.rb', line 226

def __flexr_generated_simple_fast_path?(dfa)
  @__flexr_generated_simple_fast_paths ||= {}
  return @__flexr_generated_simple_fast_paths[dfa] if @__flexr_generated_simple_fast_paths.key?(dfa)
  rules = dfa.rule_ids.map { |rule_index| self.class.__flexr_rules.fetch(rule_index) }
  @__flexr_generated_simple_fast_paths[dfa] = rules.all? do |rule|
    rule.pattern_conditions.all? { |condition| condition && !condition.bol_only && !condition.end_anchor }
  end
end

#scan_one ⇒ Object 



21
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
 
# File 'lib/flexr/regexp/tokenizer.rb', line 21

def scan_one
  if @__flexr_generated_scan_state != @state
    machine = self.class.__flexr_compiled.machines.fetch(@state)
    @__flexr_generated_scan_state = @state
    @__flexr_generated_scan_dfa = machine.dfa
    @__flexr_generated_scan_fast = __flexr_generated_fast_path?(@__flexr_generated_scan_dfa)
    @__flexr_generated_scan_simple = __flexr_generated_simple_fast_path?(@__flexr_generated_scan_dfa)
  end
  dfa = @__flexr_generated_scan_dfa
  return Flexr::Runtime::Interpreter.new(self).scan unless @__flexr_generated_scan_fast
  position = byte_pos
  return nil unless valid_utf8_at?(position)
  return __flexr_generated_scan_simple(dfa, position) if @__flexr_generated_scan_simple
  cursor = position
  current = dfa.start
  direct = dfa.direct
  direct_nxt = direct&.fetch(:nxt)
  direct_classes = direct&.fetch(:classes)
  ec = dfa.ec
  source = buffer.stable_source
  guarded_steps = scan_steps_guarded?
  token_start = more_text_start || position
  token_limit = max_token_size
  check_token_limit = token_limit_required?
  best = nil
  accel_mode = @accel_mode
  accelerate = accel_mode != :none && !(accel_mode == :auto && @__flexr_generated_auto_accel_off)
  while cursor < (source ? source.bytesize : buffer.bytesize) || buffer.ensure_available?(cursor + 1)
    if accelerate
      region = __flexr_generated_acceleration_region(dfa, current)
      acceptance = dfa.accepts[current].first
      token_rule = acceptance&.rule_index
      accelerated_end = __flexr_generated_accelerate(region, cursor, token_rule, position) if region
      if accelerated_end && accelerated_end > cursor
        cursor = accelerated_end
        best = __flexr_generated_acceptance(dfa, current, position, cursor, best)
        next
      end
    end
    byte = source ? source.getbyte(cursor) : buffer.getbyte(cursor)
    consume_step! if guarded_steps
    current = if direct_nxt
      class_id = ec[byte]
      value = direct_nxt[(current * direct_classes) + class_id]
      value >= 0 ? value : nil
    else
      dfa.transition(current, byte)
    end
    break unless current
    cursor += 1
    resource_acceptance = dfa.accepts[current].first
    resource_rule = resource_acceptance&.rule_index
    if check_token_limit
      size = cursor - token_start
      defer_token_size_check!(size, rule: resource_rule) if size > token_limit
    end
    best = __flexr_generated_acceptance(dfa, current, position, cursor, best)
  end
  best
end