-
Notifications
You must be signed in to change notification settings - Fork 35
/
parser.rb
1685 lines (1389 loc) · 49.3 KB
/
parser.rb
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
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
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
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
# Author: Chris Wailes <chris.wailes@gmail.com>
# Project: Ruby Language Toolkit
# Date: 2011/01/19
# Description: This file contains the base class for parsers that use RLTK.
############
# Requires #
############
# Ruby Language Toolkit
require 'rltk/cfg'
#######################
# Classes and Modules #
#######################
# The RLTK root module
module RLTK
# A BadToken error indicates that a token was observed in the input stream
# that wasn't used in the grammar's definition.
class BadToken < StandardError
# @return [String] String representation of the error.
def to_s
'Unexpected token. Token not present in grammar definition.'
end
end
# A NotInLanguage error is raised whenever there is no valid parse tree
# for a given token stream. In other words, the input string is not in the
# defined language.
class NotInLanguage < StandardError
# @return [Array<Token>] List of tokens that have been successfully parsed
attr_reader :seen
# @return [Token] Token that caused the parser to stop
attr_reader :current
# @return [Array<Token>] List of tokens that have yet to be seen
attr_reader :remaining
# @param [Array<Token>] seen Tokens that have been successfully parsed
# @param [Token] current Token that caused the parser to stop
# @param [Array<Token>] remaining Tokens that have yet to be seen
def initialize(seen, current, remaining)
@seen = seen
@current = current
@remaining = remaining
end
# @return [String] String representation of the error.
def to_s
"String not in language. Token info:\n\tSeen: #{@seen}\n\tCurrent: #{@current}\n\tRemaining: #{@remaining}"
end
end
# An error of this type is raised when the parser encountered a error that
# was handled by an error production.
class HandledError < StandardError
# The errors as reported by the parser.
#
# @return [Array<Object>]
attr_reader :errors
# The result that would have been returned by the call to *parse*.
attr_reader :result
# Instantiate a new HandledError object with *errors*.
#
# @param [Array<Object>] errors Errors added to the parsing environment by calls to {Parser::Environment#error}.
# @param [Object] result Object resulting from parsing Tokens before the error occurred.
def initialize(errors, result)
@errors = errors
@result = result
end
end
# Used for exceptions that occure during parser construction.
class ParserConstructionException < Exception; end
# Used for runtime exceptions that are the parsers fault. These should
# never be observed in the wild.
class InternalParserException < Exception; end
# Used to indicate that a parser is empty or hasn't been finalized.
class UselessParserException < Exception
# Sets the error messsage for this exception.
def initialize
super('Parser has not been finalized.')
end
end
# The Parser class may be sub-classed to produce new parsers. These
# parsers have a lot of features, and are described in the main
# documentation.
class Parser
# @return [Environment] Environment used by the instantiated parser.
attr_reader :env
#################
# Class Methods #
#################
class << self
# The overridden new prevents un-finalized parsers from being
# instantiated.
def new(*args)
if not @symbols
raise UselessParserException
else
super(*args)
end
end
# Installs instance class varialbes into a class.
#
# @return [void]
def install_icvars
@curr_lhs = nil
@curr_prec = nil
@conflicts = Hash.new {|h, k| h[k] = Array.new}
@grammar = CFG.new
@grammar_prime = nil
@lh_sides = Hash.new
@procs = Array.new
@states = Array.new
@symbols = nil
# Variables for dealing with precedence.
@prec_counts = {:left => 0, :right => 0, :non => 0}
@production_precs = Array.new
@token_precs = Hash.new
@token_hooks = Hash.new {|h, k| h[k] = []}
# Set the default argument handling policy. Valid values
# are :array and :splat.
@default_arg_type = :splat
@grammar.callback do |type, which, p, sels = []|
@procs[p.id] = [
case type
when :optional
case which
when :empty then ProdProc.new { || nil }
else ProdProc.new { |o| o }
end
when :elp
case which
when :empty then ProdProc.new { || [] }
else ProdProc.new { |prime| prime }
end
when :nelp
case which
when :single
ProdProc.new { |el| [el] }
when :multiple
ProdProc.new(:splat, sels) do |*syms|
el = syms[1..-1]
syms.first << (el.length == 1 ? el.first : el)
end
else
ProdProc.new { |*el| el.length == 1 ? el.first : el }
end
end,
p.rhs.length
]
@production_precs[p.id] = p.last_terminal
end
end
# Called when the Lexer class is sub-classed, it installes
# necessary instance class variables.
#
# @return [void]
def inherited(klass)
klass.install_icvars
end
# If *state* (or its equivalent) is not in the state list it is
# added and it's ID is returned. If there is already a state
# with the same items as *state* in the state list its ID is
# returned and *state* is discarded.
#
# @param [State] state State to add to the parser.
#
# @return [Integer] The ID of the state.
def add_state(state)
if (id = @states.index(state))
id
else
state.id = @states.length
@states << state
@states.length - 1
end
end
# Build a hash with the default options for Parser.finalize
# and then update it with the values from *opts*.
#
# @param [Hash{Symbol => Object}] opts Hash containing options for finalize.
#
# @return [Hash{Symbol => Object}]
def build_finalize_opts(opts)
opts[:explain] = self.get_io(opts[:explain])
{
explain: false,
lookahead: true,
precedence: true,
use: false
}.update(opts)
end
private :build_finalize_opts
# Build a hash with the default options for Parser.parse and
# then update it with the values from *opts*.
#
# @param [Hash{Symbol => Object}] opts Hash containing options for parse.
#
# @return [Hash{Symbol => Object}]
def build_parse_opts(opts)
opts[:parse_tree] = self.get_io(opts[:parse_tree])
opts[:verbose] = self.get_io(opts[:verbose])
{
accept: :first,
env: self::Environment.new,
parse_tree: false,
verbose: false
}.update(opts)
end
private :build_parse_opts
# This method is used to (surprise) check the sanity of the
# constructed parser. It checks to make sure all non-terminals
# used in the grammar definition appear on the left-hand side of
# one or more productions, and that none of the parser's states
# have invalid actions. If a problem is encountered a
# ParserConstructionException is raised.
#
# @return [void]
def check_sanity
# Check to make sure all non-terminals appear on the
# left-hand side of some production.
@grammar.nonterms.each do |sym|
if not @lh_sides.values.include?(sym)
raise ParserConstructionException, "Non-terminal #{sym} does not appear on the left-hand side of any production."
end
end
# Check the actions in each state.
each_state do |state|
state.actions.each do |sym, actions|
if CFG::is_terminal?(sym)
# Here we check actions for terminals.
actions.each do |action|
if action.is_a?(Accept)
if sym != :EOS
raise ParserConstructionException, "Accept action found for terminal #{sym} in state #{state.id}."
end
elsif not (action.is_a?(GoTo) or action.is_a?(Reduce) or action.is_a?(Shift))
raise ParserConstructionException, "Object of type #{action.class} found in actions for terminal " +
"#{sym} in state #{state.id}."
end
end
if (conflict = state.conflict_on?(sym))
self.inform_conflict(state.id, conflict, sym)
end
else
# Here we check actions for non-terminals.
if actions.length > 1
raise ParserConstructionException, "State #{state.id} has multiple GoTo actions for non-terminal #{sym}."
elsif actions.length == 1 and not actions.first.is_a?(GoTo)
raise ParserConstructionException, "State #{state.id} has non-GoTo action for non-terminal #{sym}."
end
end
end
end
end
# This method checks to see if the parser would be in parse state
# *dest* after starting in state *start* and reading *symbols*.
#
# @param [Symbol] start Symbol representing a CFG production.
# @param [Symbol] dest Symbol representing a CFG production.
# @param [Array<Symbol>] symbols Grammar symbols.
#
# @return [Boolean] If the destination symbol is reachable from the start symbol after reading *symbols*.
def check_reachability(start, dest, symbols)
path_exists = true
cur_state = start
symbols.each do |sym|
actions = @states[cur_state.id].on?(sym)
actions = actions.select { |a| a.is_a?(Shift) } if CFG::is_terminal?(sym)
if actions.empty?
path_exists = false
break
end
# There can only be one Shift action for terminals and
# one GoTo action for non-terminals, so we know the
# first action is the only one in the list.
cur_state = @states[actions.first.id]
end
path_exists and cur_state.id == dest.id
end
# Declares a new clause inside of a production. The right-hand
# side is specified by *expression* and the precedence of this
# production can be changed by setting the *precedence* argument
# to some terminal symbol.
#
# @param [String, Symbol] expression Right-hand side of a production.
# @param [Symbol] precedence Symbol representing the precedence of this production.
# @param [:array, :splat] arg_type Method to use when passing arguments to the action.
# @param [Proc] action Action to be taken when the production is reduced.
#
# @return [void]
def clause(expression, precedence = nil, arg_type = @default_arg_type, &action)
# Use the curr_prec only if it isn't overridden for this
# clause.
precedence ||= @curr_prec
production, selections = @grammar.clause(expression)
# Check to make sure the action's arity matches the number
# of symbols on the right-hand side.
expected_arity = (selections.empty? ? production.rhs.length : selections.length)
if arg_type == :splat and action.arity != expected_arity
raise ParserConstructionException,
"Incorrect number of action parameters. Expected #{expected_arity} but got #{action.arity}." +
' Action arity must match the number of terminals and non-terminals in the clause.'
end
# Add the action to our proc list.
@procs[production.id] = [ProdProc.new(arg_type, selections, &action), production.rhs.length]
# If no precedence is specified use the precedence of the
# last terminal in the production.
@production_precs[production.id] = precedence || production.last_terminal
end
alias :c :clause
# Removes resources that were needed to generate the parser but
# aren't needed when actually parsing input.
#
# @return [void]
def clean
# We've told the developer about conflicts by now.
@conflicts = nil
# Drop the grammar and the grammar'.
@grammar = nil
@grammar_prime = nil
# Drop precedence and bookkeeping information.
@cur_lhs = nil
@cur_prec = nil
@prec_counts = nil
@production_precs = nil
@token_precs = nil
# Drop the items from each of the states.
each_state { |state| state.clean }
end
# Set the default argument type for the actions associated with
# clauses. All actions defined after this call will be passed
# arguments in the way specified here, unless overridden in the
# call to {Parser.clause}.
#
# @param [:array, :splat] type The default argument type.
#
# @return [void]
def default_arg_type(type)
@default_arg_type = type if type == :array or type == :splat
end
alias :dat :default_arg_type
# Adds productions and actions for parsing empty lists.
#
# @see CFG#empty_list_production
def build_list_production(symbol, list_elements, separator = '')
@grammar.build_list_production(symbol, list_elements, separator)
end
alias :list :build_list_production
# This function will print a description of the parser to the
# provided IO object.
#
# @param [IO] io Input/Output object used for printing the parser's explanation.
#
# @return [void]
def explain(io)
if @grammar and not @states.empty?
io.puts('###############')
io.puts('# Productions #')
io.puts('###############')
io.puts
max_id_length = @grammar.productions(:id).length.to_s.length
# Print the productions.
@grammar.productions.each do |sym, productions|
max_rhs_length = productions.inject(0) { |m, p| if (len = p.to_s.length) > m then len else m end }
productions.each do |production|
p_string = production.to_s
io.print("\tProduction #{sprintf("%#{max_id_length}d", production.id)}: #{p_string}")
if (prec = @production_precs[production.id])
io.print(' ' * (max_rhs_length - p_string.length))
io.print(" : (#{sprintf("%-5s", prec.first)}, #{prec.last})")
end
io.puts
end
io.puts
end
io.puts('##########')
io.puts('# Tokens #')
io.puts('##########')
io.puts
max_token_len = @grammar.terms.inject(0) { |m, t| if t.length > m then t.length else m end }
@grammar.terms.sort {|a,b| a.to_s <=> b.to_s }.each do |term|
io.print("\t#{term}")
if (prec = @token_precs[term])
io.print(' ' * (max_token_len - term.length))
io.print(" : (#{sprintf("%-5s", prec.first)}, #{prec.last})")
end
io.puts
end
io.puts
io.puts('#####################')
io.puts('# Table Information #')
io.puts('#####################')
io.puts
io.puts("\tStart symbol: #{@grammar.start_symbol}'")
io.puts
io.puts("\tTotal number of states: #{@states.length}")
io.puts
io.puts("\tTotal conflicts: #{@conflicts.values.flatten(1).length}")
io.puts
@conflicts.each do |state_id, conflicts|
io.puts("\tState #{state_id} has #{conflicts.length} conflict(s)")
end
io.puts if not @conflicts.empty?
# Print the parse table.
io.puts('###############')
io.puts('# Parse Table #')
io.puts('###############')
io.puts
each_state do |state|
io.puts("State #{state.id}:")
io.puts
io.puts("\t# ITEMS #")
max = state.items.inject(0) do |inner_max, item|
if item.lhs.to_s.length > inner_max then item.lhs.to_s.length else inner_max end
end
state.each do |item|
io.puts("\t#{item.to_s(max)}")
end
io.puts
io.puts("\t# ACTIONS #")
state.actions.keys.sort {|a,b| a.to_s <=> b.to_s}.each do |sym|
state.actions[sym].each do |action|
io.puts("\tOn #{sym} #{action}")
end
end
io.puts
io.puts("\t# CONFLICTS #")
if @conflicts[state.id].length == 0
io.puts("\tNone\n\n")
else
@conflicts[state.id].each do |conflict|
type, sym = conflict
io.print("\t#{if type == :SR then "Shift/Reduce" else "Reduce/Reduce" end} conflict")
io.puts(" on #{sym}")
end
io.puts
end
end
# Close any IO objects that aren't $stdout.
io.close if io.is_a?(IO) and io != $stdout
else
raise ParserConstructionException, 'Parser.explain called outside of finalize.'
end
end
# This method will finalize the parser causing the construction
# of states and their actions, and the resolution of conflicts
# using lookahead and precedence information.
#
# No calls to {Parser.production} may appear after the call to
# Parser.finalize.
#
# @param [Hash] opts Options describing how to finalize the parser.
#
# @option opts [Boolean,String,IO] :explain To explain the parser or not.
# @option opts [Boolean] :lookahead To use lookahead info for conflict resolution.
# @option opts [Boolean] :precedence To use precedence info for conflict resolution.
# @option opts [String,IO] :use A file name or object that is used to load/save the parser.
#
# @return [void]
def finalize(opts = {})
if @grammar.productions.empty?
raise ParserConstructionException,
"Parser has no productions. Cowardly refusing to construct an empty parser."
end
# Get the full options hash.
opts = build_finalize_opts(opts)
# Get the name of the file in which the parser is defined.
#
# FIXME: See why this is failing for the simple ListParser example.
def_file = caller()[2].split(':')[0] if opts[:use]
# Check to make sure we can load the necessary information
# from the specified object.
if opts[:use] and (
(opts[:use].is_a?(String) and File.exist?(opts[:use]) and File.mtime(opts[:use]) > File.mtime(def_file)) or
(opts[:use].is_a?(File) and opts[:use].mtime > File.mtime(def_file))
)
file = self.get_io(opts[:use], 'r')
# Un-marshal our saved data structures.
file.flock(File::LOCK_SH)
@lh_sides, @states, @symbols = Marshal.load(file)
file.flock(File::LOCK_UN)
# Close the file if we opened it.
file.close if opts[:use].is_a?(String)
# Remove any un-needed data and return.
return self.clean
end
# Grab all of the symbols that comprise the grammar
# (besides the start symbol).
@symbols = @grammar.symbols << :ERROR
# Add our starting state to the state list.
@start_symbol = (@grammar.start_symbol.to_s + '\'').to_sym
start_production, _ = @grammar.production(@start_symbol, @grammar.start_symbol).first
start_state = State.new(@symbols, [start_production.to_item])
start_state.close(@grammar.productions)
self.add_state(start_state)
# Translate the precedence of productions from tokens to
# (associativity, precedence) pairs.
@production_precs.map! { |prec| @token_precs[prec] }
# Build the rest of the transition table.
each_state do |state|
#Transition states.
tstates = Hash.new { |h,k| h[k] = State.new(@symbols) }
#Bin each item in this set into reachable transition
#states.
state.each do |item|
if (next_symbol = item.next_symbol)
tstates[next_symbol] << item.copy
end
end
# For each transition state:
# 1) Get transition symbol
# 2) Advance dot
# 3) Close it
# 4) Get state id and add transition
tstates.each do |symbol, tstate|
tstate.each { |item| item.advance }
tstate.close(@grammar.productions)
id = self.add_state(tstate)
# Add Goto and Shift actions.
state.on(symbol, CFG::is_nonterminal?(symbol) ? GoTo.new(id) : Shift.new(id))
end
# Find the Accept and Reduce actions for this state.
state.each do |item|
if item.at_end?
if item.lhs == @start_symbol
state.on(:EOS, Accept.new)
else
state.add_reduction(@grammar.productions(:id)[item.id])
end
end
end
end
# Build the production.id -> production.lhs map.
@grammar.productions(:id).each { |id, production| @lh_sides[id] = production.lhs }
# Prune the parsing table for unnecessary reduce actions.
self.prune(opts[:lookahead], opts[:precedence])
# Check the parser for inconsistencies.
self.check_sanity
# Print the table if requested.
self.explain(opts[:explain]) if opts[:explain]
# Remove any data that is no longer needed.
self.clean
# Store the parser's final data structures if requested.
if opts[:use]
io = self.get_io(opts[:use])
io.flock(File::LOCK_EX) if io.is_a?(File)
Marshal.dump([@lh_sides, @states, @symbols], io)
io.flock(File::LOCK_UN) if io.is_a?(File)
# Close the IO object if we opened it.
io.close if opts[:use].is_a?(String)
end
end
# Converts an object into an IO object as appropriate.
#
# @param [Object] o Object to be converted into an IO object.
# @param [String] mode String representing the mode to open the IO object in.
#
# @return [IO, false] The IO object or false if a conversion wasn't possible.
def get_io(o, mode = 'w')
if o.is_a?(TrueClass)
$stdout
elsif o.is_a?(String)
File.open(o, mode)
elsif o.is_a?(IO)
o
else
false
end
end
# Iterate over the parser's states.
#
# @yieldparam [State] state One of the parser automaton's state objects
#
# @return [void]
def each_state
current_state = 0
while current_state < @states.count
yield @states.at(current_state)
current_state += 1
end
end
# @return [CFG] The grammar that can be parsed by this Parser.
def grammar
@grammar.clone
end
# This method generates and memoizes the G' grammar used to
# calculate the LALR(1) lookahead sets. Information about this
# grammar and its use can be found in the following paper:
#
# Simple Computation of LALR(1) Lookahead Sets
# Manuel E. Bermudez and George Logothetis
# Information Processing Letters 31 - 1989
#
# @return [CFG]
def grammar_prime
if not @grammar_prime
@grammar_prime = CFG.new
each_state do |state|
state.each do |item|
lhs = "#{state.id}_#{item.next_symbol}".to_sym
next unless CFG::is_nonterminal?(item.next_symbol) and not @grammar_prime.productions.keys.include?(lhs)
@grammar.productions[item.next_symbol].each do |production|
rhs = ''
cstate = state
production.rhs.each do |symbol|
rhs += "#{cstate.id}_#{symbol} "
cstate = @states[cstate.on?(symbol).first.id]
end
@grammar_prime.production(lhs, rhs)
end
end
end
end
@grammar_prime
end
# Inform the parser core that a conflict has been detected.
#
# @param [Integer] state_id ID of the state where the conflict was encountered.
# @param [:RR, :SR] type Reduce/Reduce or Shift/Reduce conflict.
# @param [Symbol] sym Symbol that caused the conflict.
#
# @return [void]
def inform_conflict(state_id, type, sym)
@conflicts[state_id] << [type, sym]
end
# This method is used to specify that the symbols in *symbols*
# are left-associative. Subsequent calls to this method will
# give their arguments higher precedence.
#
# @param [Array<Symbol>] symbols Symbols that are left associative.
#
# @return [void]
def left(*symbols)
prec_level = @prec_counts[:left] += 1
symbols.map { |s| s.to_sym }.each do |sym|
@token_precs[sym] = [:left, prec_level]
end
end
# This method is used to specify that the symbols in *symbols*
# are non-associative.
#
# @param [Array<Symbol>] symbols Symbols that are non-associative.
#
# @return [void]
def nonassoc(*symbols)
prec_level = @prec_counts[:non] += 1
symbols.map { |s| s.to_sym }.each do |sym|
@token_precs[sym] = [:non, prec_level]
end
end
# Adds productions and actions for parsing nonempty lists.
#
# @see CFG#nonempty_list_production
def build_nonempty_list_production(symbol, list_elements, separator = '')
@grammar.build_nonempty_list_production(symbol, list_elements, separator)
end
alias :nonempty_list :build_nonempty_list_production
# This function is where actual parsing takes place. The
# _tokens_ argument must be an array of Token objects, the last
# of which has type EOS. By default this method will return the
# value computed by the first successful parse tree found.
#
# Additional information about the parsing options can be found in
# the main documentation.
#
# @param [Array<Token>] tokens Tokens to be parsed.
# @param [Hash] opts Options to use when parsing input.
#
# @option opts [:first, :all] :accept Either :first or :all.
# @option opts [Object] :env The environment in which to evaluate the production action.
# @option opts [Boolean,String,IO] :parse_tree To print parse trees in the DOT language or not.
# @option opts [Boolean,String,IO] :verbose To be verbose or not.
#
# @return [Object, Array<Object>] Result or results of parsing the given tokens.
def parse(tokens, opts = {})
# Get the full options hash.
opts = build_parse_opts(opts)
v = opts[:verbose]
if opts[:verbose]
v.puts("Input tokens:")
v.puts(tokens.map { |t| t.type }.inspect)
v.puts
end
# Stack IDs to keep track of them during parsing.
stack_id = 0
# Error mode indicators.
error_mode = false
reduction_guard = false
# Our various list of stacks.
accepted = []
moving_on = []
processing = [ParseStack.new(stack_id += 1)]
# Iterate over the tokens. We don't procede to the
# next token until every stack is done with the
# current one.
tokens.each_with_index do |token, index|
# Check to make sure this token was seen in the
# grammar definition.
raise BadToken if not @symbols.include?(token.type)
v.puts("Current token: #{token.type}#{if token.value then "(#{token.value})" end}") if v
# Iterate over the stacks until each one is done.
while (stack = processing.shift)
# Execute any token hooks in this stack's environment.
@token_hooks[token.type].each { |hook| opts[:env].instance_exec(&hook)}
# Get the available actions for this stack.
actions = @states[stack.state].on?(token.type)
if actions.empty?
# If we are already in error mode and there
# are no actions we skip this token.
if error_mode
v.puts("Discarding token: #{token.type}#{if token.value then "(#{token.value})" end}") if v
# Add the current token to the array
# that corresponds to the output value
# for the ERROR token.
stack.output_stack.last << token
moving_on << stack
next
end
# We would be dropping the last stack so we
# are going to go into error mode.
if accepted.empty? and moving_on.empty? and processing.empty?
if v
v.puts
v.puts('Current stack:')
v.puts("\tID: #{stack.id}")
v.puts("\tState stack:\t#{stack.state_stack.inspect}")
v.puts("\tOutput Stack:\t#{stack.output_stack.inspect}")
v.puts
end
# Try and find a valid error state.
while stack.state
if (actions = @states[stack.state].on?(:ERROR)).empty?
# This state doesn't have an
# error production. Moving on.
stack.pop
else
# Enter the found error state.
stack.push(actions.first.id, [token], :ERROR, token.position)
break
end
end
if stack.state
# We found a valid error state.
error_mode = reduction_guard = true
opts[:env].he = true
moving_on << stack
if v
v.puts('Invalid input encountered. Entering error handling mode.')
v.puts("Discarding token: #{token.type}#{if token.value then "(#{token.value})" end}")
end
else
# No valid error states could be
# found. Time to print a message
# and leave.
v.puts("No more actions for stack #{stack.id}. Dropping stack.") if v
end
else
v.puts("No more actions for stack #{stack.id}. Dropping stack.") if v
end
next
end
# Make (stack, action) pairs, duplicating the
# stack as necessary.
pairs = [[stack, actions.pop]] + actions.map {|action| [stack.branch(stack_id += 1), action] }
pairs.each do |new_stack, action|
if v
v.puts
v.puts('Current stack:')
v.puts("\tID: #{new_stack.id}")
v.puts("\tState stack:\t#{new_stack.state_stack.inspect}")
v.puts("\tOutput Stack:\t#{new_stack.output_stack.inspect}")
v.puts
v.puts("Action taken: #{action.to_s}")
end
if action.is_a?(Accept)
if opts[:accept] == :all
accepted << new_stack
else
v.puts('Accepting input.') if v
opts[:parse_tree].puts(new_stack.tree) if opts[:parse_tree]
if opts[:env].he
raise HandledError.new(opts[:env].errors, new_stack.result)
else
return new_stack.result
end
end
elsif action.is_a?(Reduce)
# Get the production associated with this reduction.
production_proc, pop_size = @procs[action.id]
if not production_proc
raise InternalParserException, "No production #{action.id} found."
end
args, positions = new_stack.pop(pop_size)
opts[:env].set_positions(positions)
if not production_proc.selections.empty?
args = args.values_at(*production_proc.selections)
end
result =
if production_proc.arg_type == :array
opts[:env].instance_exec(args, &production_proc)
else
opts[:env].instance_exec(*args, &production_proc)
end
if (goto = @states[new_stack.state].on?(@lh_sides[action.id]).first)
v.puts("Going to state #{goto.id}.\n") if v
pos0 = nil
if args.empty?
# Empty productions need to be
# handled specially.
pos0 = new_stack.position
pos0.stream_offset += pos0.length + 1
pos0.line_offset += pos0.length + 1
pos0.length = 0
else
pos0 = opts[:env].pos( 0)
pos1 = opts[:env].pos(-1)
pos0.length = (pos1.stream_offset + pos1.length) - pos0.stream_offset
end
new_stack.push(goto.id, result, @lh_sides[action.id], pos0)
else
raise InternalParserException, "No GoTo action found in state #{stack.state} " +
"after reducing by production #{action.id}"
end