1 $ // This file is part of NIT ( http://www.nitlanguage.org ).
3 $ // Copyright 2008 Jean Privat <jean@pryen.org>
4 $ // Based on algorithms developped for ( http://www.sablecc.org/ ).
6 $ // Licensed under the Apache License, Version 2.0 (the "License");
7 $ // you may not use this file except in compliance with the License.
8 $ // You may obtain a copy of the License at
10 $ // http://www.apache.org/licenses/LICENSE-2.0
12 $ // Unless required by applicable law or agreed to in writing, software
13 $ // distributed under the License is distributed on an "AS IS" BASIS,
14 $ // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
15 $ // See the License for the specific language governing permissions and
16 $ // limitations under the License.
18 $ template make_parser()
20 # State of the parser automata as stored in the parser stack.
22 # The internal state number
23 readable writable var _state: Int
25 # The node stored with the state in the stack
26 readable writable var _nodes: nullable Object
28 init(state: Int, nodes: nullable Object)
40 # Stack of pushed states and productions
41 var _stack: Array[State]
43 # Position in the stack
46 # Create a new parser based on a given lexer
50 _stack = new Array[State]
57 # Do a transition in the automata
58 private fun go_to(index: Int): Int
61 var table = _goto_table[index]
63 var high = table.length/2 - 1
66 var middle = (low + high) / 2
67 var subindex = middle * 2
69 if state < table[subindex] then
71 else if state > table[subindex] then
74 return table[subindex + 1]
78 return table[1] # Default value
81 # Push someting in the state stack
82 private fun push(numstate: Int, list_node: nullable Object)
84 var pos = _stack_pos + 1
86 if pos < _stack.length then
87 var state = _stack[pos]
88 state.state = numstate
89 state.nodes = list_node
91 _stack.push(new State(numstate, list_node))
96 private fun state: Int
98 return _stack[_stack_pos].state
101 # Pop something from the stack state
102 private fun pop: nullable Object
104 var res = _stack[_stack_pos].nodes
105 _stack_pos = _stack_pos -1
109 # Build and return a full AST.
116 var token = lexer.peek
117 var last_pos = token.location.column_start
118 var last_line = token.location.line_start
120 if token isa PError then
121 return new Start(null, token)
124 var index = token.parser_index
125 var table = _action_table[state]
126 var action_type = table[1]
127 var action_value = table[2]
130 var high = table.length/3 - 1
133 var middle = (low + high) / 2
134 var subindex = middle * 3
136 if index < table[subindex] then
138 else if index > table[subindex] then
141 action_type = table[subindex + 1]
142 action_value = table[subindex + 2]
143 high = low -1 # break
147 if action_type == 0 then # SHIFT
148 push(action_value, lexer.next)
149 else if action_type == 1 then # REDUCE
150 _reduce_table[action_value].action(self)
151 else if action_type == 2 then # ACCEPT
152 var node2 = lexer.next
155 assert node1 isa ${/parser/prods/prod/@ename}
156 var node = new Start(node1, node2)
157 (new ComputeProdLocationVisitor).enter_visit(node)
159 else if action_type == 3 then # ERROR
160 var location = new Location(lexer.filename, last_line, last_line, last_pos, last_pos)
161 var node2 = new PError.init_error(error_messages[errors[action_value]],location)
162 var node = new Start(null, node2)
169 var _reduce_table: Array[ReduceAction]
170 private fun build_reduce_table
172 _reduce_table = new Array[ReduceAction].with_items(
173 $ foreach {rules/rule}
174 new ReduceAction@index[-sep ','-]
181 # Location on the first token after the start of a production
182 # So outside the production for epilon production
183 var _first_location: nullable Location
185 # Location of the last token before the end of a production
186 # So outside the production for epilon production
187 var _last_location: nullable Location
190 # Find location of production nodes
191 # Uses existing token locations to infer location of productions.
192 private class ComputeProdLocationVisitor
194 # Currenlty visited productions that need a first token
195 var _need_first_prods: Array[Prod] = new Array[Prod]
197 # Already visited epsilon productions that waits something after them
198 var _need_after_epsilons: Array[Prod] = new Array[Prod]
200 # Already visited epsilon production that waits something before them
201 var _need_before_epsilons: Array[Prod] = new Array[Prod]
203 # Location of the last visited token in the current production
204 var _last_location: nullable Location = null
206 redef fun visit(n: nullable PNode)
210 else if n isa Token then
214 # Add a first token to productions that need one
215 for no in _need_first_prods do
216 no._first_location = loc
218 _need_first_prods.clear
220 # Find location for already visited epsilon production that need one
221 for no in _need_after_epsilons do
222 # Epsilon production that is in the middle of a non-epsilon production
223 # The epsilon production has both a token before and after it
225 var startl = no._last_location
226 no.location = new Location(endl.file, startl.line_end, endl.line_start, startl.column_end, endl.column_start)
228 _need_after_epsilons.clear
231 _need_first_prods.add(n)
233 var old_last = _last_location
234 _last_location = null
236 var endl = _last_location
237 if endl == null then _last_location = old_last
239 n._last_location = endl
240 var startl = n._first_location
241 if startl != null then
242 # Non-epsilon production
245 n.location = new Location(startl.file, startl.line_start, endl.line_end, startl.column_start, endl.column_end)
247 for no in _need_before_epsilons do
248 # Epsilon production that starts the current non-epsilon production
249 #var startl = n.location
250 no.location = new Location(startl.file, startl.line_start, startl.line_start, startl.column_start, startl.column_start)
252 _need_before_epsilons.clear
254 for no in _need_after_epsilons do
255 # Epsilon production that finishes the current non-epsilon production
256 #var endl = n.location
257 no.location = new Location(endl.file, endl.line_end, endl.line_end, endl.column_end, endl.column_end)
259 _need_after_epsilons.clear
261 # No first token means epsilon production (or "throw all my tokens" production)
262 # So, it must be located it later
264 # Epsilon production that starts a parent non-epsilon production
265 _need_before_epsilons.add(n)
267 # Epsilon production in the middle or that finishes a parent non-epsilon production
268 _need_after_epsilons.add(n)
277 # Each reduca action has its own class, this one is the root of the hierarchy.
278 private abstract class ReduceAction
279 fun action(p: Parser) is abstract
282 $ foreach {rules/rule}
283 private class ReduceAction@index
285 redef fun action(p: Parser)
287 var node_list: nullable Object = null
291 var ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} = p.pop
293 $ when {@cmd='FETCHLIST'}
294 var ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} = ${translate(@from,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}
295 assert ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} isa Array[Object]
297 $ when {@cmd='FETCHNODE'}
298 var ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} = ${translate(@from,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}
299 assert ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} isa nullable @etype
301 $ when {@cmd='ADDNODE'}
302 if ${translate(@node,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} != null then
303 ${translate(@tolist,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}.add(${translate(@node,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")})
306 $ when {@cmd='ADDLIST'}
307 # if ${translate(@fromlist,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} != null then
308 if ${translate(@tolist,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}.is_empty then
309 ${translate(@tolist,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} = ${translate(@fromlist,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}
311 ${translate(@tolist,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}.append(${translate(@fromlist,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")})
315 $ when {@cmd='MAKELIST'}
316 var ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} = new Array[Object]
318 $ when {@cmd='MAKENODE'}
319 var ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}: nullable @etype = new @etype.init_${translate(@etype,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}(
324 ${translate(.,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}[-sep ','-]
329 $ when {@cmd='RETURNNODE'}
333 node_list = ${translate(@node,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}
336 $ when {@cmd='RETURNLIST'}
337 node_list = ${translate(@list,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}
341 p.push(p.go_to(@leftside), node_list)
348 $ template make_parser_tables()
349 # Parser that build a full AST
350 abstract class ParserTable
351 var _action_table: Array[Array[Int]]
352 private fun build_action_table
354 _action_table = once [
355 $ foreach {parser_data/action_table/row}
356 action_table_row${position()}[-sep ','-]
361 $ foreach {parser_data/action_table/row}
362 private fun action_table_row${position()}: Array[Int]
366 @from, @action, @to[-sep ','-]
372 var _goto_table: Array[Array[Int]]
373 private fun build_goto_table
376 $ foreach {parser_data/goto_table/row}
379 @from, @to[-sep ','-]
386 private fun error_messages: Array[String]
389 $ foreach {parser_data/error_messages/msg}
390 "${sablecc:string2escaped_unicode(.)}"[-sep ','-]
395 private fun errors: Array[Int]
398 [-foreach {parser_data/errors/i}-]${.}[-sep ','-][-end-]