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]
55 # Do a transition in the automata
56 private fun go_to(index: Int): Int
60 var high = parser_goto(index, 0) - 1
63 var middle = (low + high) / 2
64 var subindex = middle * 2 + 1 # +1 because parser_goto(index, 0) is the length
66 var goal = parser_goto(index, subindex)
69 else if state > goal then
72 return parser_goto(index, subindex+1)
76 return parser_goto(index, 2) # Default value
79 # Push someting in the state stack
80 private fun push(numstate: Int, list_node: nullable Object)
82 var pos = _stack_pos + 1
84 if pos < _stack.length then
85 var state = _stack[pos]
86 state.state = numstate
87 state.nodes = list_node
89 _stack.push(new State(numstate, list_node))
94 private fun state: Int
96 return _stack[_stack_pos].state
99 # Pop something from the stack state
100 private fun pop: nullable Object
102 var res = _stack[_stack_pos].nodes
103 _stack_pos = _stack_pos -1
107 # Build and return a full AST.
114 var token = lexer.peek
115 if token isa PError then
116 return new Start(null, token)
119 var state = self.state
120 var index = token.parser_index
121 var action_type = parser_action(state, 2)
122 var action_value = parser_action(state, 3)
125 var high = parser_action(state, 0) - 1
128 var middle = (low + high) / 2
129 var subindex = middle * 3 + 1 # +1 because parser_action(state, 0) is the length
131 var goal = parser_action(state, subindex)
134 else if index > goal then
137 action_type = parser_action(state, subindex+1)
138 action_value = parser_action(state, subindex+2)
143 if action_type == 0 then # SHIFT
144 push(action_value, lexer.next)
145 else if action_type == 1 then # REDUCE
146 _reduce_table[action_value].action(self)
147 else if action_type == 2 then # ACCEPT
148 var node2 = lexer.next
151 assert node1 isa ${/parser/prods/prod/@ename}
152 var node = new Start(node1, node2)
153 (new ComputeProdLocationVisitor).enter_visit(node)
155 else if action_type == 3 then # ERROR
156 var node2 = new PParserError.init_parser_error("Syntax error: unexpected {token}.", token.location, token)
157 var node = new Start(null, node2)
163 var _reduce_table: Array[ReduceAction]
164 private fun build_reduce_table
166 _reduce_table = new Array[ReduceAction].with_items(
167 $ foreach {rules/rule}
168 new ReduceAction@index(@leftside)[-sep ','-]
175 # Location on the first token after the start of a production
176 # So outside the production for epilon production
177 var _first_location: nullable Location
180 # Find location of production nodes
181 # Uses existing token locations to infer location of productions.
182 private class ComputeProdLocationVisitor
184 # Currenlty visited productions that need a first token
185 var _need_first_prods: Array[Prod] = new Array[Prod]
187 # Already visited epsilon productions that waits something after them
188 var _need_after_epsilons: Array[Prod] = new Array[Prod]
190 # Location of the last visited token in the current production
191 var _last_location: nullable Location = null
193 redef fun visit(n: ANode)
199 # Add a first token to productions that need one
200 if not _need_first_prods.is_empty then
201 for no in _need_first_prods do
202 no._first_location = loc
204 _need_first_prods.clear
207 # Find location for already visited epsilon production that need one
208 if not _need_after_epsilons.is_empty then
209 var loco = new Location(loc.file, loc.line_start, loc.line_start, loc.column_start, loc.column_start)
210 for no in _need_after_epsilons do
213 _need_after_epsilons.clear
217 _need_first_prods.add(n)
221 var startl = n._first_location
222 if startl != null then
223 # Non-epsilon production
224 var endl = _last_location
227 n.location = new Location(startl.file, startl.line_start, endl.line_end, startl.column_start, endl.column_end)
229 if not _need_after_epsilons.is_empty then
230 var loc = new Location(endl.file, endl.line_end, endl.line_end, endl.column_end, endl.column_end)
231 for no in _need_after_epsilons do
232 # Epsilon production that finishes the current non-epsilon production
235 _need_after_epsilons.clear
238 # Epsilon production in the middle or that finishes a parent non-epsilon production
239 _need_after_epsilons.add(n)
247 # Each reduca action has its own class, this one is the root of the hierarchy.
248 private abstract class ReduceAction
249 fun action(p: Parser) is abstract
250 fun concat(l1, l2 : Array[Object]): Array[Object]
252 if l1.is_empty then return l2
257 init(g: Int) do _goto = g
260 $ foreach {rules/rule}
261 private class ReduceAction@index
263 redef fun action(p: Parser)
265 var node_list: nullable Object = null
269 var ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} = p.pop
271 $ when {@cmd='FETCHLIST'}
272 var ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} = ${translate(@from,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}
273 assert ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} isa Array[Object]
275 $ when {@cmd='FETCHNODE'}
276 var ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} = ${translate(@from,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}
277 assert ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} isa nullable @etype
279 $ when {@cmd='ADDNODE'}
280 if ${translate(@node,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} != null then
281 ${translate(@tolist,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}.add(${translate(@node,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")})
284 $ when {@cmd='ADDLIST'}
285 ${translate(@tolist,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} = concat(${translate(@tolist,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}, ${translate(@fromlist,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")})
287 $ when {@cmd='MAKELIST'}
288 var ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")} = new Array[Object]
290 $ when {@cmd='MAKENODE'}
292 var ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}: nullable @etype = new @etype.init_${translate(@etype,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}(
297 ${translate(.,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}[-sep ','-]
302 var ${translate(@result,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}: nullable @etype = new @etype.init_${translate(@etype,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}
305 $ when {@cmd='RETURNNODE'}
309 node_list = ${translate(@node,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}
312 $ when {@cmd='RETURNLIST'}
313 node_list = ${translate(@list,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}
317 p.push(p.go_to(_goto), node_list)
323 $ template make_parser_table()
324 $ foreach {parser_data/action_table/row}
325 static int parser_action_row${position()}[] = {
328 @from, @action, @to[-sep ','-]
333 const int* const parser_action_table[] = {
334 $ foreach {parser_data/action_table/row}
335 parser_action_row${position()}[-sep ','-]
339 $ foreach {parser_data/goto_table/row}
340 static int parser_goto_row${position()}[] = {
343 @from, @to[-sep ','-]
348 const int* const parser_goto_table[] = {
349 $ foreach {parser_data/goto_table/row}
350 parser_goto_row${position()}[-sep ','-]