-/* This file is part of NIT ( http://www.nitlanguage.org ).
- *
- * Copyright 2008 Jean Privat <jean@pryen.org>
- * Based on algorithms developped for ( http://www.sablecc.org/ ).
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- * http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
+$ // This file is part of NIT ( http://www.nitlanguage.org ).
+$ //
+$ // Copyright 2008 Jean Privat <jean@pryen.org>
+$ // Based on algorithms developped for ( http://www.sablecc.org/ ).
+$ //
+$ // Licensed under the Apache License, Version 2.0 (the "License");
+$ // you may not use this file except in compliance with the License.
+$ // You may obtain a copy of the License at
+$ //
+$ // http://www.apache.org/licenses/LICENSE-2.0
+$ //
+$ // Unless required by applicable law or agreed to in writing, software
+$ // distributed under the License is distributed on an "AS IS" BASIS,
+$ // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+$ // See the License for the specific language governing permissions and
+$ // limitations under the License.
$ template make_parser()
readable writable var _state: Int
# The node stored with the state in the stack
- readable writable var _nodes: nullable Object
+ readable writable var _nodes: nullable Object
init(state: Int, nodes: nullable Object)
do
var lexer = _lexer
while true do
var token = lexer.peek
- var last_pos = token.pos
- var last_line = token.line
+ var last_pos = token.location.column_start
+ var last_line = token.location.line_start
if token isa PError then
return new Start(null, token)
var node1 = pop
assert node1 isa ${/parser/prods/prod/@ename}
var node = new Start(node1, node2)
- (new SearchTokensVisitor).visit(node)
+ (new ComputeProdLocationVisitor).enter_visit(node)
return node
else if action_type == 3 then # ERROR
- var node2 = new PError.init_error(lexer.filename, last_line, last_pos, error_messages[errors[action_value]])
- node2._location = new Location(lexer.filename, last_line, last_line, last_pos, last_pos)
+ var location = new Location(lexer.filename, last_line, last_line, last_pos, last_pos)
+ var node2 = new PError.init_error(error_messages[errors[action_value]],location)
var node = new Start(null, node2)
return node
end
end
end
-# Find first and last tokens of production nodes
-private class SearchTokensVisitor
+redef class Prod
+ # Location on the first token after the start of a production
+ # So outside the production for epilon production
+ var _first_location: nullable Location
+
+ # Location of the last token before the end of a production
+ # So outside the production for epilon production
+ var _last_location: nullable Location
+end
+
+# Find location of production nodes
+# Uses existing token locations to infer location of productions.
+private class ComputeProdLocationVisitor
special Visitor
- var _untokenned_nodes: Array[Prod]
- var _last_token: nullable Token = null
+ # Currenlty visited productions that need a first token
+ var _need_first_prods: Array[Prod] = new Array[Prod]
+
+ # Already visited epsilon productions that waits something after them
+ var _need_after_epsilons: Array[Prod] = new Array[Prod]
+
+ # Already visited epsilon production that waits something before them
+ var _need_before_epsilons: Array[Prod] = new Array[Prod]
+
+ # Location of the last visited token in the current production
+ var _last_location: nullable Location = null
+
redef fun visit(n: nullable PNode)
do
if n == null then
return
else if n isa Token then
- _last_token = n
- for no in _untokenned_nodes do
- no.first_token = n
+ var loc = n.location
+ _last_location = loc
+
+ # Add a first token to productions that need one
+ for no in _need_first_prods do
+ no._first_location = loc
+ end
+ _need_first_prods.clear
+
+ # Find location for already visited epsilon production that need one
+ for no in _need_after_epsilons do
+ # Epsilon production that is in the middle of a non-epsilon production
+ # The epsilon production has both a token before and after it
+ var endl = loc
+ var startl = no._last_location
+ no.location = new Location(endl.file, startl.line_end, endl.line_start, startl.column_end, endl.column_start)
end
- _untokenned_nodes.clear
+ _need_after_epsilons.clear
else
assert n isa Prod
- _untokenned_nodes.add(n)
+ _need_first_prods.add(n)
+
+ var old_last = _last_location
+ _last_location = null
n.visit_all(self)
- n.last_token = _last_token
-
- if n.first_token != null then
- var start_location = n.first_token.location
- var end_location = _last_token.location
-
- if start_location != null and end_location != null then
- var file = end_location.file
- var line_start = start_location.line_start
- var line_end = end_location.line_end
- var column_start = start_location.column_start
- var column_end = end_location.column_end
- n._location = new Location(file, line_start, line_end, column_start, column_end)
+ var endl = _last_location
+ if endl == null then _last_location = old_last
+
+ n._last_location = endl
+ var startl = n._first_location
+ if startl != null then
+ # Non-epsilon production
+ assert endl != null
+
+ n.location = new Location(startl.file, startl.line_start, endl.line_end, startl.column_start, endl.column_end)
+
+ for no in _need_before_epsilons do
+ # Epsilon production that starts the current non-epsilon production
+ #var startl = n.location
+ no.location = new Location(startl.file, startl.line_start, startl.line_start, startl.column_start, startl.column_start)
+ end
+ _need_before_epsilons.clear
+
+ for no in _need_after_epsilons do
+ # Epsilon production that finishes the current non-epsilon production
+ #var endl = n.location
+ no.location = new Location(endl.file, endl.line_end, endl.line_end, endl.column_end, endl.column_end)
+ end
+ _need_after_epsilons.clear
+ else
+ # No first token means epsilon production (or "throw all my tokens" production)
+ # So, it must be located it later
+ if endl == null then
+ # Epsilon production that starts a parent non-epsilon production
+ _need_before_epsilons.add(n)
+ else
+ # Epsilon production in the middle or that finishes a parent non-epsilon production
+ _need_after_epsilons.add(n)
end
end
end
end
- init
- do
- _untokenned_nodes = new Array[Prod]
- end
+
+ init do end
end
# Each reduca action has its own class, this one is the root of the hierarchy.
var _action_table: Array[Array[Int]]
private fun build_action_table
do
- _action_table = once [
+ _action_table = once [
$ foreach {parser_data/action_table/row}
- action_table_row${position()}[-sep ','-]
+ action_table_row${position()}[-sep ','-]
$ end foreach
]
end
do
return [
$ foreach {action}
- @from, @action, @to [-sep ','-]
+ @from, @action, @to[-sep ','-]
$ end foreach
- ]
+ ]
end
$ end foreach
var _goto_table: Array[Array[Int]]
private fun build_goto_table
do
- _goto_table = once [
+ _goto_table = once [
$ foreach {parser_data/goto_table/row}
[
$ foreach {goto}
- @from, @to [-sep ','-]
+ @from, @to[-sep ','-]
$ end foreach
- ] [-sep ','-]
+ ][-sep ','-]
$ end foreach
]
end
do
return once [
$ foreach {parser_data/error_messages/msg}
- "${sablecc:string2escaped_unicode(.)}" [-sep ','-]
+ "${sablecc:string2escaped_unicode(.)}"[-sep ','-]
$ end
]
end
private fun errors: Array[Int]
do
return once [
- [-foreach {parser_data/errors/i}-]${.} [-sep ','-] [-end-]
+ [-foreach {parser_data/errors/i}-]${.}[-sep ','-][-end-]
]
end