NitCC LALR and LR(1) #2843

I'ts happening!  nitcc can do LALR(1) (by default) and LR(1).

We also split the grammar.nit module into two by extracting a lrautomaton.nit module.

contrib/nitcc/src/nitcc.nit

@@ -16,6 +16,7 @@
 module nitcc
 
 import nitcc_semantic
+import lrautomaton
 
 # Load the grammar file
 
@@ -69,7 +70,9 @@ end
 
 # Generate the LR automaton
 
-var lr = gram.lr0
+gram.prepare_for_automaton
+var lr = new LALR1Automaton(gram)
+lr.build
 
 var conflitcs = new ArraySet[Production]
 for s in lr.states do
@@ -80,13 +83,6 @@ if not conflitcs.is_empty then
 	print "Error: there is conflicts"
 end
 
-if false then loop
-if conflitcs.is_empty then break
-print "Inline {conflitcs.join(" ")}"
-gram.inline(conflitcs)
-lr=gram.lr0
-end
-
 # Output concrete grammar and LR automaton
 
 var nbalts = 0

contrib/nitcc/src/nitcc_parser_gen.nit

@@ -29,7 +29,7 @@
 #
 module nitcc_parser_gen
 
-import grammar
+import lrautomaton
 
 var g = new Gram
 var p_gr = new Production("grammar")
@@ -195,7 +195,9 @@ p_pri.new_alt0("priority_left").phony = true
 p_pri.new_alt0("priority_right").phony = true
 p_pri.new_alt0("priority_unary").phony = true
 
-var a = g.lr0
+g.prepare_for_automaton
+var a = new LALR1Automaton(g)
+a.build
 
 print "LR automaton: {a.states.length} states (see nitcc0.lr.dot)"
 a.to_dot("nitcc0.lr.dot")

contrib/nitcc/tests/amb_repeat.input

@@ -0,0 +1 @@
+c x

contrib/nitcc/tests/amb_repeat.sablecc

@@ -0,0 +1,6 @@
+Grammar amb;
+Parser
+Ignored #10, #32;
+s = a | a 'x' ;
+a = b | b 'x' ;
+b = 'c' ;

contrib/nitcc/tests/amb_repeat0.input

@@ -0,0 +1 @@
+hello

contrib/nitcc/tests/amb_repeat0.sablecc

@@ -0,0 +1,4 @@
+Grammar amb;
+Parser
+Ignored #10, #32;
+s = 'hello' | 'hello' ;

contrib/nitcc/tests/amb_repeat2.sablecc

@@ -0,0 +1,6 @@
+Grammar amb;
+Parser
+Ignored #10, #32;
+s = a | Empty ;
+a = b | Empty ;
+b = 'c' ;

contrib/nitcc/tests/calc1.sablecc

@@ -0,0 +1,13 @@
+Grammar calc;
+Lexer
+n = ('0'..'9')+ ;
+Parser
+Ignored #10, #32;
+e
+	= e '+' a
+	| e '-' a
+	;
+a	=
+	| '(' e ')'
+	| n
+	;

contrib/nitcc/tests/dangling_else.input

@@ -0,0 +1 @@
+i a i a e a

contrib/nitcc/tests/dangling_else.sablecc

@@ -0,0 +1,9 @@
+Grammar dangling;
+Parser
+Ignored #10, #32;
+ss = s*;
+s
+	= 'i' s 'e' s
+	| 'i' s
+	| 'a'
+	;

contrib/nitcc/tests/dangling_else2.input

@@ -0,0 +1 @@
+i a i a e a

contrib/nitcc/tests/dangling_else2.sablecc

@@ -0,0 +1,14 @@
+Grammar dangling;
+Parser
+Ignored #10, #32;
+ss = s*;
+
+s
+	= 'a'
+	| 'i' se 'e' s
+	| 'i' s
+	;
+se
+	= 'a'
+	| 'i' se 'e' se
+	;

contrib/nitcc/tests/dangling_end.input

@@ -0,0 +1,5 @@
+a id
+a id end
+a a id end end
+a a id end
+a a id

contrib/nitcc/tests/dangling_end.sablecc

@@ -0,0 +1,9 @@
+Grammar dangling;
+Parser
+Ignored #10, #32;
+es = e*;
+e
+	= 'a' e 'end'
+	| 'a' e
+	| 'id'
+	;

contrib/nitcc/tests/dragon_4-55.input

@@ -0,0 +1 @@
+ccdd

contrib/nitcc/tests/dragon_4-55.sablecc

@@ -0,0 +1,5 @@
+Grammar dragon;
+Parser
+Ignored #10, #32;
+s = c c ;
+c = 'c' c | 'd' ;

contrib/nitcc/tests/oneliner.input

@@ -0,0 +1,4 @@
+if id then id
+if id then id end
+if if id then id then id end
+if id then if id then id end

contrib/nitcc/tests/oneliner.sablecc

@@ -0,0 +1,34 @@
+Grammar oneliner;
+Lexer
+eol = #10 | #13 ;
+
+Parser
+Ignored #32;
+
+prog
+	= stmts
+	;
+stmt
+	= if_block stmt_or_end
+	| if_block stmt_else stmt_or_end
+	| 'id'
+	;
+if_block
+	= 'if' n? stmt no 'then'
+	;
+stmt_or_end
+	= stmt 'end'?
+	| n stmts no 'end'
+	| no 'end'
+	;
+stmt_else
+	= stmt 'else'
+	| n stmts no 'else'
+	| 'else'
+	;
+stmts
+	= stmts n stmt
+	| stmt
+	;
+no	= n?;
+n	= eol+;

contrib/nitcc/tests/oneliner2.sablecc

@@ -0,0 +1,42 @@
+Grammar oneliner;
+Lexer
+eol = #10 | #13 ;
+
+Parser
+Ignored #32;
+
+prog
+	= stmts
+	;
+stmt
+	= if_block stmt_or_end
+	| if_block stmt_else stmt_or_end
+	| 'id'
+	;
+stmt_with_else
+	= if_block stmt_else stmt_or_end
+	| 'id'
+	;
+if_block
+	= 'if' n? stmt no 'then'
+	;
+stmt_or_end
+	= stmt 'end'?
+	| n stmts no 'end'
+	| no 'end'
+	;
+stmt_end
+	= stmt 'else'
+	| n stmts no 'else'
+	| 'else'
+	;
+stmts_with_else
+	= stmts_with_else n stmt_with_else
+	| stmt_with_else
+	;
+stmts
+	= stmts n stmt
+	| stmt
+	;
+no	= n?;
+n	= eol+;

contrib/nitcc/tests/regex0.input

@@ -0,0 +1 @@
+aa|aa*

contrib/nitcc/tests/regex0.sablecc

@@ -0,0 +1,5 @@
+Grammar regex;
+Parser
+Ignored #10, #32;
+e = a | e '|' e | e e | e '*' ;
+a = 'a' ;

contrib/nitcc/tests/regex1.input

@@ -0,0 +1 @@
+aa|aa*

contrib/nitcc/tests/regex1.sablecc

@@ -0,0 +1,13 @@
+Grammar regex;
+Parser
+Ignored #10, #32;
+e
+	= a
+Unary
+	e '*'
+Left
+	e e
+Left
+	e '|' e
+	;
+a = 'a' ;

contrib/nitcc/tests/regex2.input

@@ -0,0 +1 @@
+aa|aa*

contrib/nitcc/tests/regex2.sablecc

@@ -0,0 +1,7 @@
+Grammar regex;
+Parser
+Ignored #10, #32;
+e  = e '|' e2 | e2;
+e2 = e2 e3 | e3 ;
+e3 = e4 '*' | e4 ;
+e4 = 'a' ;

contrib/nitcc/tests/sav/amb.res

@@ -1 +1,7 @@
+REDUCE/REDUCE Conflict on state 1 e for token Eof:
+REDUCE on item: e→ e·, Eof
+REDUCE on item: e→ e·, Eof
+SHIFT/REDUCE Conflict on state 1 e for token Eof:
+SHIFT on item: _start→ e·Eof
+REDUCE on item: e→ e·, Eof
 Error: there is conflicts

contrib/nitcc/tests/sav/amb2.res

@@ -1 +1,4 @@
+REDUCE/REDUCE Conflict on state 7 '-' e for token '+':
+REDUCE on item: a→ e·, Eof/'+'
+REDUCE on item: b→ e·, '+'
 Error: there is conflicts

contrib/nitcc/tests/sav/amb3.res

@@ -1 +1,4 @@
+SHIFT/REDUCE Conflict on state 3 'a' for token 'b':
+SHIFT on item: x→·'b', Eof
+REDUCE on item: x→ 'a'·, 'b'
 Error: there is conflicts

contrib/nitcc/tests/sav/amb_repeat.input.res

@@ -0,0 +1,7 @@
+Start
+  s_0
+    a_1
+      b
+        'c'@(1:1-1:2)
+      'x'@(1:3-1:4)
+  Eof@(2:1-2:1)=''

contrib/nitcc/tests/sav/amb_repeat.res

@@ -0,0 +1,4 @@
+SHIFT/REDUCE Conflict on state 3 b for token 'x':
+SHIFT on item: a→ b·'x', Eof/'x'
+REDUCE on item: a→ b·, Eof/'x'
+Error: there is conflicts

contrib/nitcc/tests/sav/amb_repeat0.input.res

@@ -0,0 +1,4 @@
+Start
+  s_0
+    'hello'@(1:1-1:6)
+  Eof@(2:1-2:1)=''

contrib/nitcc/tests/sav/amb_repeat0.res

@@ -0,0 +1,4 @@
+REDUCE/REDUCE Conflict on state 2 'hello' for token Eof:
+REDUCE on item: s→ 'hello'·, Eof
+REDUCE on item: s→ 'hello'·, Eof
+Error: there is conflicts

contrib/nitcc/tests/sav/amb_repeat2.input.res

@@ -0,0 +1,4 @@
+Start
+  s_0
+    a_1
+  Eof@(1:1-1:1)=''

contrib/nitcc/tests/sav/amb_repeat2.res

@@ -0,0 +1,4 @@
+REDUCE/REDUCE Conflict on state 0  for token Eof:
+REDUCE on item: a→·, Eof
+REDUCE on item: s→·, Eof
+Error: there is conflicts

contrib/nitcc/tests/sav/calc0.res

@@ -0,0 +1,12 @@
+Automatic Dangling SHIFT/REDUCE on state 8 e '+' e for token '+':
+	reduce: e→ e '+' e·, Eof/'+'/'*'/')'
+	shift:  e→ e·'+' e
+Automatic Dangling SHIFT/REDUCE on state 8 e '+' e for token '*':
+	reduce: e→ e '+' e·, Eof/'+'/'*'/')'
+	shift:  e→ e·'*' e
+Automatic Dangling SHIFT/REDUCE on state 9 e '*' e for token '+':
+	reduce: e→ e '*' e·, '+'/'*'/Eof/')'
+	shift:  e→ e·'+' e
+Automatic Dangling SHIFT/REDUCE on state 9 e '*' e for token '*':
+	reduce: e→ e '*' e·, '+'/'*'/Eof/')'
+	shift:  e→ e·'*' e

contrib/nitcc/tests/sav/calc1.res

@@ -0,0 +1 @@
+Runtime error: Aborted (./lrautomaton.nit:507)

contrib/nitcc/tests/sav/conflict-bracket.res

@@ -1 +1,4 @@
+SHIFT/REDUCE Conflict on state 8 '[' e '[' for token '[':
+SHIFT on item: e→·'[' e '[', ']'/'['
+REDUCE on item: e→ '[' e '['·, Eof/'['/']'
 Error: there is conflicts

contrib/nitcc/tests/sav/conflict-dangling.1alt1.alt1.res

@@ -1 +1,4 @@
+SHIFT/REDUCE Conflict on state 7 '1' e for token '2':
+SHIFT on item: x→·'2', Eof/'2'
+REDUCE on item: e→ '1' e·, Eof/'2'
 Error: there is conflicts

contrib/nitcc/tests/sav/conflict-dangling.1alt1.alt2.res

@@ -1 +1,4 @@
+SHIFT/REDUCE Conflict on state 7 '1' e for token '2':
+SHIFT on item: x→·'2', Eof/'2'
+REDUCE on item: y→·, Eof/'2'
 Error: there is conflicts

contrib/nitcc/tests/sav/conflict-dangling.1alt1.res

@@ -1 +1,4 @@
+SHIFT/REDUCE Conflict on state 7 '1' e for token '2':
+SHIFT on item: e→ '1' e·'2', Eof/'2'
+REDUCE on item: e→ '1' e·, Eof/'2'
 Error: there is conflicts

contrib/nitcc/tests/sav/conflict-dangling.alt1.res

@@ -0,0 +1,4 @@
+Automatic Dangling SHIFT/REDUCE on state 6 '1' e for token '2':
+	reduce: e→ '1' e·, Eof/'2'
+	shift:  x→·'2'
+	core shift: e→ '1' e·x, Eof/'2'

contrib/nitcc/tests/sav/conflict-dangling.alt2.res

@@ -0,0 +1,4 @@
+Automatic Dangling SHIFT/REDUCE on state 6 '1' e for token '2':
+	reduce: y→·, Eof/'2'
+	shift:  x→·'2'
+	core shift: y→·x, Eof/'2'

contrib/nitcc/tests/sav/conflict-dangling.res

@@ -0,0 +1,3 @@
+Automatic Dangling SHIFT/REDUCE on state 6 '1' e for token '2':
+	reduce: e→ '1' e·, Eof/'2'
+	shift:  e→ '1' e·'2'

contrib/nitcc/tests/sav/dangling_else.input.res

@@ -0,0 +1,15 @@
+Start
+  ss
+    Nodes[Ns]
+      s_1
+        'i'@(1:1-1:2)
+        s_2
+          'a'@(1:3-1:4)
+      s_0
+        'i'@(1:5-1:6)
+        s_2
+          'a'@(1:7-1:8)
+        'e'@(1:9-1:10)
+        s_2
+          'a'@(1:11-1:12)
+  Eof@(2:1-2:1)=''

contrib/nitcc/tests/sav/dangling_else.res

@@ -0,0 +1,3 @@
+Automatic Dangling SHIFT/REDUCE on state 8 'i' s for token 'e':
+	reduce: s→ 'i' s·, Eof/'i'/'a'/'e'
+	shift:  s→ 'i' s·'e' s

contrib/nitcc/tests/sav/dangling_else2.input.res

@@ -0,0 +1,15 @@
+Start
+  ss
+    Nodes[Ns]
+      s_2
+        'i'@(1:1-1:2)
+        s_0
+          'a'@(1:3-1:4)
+      s_1
+        'i'@(1:5-1:6)
+        se_0
+          'a'@(1:7-1:8)
+        'e'@(1:9-1:10)
+        s_0
+          'a'@(1:11-1:12)
+  Eof@(2:1-2:1)=''