# Exercising Bison on conflicts. -*- Autotest -*-
# Copyright (C) 2002-2005, 2007-2015 Free Software Foundation, Inc.
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
AT_BANNER([[Conflicts.]])
## ------------------------- ##
## Token declaration order. ##
## ------------------------- ##
# This test checks that token are declared left to right when in a precedence
# statement.
AT_SETUP([Token declaration order])
AT_BISON_OPTION_PUSHDEFS
AT_DATA_GRAMMAR([[input.y]],
[[%code {
#include <stdio.h>
]AT_YYERROR_DECLARE[
]AT_YYLEX_DECLARE[
}
%token A B C
%token D
%right E F G
%right H I
%right J
%left K
%left L M N
%nonassoc O P Q
%precedence R S T U
%precedence V W
%%
exp: A
%%
]AT_YYERROR_DEFINE[
]AT_YYLEX_DEFINE[
int main (void)
{
assert (A < B);
assert (B < C);
assert (C < D);
assert (D < E);
assert (E < F);
assert (F < G);
assert (G < H);
assert (H < I);
assert (I < J);
assert (J < K);
assert (K < L);
assert (L < M);
assert (M < N);
assert (N < O);
assert (O < P);
assert (P < Q);
assert (Q < R);
assert (R < S);
assert (S < T);
assert (T < U);
assert (U < V);
assert (V < W);
return 0;
}
]])
AT_BISON_CHECK([-o input.c input.y])
AT_COMPILE([input])
AT_PARSER_CHECK([./input])
AT_BISON_OPTION_POPDEFS
AT_CLEANUP
## --------------------------------------------------- ##
## Token declaration order: literals vs. identifiers. ##
## --------------------------------------------------- ##
# This test checks that when several tokens are declared by the same keyword,
# some of them defined as a character ('a'), others as simple textual reference
# (A), they are declared correctly left to right.
# Previously, the following test would declare the states in the order 'o' 'p'
# M N, instead of M N 'o' 'p'.
AT_SETUP([Token declaration order: literals vs. identifiers])
AT_DATA_GRAMMAR([[input.y]],
[[%token 'a' 'b' C D
%token E F 'g' 'h'
%right 'i' 'j' K L
%right M N 'o' 'p'
%%
exp: 'a'
| 'b'
| C
| D
| E
| F
| 'g'
| 'h'
| 'i'
| 'j'
| K
| L
| M
| N
| 'o'
| 'p'
;
%%
]])
AT_BISON_CHECK([[--report=all -o input.c input.y]], 0, [], [ignore])
AT_CHECK([[cat input.output | sed -n '/^State 0$/,/^State 1$/p']], 0,
[[State 0
0 $accept: . exp $end
1 exp: . 'a'
2 | . 'b'
3 | . C
4 | . D
5 | . E
6 | . F
7 | . 'g'
8 | . 'h'
9 | . 'i'
10 | . 'j'
11 | . K
12 | . L
13 | . M
14 | . N
15 | . 'o'
16 | . 'p'
'a' shift, and go to state 1
'b' shift, and go to state 2
C shift, and go to state 3
D shift, and go to state 4
E shift, and go to state 5
F shift, and go to state 6
'g' shift, and go to state 7
'h' shift, and go to state 8
'i' shift, and go to state 9
'j' shift, and go to state 10
K shift, and go to state 11
L shift, and go to state 12
M shift, and go to state 13
N shift, and go to state 14
'o' shift, and go to state 15
'p' shift, and go to state 16
exp go to state 17
State 1
]])
AT_CLEANUP
## ------------------------------- ##
## Useless associativity warning. ##
## ------------------------------- ##
AT_SETUP([Useless associativity warning])
AT_DATA([[input.y]],
[[%token EQ "=" PL "+" ST "*" LP "("
%nonassoc "="
%left "+"
%left "*"
%precedence "("
%%
stmt:
exp
| "var" "=" exp
;
exp:
exp "+" exp
| exp "*" "num"
| "(" exp ")"
| "num"
;
]])
AT_BISON_CHECK([-Wprecedence input.y], 0, [],
[[input.y:2.1-9: warning: useless precedence and associativity for "=" [-Wprecedence]
input.y:4.1-5: warning: useless associativity for "*", use %precedence [-Wprecedence]
input.y:5.1-11: warning: useless precedence for "(" [-Wprecedence]
]])
AT_CLEANUP
## ---------------------------- ##
## Useless precedence warning. ##
## ---------------------------- ##
AT_SETUP([Useless precedence warning])
AT_DATA([[input.y]],
[[%token A B U V W X Y Z
%precedence Z
%left X
%precedence Y
%left W
%right V
%nonassoc U
%%
a: b
| a U b
| f
;
b: c
| b V c
;
c: d
| c W d
;
d: A
| d X d
| d Y A
;
f: B
| f Z B
;
]])
AT_BISON_CHECK([-Wprecedence -fcaret -o input.c input.y], 0, [],
[[input.y:7.1-9: warning: useless precedence and associativity for U [-Wprecedence]
%nonassoc U
^^^^^^^^^
input.y:6.1-6: warning: useless precedence and associativity for V [-Wprecedence]
%right V
^^^^^^
input.y:5.1-5: warning: useless precedence and associativity for W [-Wprecedence]
%left W
^^^^^
input.y:2.1-11: warning: useless precedence for Z [-Wprecedence]
%precedence Z
^^^^^^^^^^^
]])
AT_CLEANUP
## ---------------- ##
## S/R in initial. ##
## ---------------- ##
# I once hacked Bison in such a way that it lost its reductions on the
# initial state (because it was confusing it with the last state). It
# took me a while to strip down my failures to this simple case. So
# make sure it finds the s/r conflict below.
AT_SETUP([S/R in initial])
AT_DATA([[input.y]],
[[%expect 1
%%
exp: e 'e';
e: 'e' | /* Nothing. */;
]])
AT_BISON_CHECK([-o input.c input.y], 0, [],
[[input.y:4.9: warning: rule useless in parser due to conflicts [-Wother]
]])
AT_BISON_CHECK([-fcaret -o input.c input.y], 0, [],
[[input.y:4.9: warning: rule useless in parser due to conflicts [-Wother]
e: 'e' | /* Nothing. */;
^
]])
AT_CLEANUP
## ------------------- ##
## %nonassoc and eof. ##
## ------------------- ##
AT_SETUP([%nonassoc and eof])
AT_BISON_OPTION_PUSHDEFS
AT_DATA_GRAMMAR([input.y],
[[
%{
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <assert.h>
#define YYERROR_VERBOSE 1
]AT_YYERROR_DEFINE[
/* The current argument. */
static const char *input;
static int
yylex (void)
{
static size_t toknum;
assert (toknum <= strlen (input));
return input[toknum++];
}
%}
%nonassoc '<' '>'
%%
expr: expr '<' expr
| expr '>' expr
| '0'
;
%%
int
main (int argc, const char *argv[])
{
input = argc <= 1 ? "" : argv[1];
return yyparse ();
}
]])
AT_BISON_OPTION_POPDEFS
m4_pushdef([AT_NONASSOC_AND_EOF_CHECK],
[AT_BISON_CHECK([$1[ -o input.c input.y]])
AT_COMPILE([input])
m4_pushdef([AT_EXPECTING], [m4_if($2, [correct], [[, expecting $end]])])
AT_PARSER_CHECK([./input '0<0'])
AT_PARSER_CHECK([./input '0<0<0'], [1], [],
[syntax error, unexpected '<'AT_EXPECTING
])
AT_PARSER_CHECK([./input '0>0'])
AT_PARSER_CHECK([./input '0>0>0'], [1], [],
[syntax error, unexpected '>'AT_EXPECTING
])
AT_PARSER_CHECK([./input '0<0>0'], [1], [],
[syntax error, unexpected '>'AT_EXPECTING
])
m4_popdef([AT_EXPECTING])])
# Expected token list is missing.
AT_NONASSOC_AND_EOF_CHECK([], [[incorrect]])
# We must disable default reductions in inconsistent states in order to
# have an explicit list of all expected tokens.
AT_NONASSOC_AND_EOF_CHECK([[-Dlr.default-reduction=consistent]],
[[correct]])
# lr.default-reduction=consistent happens to work for this test case.
# However, for other grammars, lookahead sets can be merged for
# different left contexts, so it is still possible to have an incorrect
# expected list. Canonical LR is almost a general solution (that is, it
# can fail only when %nonassoc is used), so make sure it gives the same
# result as above.
AT_NONASSOC_AND_EOF_CHECK([[-Dlr.type=canonical-lr]], [[correct]])
# parse.lac=full is a completely general solution that does not require
# any of the above sacrifices. Of course, it does not extend the
# language-recognition power of LALR to (IE)LR, but it does ensure that
# the reported list of expected tokens matches what the given parser
# would have accepted in place of the unexpected token.
AT_NONASSOC_AND_EOF_CHECK([[-Dparse.lac=full]], [[correct]])
m4_popdef([AT_NONASSOC_AND_EOF_CHECK])
AT_CLEANUP
## ------------------------------------------- ##
## parse.error=verbose and consistent errors. ##
## ------------------------------------------- ##
m4_pushdef([AT_CONSISTENT_ERRORS_CHECK], [
AT_SETUP([[parse.error=verbose and consistent errors: ]$1])
AT_BISON_OPTION_PUSHDEFS([$1])
m4_pushdef([AT_YYLEX_PROTOTYPE],
[AT_SKEL_CC_IF([[int yylex (yy::parser::semantic_type *lvalp)]],
[[int yylex (YYSTYPE *lvalp)]])])
AT_SKEL_JAVA_IF([AT_DATA], [AT_DATA_GRAMMAR])([input.y],
[AT_SKEL_JAVA_IF([[
%code imports {
import java.io.IOException;
}]], [[
%code {]AT_SKEL_CC_IF([[
#include <string>]], [[
#include <assert.h>
#include <stdio.h>
]AT_YYERROR_DECLARE])[
]AT_YYLEX_PROTOTYPE[;
#define USE(Var)
}
]AT_SKEL_CC_IF([[%defines]], [[%define api.pure]])])[
]$1[
%define parse.error verbose
%%
]$2[
]AT_SKEL_JAVA_IF([[%code lexer {]], [[%%]])[
/*--------.
| yylex. |
`--------*/]AT_SKEL_JAVA_IF([[
public String input = "]$3[";
public int index = 0;
public int yylex ()
{
if (index < input.length ())
return input.charAt (index++);
else
return 0;
}
public Object getLVal ()
{
return new Integer(1);
}]], [[
]AT_YYLEX_PROTOTYPE[
{
static char const *input = "]$3[";
*lvalp = 1;
return *input++;
}]])[
]AT_YYERROR_DEFINE[
]AT_SKEL_JAVA_IF([[
};
%%]])[
/*-------.
| main. |
`-------*/
]AT_MAIN_DEFINE
])
AT_FULL_COMPILE([[input]])
m4_pushdef([AT_EXPECTING], [m4_if($5, [ab], [[, expecting 'a' or 'b']],
$5, [a], [[, expecting 'a']],
$5, [b], [[, expecting 'b']])])
AT_SKEL_JAVA_IF([AT_JAVA_PARSER_CHECK([[input]], [[0]]],
[AT_PARSER_CHECK([[./input]], [[1]]]),
[[]],
[[syntax error, unexpected ]$4[]AT_EXPECTING[
]])
m4_popdef([AT_EXPECTING])
m4_popdef([AT_YYLEX_PROTOTYPE])
AT_BISON_OPTION_POPDEFS
AT_CLEANUP
]) dnl AT_CONSISTENT_ERRORS_CHECK
m4_pushdef([AT_PREVIOUS_STATE_GRAMMAR],
[[%nonassoc 'a';
start: consistent-error-on-a-a 'a' ;
consistent-error-on-a-a:
'a' default-reduction
| 'a' default-reduction 'a'
| 'a' shift
;
default-reduction: /*empty*/ ;
shift: 'b' ;
// Provide another context in which all rules are useful so that this
// test case looks a little more realistic.
start: 'b' consistent-error-on-a-a 'c' ;
]])
m4_pushdef([AT_PREVIOUS_STATE_INPUT], [[a]])
# Unfortunately, no expected tokens are reported even though 'b' can be
# accepted. Nevertheless, the main point of this test is to make sure
# that at least the unexpected token is reported. In a previous version
# of Bison, it wasn't reported because the error is detected in a
# consistent state with an error action, and that case always triggered
# the simple "syntax error" message.
#
# The point isn't to test IELR here, but state merging happens to
# complicate this example.
AT_CONSISTENT_ERRORS_CHECK([[%define lr.type ielr]],
[AT_PREVIOUS_STATE_GRAMMAR],
[AT_PREVIOUS_STATE_INPUT],
[[$end]], [[none]])
AT_CONSISTENT_ERRORS_CHECK([[%define lr.type ielr
%glr-parser]],
[AT_PREVIOUS_STATE_GRAMMAR],
[AT_PREVIOUS_STATE_INPUT],
[[$end]], [[none]])
AT_CONSISTENT_ERRORS_CHECK([[%define lr.type ielr
%language "c++"]],
[AT_PREVIOUS_STATE_GRAMMAR],
[AT_PREVIOUS_STATE_INPUT],
[[$end]], [[none]])
AT_CONSISTENT_ERRORS_CHECK([[%define lr.type ielr
%language "java"]],
[AT_PREVIOUS_STATE_GRAMMAR],
[AT_PREVIOUS_STATE_INPUT],
[[end of input]], [[none]])
# Even canonical LR doesn't foresee the error for 'a'!
AT_CONSISTENT_ERRORS_CHECK([[%define lr.type ielr
%define lr.default-reduction consistent]],
[AT_PREVIOUS_STATE_GRAMMAR],
[AT_PREVIOUS_STATE_INPUT],
[[$end]], [[ab]])
AT_CONSISTENT_ERRORS_CHECK([[%define lr.type ielr
%define lr.default-reduction accepting]],
[AT_PREVIOUS_STATE_GRAMMAR],
[AT_PREVIOUS_STATE_INPUT],
[[$end]], [[ab]])
AT_CONSISTENT_ERRORS_CHECK([[%define lr.type canonical-lr]],
[AT_PREVIOUS_STATE_GRAMMAR],
[AT_PREVIOUS_STATE_INPUT],
[[$end]], [[ab]])
# Only LAC gets it right.
AT_CONSISTENT_ERRORS_CHECK([[%define lr.type canonical-lr
%define parse.lac full]],
[AT_PREVIOUS_STATE_GRAMMAR],
[AT_PREVIOUS_STATE_INPUT],
[[$end]], [[b]])
AT_CONSISTENT_ERRORS_CHECK([[%define lr.type ielr
%define parse.lac full]],
[AT_PREVIOUS_STATE_GRAMMAR],
[AT_PREVIOUS_STATE_INPUT],
[[$end]], [[b]])
m4_popdef([AT_PREVIOUS_STATE_GRAMMAR])
m4_popdef([AT_PREVIOUS_STATE_INPUT])
m4_pushdef([AT_USER_ACTION_GRAMMAR],
[[%nonassoc 'a';
// If $$ = 0 here, then we know that the 'a' destructor is being invoked
// incorrectly for the 'b' set in the semantic action below. All 'a'
// tokens are returned by yylex, which sets $$ = 1.
%destructor {
if (!$$)
fprintf (stderr, "Wrong destructor.\n");
} 'a';
// Rather than depend on an inconsistent state to induce reading a
// lookahead as in the previous grammar, just assign the lookahead in a
// semantic action. That lookahead isn't needed before either error
// action is encountered. In a previous version of Bison, this was a
// problem as it meant yychar was not translated into yytoken before
// either error action. The second error action thus invoked a
// destructor that it selected according to the incorrect yytoken. The
// first error action would have reported an incorrect unexpected token
// except that, due to the bug described in the previous grammar, the
// unexpected token was not reported at all.
start: error-reduce consistent-error 'a' { USE ($][3); } ;
error-reduce:
'a' 'a' consistent-reduction consistent-error 'a'
{ USE (($][1, $][2, $][5)); }
| 'a' error
{ USE ($][1); }
;
consistent-reduction: /*empty*/ {
assert (yychar == YYEMPTY);
yylval = 0;
yychar = 'b';
} ;
consistent-error:
'a' { USE ($][1); }
| /*empty*/ %prec 'a'
;
// Provide another context in which all rules are useful so that this
// test case looks a little more realistic.
start: 'b' consistent-error 'b' ;
]])
m4_pushdef([AT_USER_ACTION_INPUT], [[aa]])
AT_CONSISTENT_ERRORS_CHECK([[]],
[AT_USER_ACTION_GRAMMAR],
[AT_USER_ACTION_INPUT],
[['b']], [[none]])
AT_CONSISTENT_ERRORS_CHECK([[%glr-parser]],
[AT_USER_ACTION_GRAMMAR],
[AT_USER_ACTION_INPUT],
[['b']], [[none]])
# No C++ or Java test because yychar cannot be manipulated by users.
AT_CONSISTENT_ERRORS_CHECK([[%define lr.default-reduction consistent]],
[AT_USER_ACTION_GRAMMAR],
[AT_USER_ACTION_INPUT],
[['b']], [[none]])
# Canonical LR doesn't foresee the error for 'a'!
AT_CONSISTENT_ERRORS_CHECK([[%define lr.default-reduction accepting]],
[AT_USER_ACTION_GRAMMAR],
[AT_USER_ACTION_INPUT],
[[$end]], [[a]])
AT_CONSISTENT_ERRORS_CHECK([[%define lr.type canonical-lr]],
[AT_USER_ACTION_GRAMMAR],
[AT_USER_ACTION_INPUT],
[[$end]], [[a]])
AT_CONSISTENT_ERRORS_CHECK([[%define parse.lac full]],
[AT_USER_ACTION_GRAMMAR],
[AT_USER_ACTION_INPUT],
[['b']], [[none]])
AT_CONSISTENT_ERRORS_CHECK([[%define parse.lac full
%define lr.default-reduction accepting]],
[AT_USER_ACTION_GRAMMAR],
[AT_USER_ACTION_INPUT],
[[$end]], [[none]])
m4_popdef([AT_USER_ACTION_GRAMMAR])
m4_popdef([AT_USER_ACTION_INPUT])
m4_popdef([AT_CONSISTENT_ERRORS_CHECK])
## ------------------------------------------------------- ##
## LAC: %nonassoc requires splitting canonical LR states. ##
## ------------------------------------------------------- ##
# This test case demonstrates that, when %nonassoc is used, canonical
# LR(1) parser table construction followed by conflict resolution
# without further state splitting is not always sufficient to produce a
# parser that can detect all syntax errors as soon as possible on one
# token of lookahead. However, LAC solves the problem completely even
# with minimal LR parser tables.
AT_SETUP([[LAC: %nonassoc requires splitting canonical LR states]])
AT_BISON_OPTION_PUSHDEFS
AT_DATA_GRAMMAR([[input.y]],
[[%code {
#include <stdio.h>
]AT_YYERROR_DECLARE[
]AT_YYLEX_DECLARE[
}
%error-verbose
%nonassoc 'a'
%%
start:
'a' problem 'a' // First context.
| 'b' problem 'b' // Second context.
| 'c' reduce-nonassoc // Just makes reduce-nonassoc useful.
;
problem:
look reduce-nonassoc
| look 'a'
| look 'b'
;
// For the state reached after shifting the 'a' in these productions,
// lookahead sets are the same in both the first and second contexts.
// Thus, canonical LR reuses the same state for both contexts. However,
// the lookahead 'a' for the reduction "look: 'a'" later becomes an
// error action only in the first context. In order to immediately
// detect the syntax error on 'a' here for only the first context, this
// canonical LR state would have to be split into two states, and the
// 'a' lookahead would have to be removed from only one of the states.
look:
'a' // Reduction lookahead set is always ['a', 'b'].
| 'a' 'b'
| 'a' 'c' // 'c' is forgotten as an expected token.
;
reduce-nonassoc: %prec 'a';
%%
]AT_YYERROR_DEFINE[
]AT_YYLEX_DEFINE(["aaa"])[
]AT_MAIN_DEFINE
])
AT_BISON_OPTION_POPDEFS
# Show canonical LR's failure.
AT_BISON_CHECK([[-Dlr.type=canonical-lr -o input.c input.y]],
[[0]], [[]],
[[input.y: warning: 2 shift/reduce conflicts [-Wconflicts-sr]
]])
AT_COMPILE([[input]])
AT_PARSER_CHECK([[./input]], [[1]], [[]],
[[syntax error, unexpected 'a', expecting 'b'
]])
# It's corrected by LAC.
AT_BISON_CHECK([[-Dlr.type=canonical-lr -Dparse.lac=full \
-o input.c input.y]], [[0]], [[]],
[[input.y: warning: 2 shift/reduce conflicts [-Wconflicts-sr]
]])
AT_COMPILE([[input]])
AT_PARSER_CHECK([[./input]], [[1]], [[]],
[[syntax error, unexpected 'a', expecting 'b' or 'c'
]])
# IELR is sufficient when LAC is used.
AT_BISON_CHECK([[-Dlr.type=ielr -Dparse.lac=full -o input.c input.y]],
[[0]], [[]],
[[input.y: warning: 2 shift/reduce conflicts [-Wconflicts-sr]
]])
AT_COMPILE([[input]])
AT_PARSER_CHECK([[./input]], [[1]], [[]],
[[syntax error, unexpected 'a', expecting 'b' or 'c'
]])
AT_CLEANUP
## ------------------------- ##
## Unresolved SR Conflicts. ##
## ------------------------- ##
AT_SETUP([Unresolved SR Conflicts])
AT_KEYWORDS([report])
AT_DATA([input.y],
[[%token NUM OP
%%
exp: exp OP exp | NUM;
]])
AT_BISON_CHECK([-o input.c --report=all input.y], 0, [],
[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
]])
# Check the contents of the report.
AT_CHECK([cat input.output], [],
[[State 5 conflicts: 1 shift/reduce
Grammar
0 $accept: exp $end
1 exp: exp OP exp
2 | NUM
Terminals, with rules where they appear
$end (0) 0
error (256)
NUM (258) 2
OP (259) 1
Nonterminals, with rules where they appear
$accept (5)
on left: 0
exp (6)
on left: 1 2, on right: 0 1
State 0
0 $accept: . exp $end
1 exp: . exp OP exp
2 | . NUM
NUM shift, and go to state 1
exp go to state 2
State 1
2 exp: NUM .
$default reduce using rule 2 (exp)
State 2
0 $accept: exp . $end
1 exp: exp . OP exp
$end shift, and go to state 3
OP shift, and go to state 4
State 3
0 $accept: exp $end .
$default accept
State 4
1 exp: . exp OP exp
1 | exp OP . exp
2 | . NUM
NUM shift, and go to state 1
exp go to state 5
State 5
1 exp: exp . OP exp
1 | exp OP exp . [$end, OP]
OP shift, and go to state 4
OP [reduce using rule 1 (exp)]
$default reduce using rule 1 (exp)
]])
AT_CLEANUP
## ----------------------- ##
## Resolved SR Conflicts. ##
## ----------------------- ##
AT_SETUP([Resolved SR Conflicts])
AT_KEYWORDS([report])
AT_DATA([input.y],
[[%token NUM OP
%left OP
%%
exp: exp OP exp | NUM;
]])
AT_BISON_CHECK([-o input.c --report=all input.y])
# Check the contents of the report.
AT_CHECK([cat input.output], [],
[[Grammar
0 $accept: exp $end
1 exp: exp OP exp
2 | NUM
Terminals, with rules where they appear
$end (0) 0
error (256)
NUM (258) 2
OP (259) 1
Nonterminals, with rules where they appear
$accept (5)
on left: 0
exp (6)
on left: 1 2, on right: 0 1
State 0
0 $accept: . exp $end
1 exp: . exp OP exp
2 | . NUM
NUM shift, and go to state 1
exp go to state 2
State 1
2 exp: NUM .
$default reduce using rule 2 (exp)
State 2
0 $accept: exp . $end
1 exp: exp . OP exp
$end shift, and go to state 3
OP shift, and go to state 4
State 3
0 $accept: exp $end .
$default accept
State 4
1 exp: . exp OP exp
1 | exp OP . exp
2 | . NUM
NUM shift, and go to state 1
exp go to state 5
State 5
1 exp: exp . OP exp
1 | exp OP exp . [$end, OP]
$default reduce using rule 1 (exp)
Conflict between rule 1 and token OP resolved as reduce (%left OP).
]])
AT_CLEANUP
## ---------------------- ##
## %precedence suffices. ##
## ---------------------- ##
AT_SETUP([%precedence suffices])
AT_DATA([input.y],
[[%precedence "then"
%precedence "else"
%%
stmt:
"if" cond "then" stmt
| "if" cond "then" stmt "else" stmt
| "stmt"
;
cond:
"exp"
;
]])
AT_BISON_CHECK([-o input.c input.y])
AT_CLEANUP
## ------------------------------ ##
## %precedence does not suffice. ##
## ------------------------------ ##
AT_SETUP([%precedence does not suffice])
AT_DATA([input.y],
[[%precedence "then"
%precedence "else"
%%
stmt:
"if" cond "then" stmt
| "if" cond "then" stmt "else" stmt
| "stmt"
;
cond:
"exp"
| cond "then" cond
;
]])
AT_BISON_CHECK([-o input.c input.y], 0, [],
[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
input.y:12.3-18: warning: rule useless in parser due to conflicts [-Wother]
]])
AT_CLEANUP
## -------------------------------- ##
## Defaulted Conflicted Reduction. ##
## -------------------------------- ##
# When there are RR conflicts, some rules are disabled. Usually it is
# simply displayed as:
#
# $end reduce using rule 3 (num)
# $end [reduce using rule 4 (id)]
#
# But when 'reduce 3' is the default action, we'd produce:
#
# $end [reduce using rule 4 (id)]
# $default reduce using rule 3 (num)
#
# In this precise case (a reduction is masked by the default
# reduction), we make the 'reduce 3' explicit:
#
# $end reduce using rule 3 (num)
# $end [reduce using rule 4 (id)]
# $default reduce using rule 3 (num)
#
# Maybe that's not the best display, but then, please propose something
# else.
AT_SETUP([Defaulted Conflicted Reduction])
AT_KEYWORDS([report])
AT_DATA([input.y],
[[%%
exp: num | id;
num: '0';
id : '0';
%%
]])
AT_BISON_CHECK([-o input.c --report=all input.y], 0, [],
[[input.y: warning: 1 reduce/reduce conflict [-Wconflicts-rr]
input.y:4.6-8: warning: rule useless in parser due to conflicts [-Wother]
]])
# Check the contents of the report.
AT_CHECK([cat input.output], [],
[[Rules useless in parser due to conflicts
4 id: '0'
State 1 conflicts: 1 reduce/reduce
Grammar
0 $accept: exp $end
1 exp: num
2 | id
3 num: '0'
4 id: '0'
Terminals, with rules where they appear
$end (0) 0
'0' (48) 3 4
error (256)
Nonterminals, with rules where they appear
$accept (4)
on left: 0
exp (5)
on left: 1 2, on right: 0
num (6)
on left: 3, on right: 1
id (7)
on left: 4, on right: 2
State 0
0 $accept: . exp $end
1 exp: . num
2 | . id
3 num: . '0'
4 id: . '0'
'0' shift, and go to state 1
exp go to state 2
num go to state 3
id go to state 4
State 1
3 num: '0' . [$end]
4 id: '0' . [$end]
$end reduce using rule 3 (num)
$end [reduce using rule 4 (id)]
$default reduce using rule 3 (num)
State 2
0 $accept: exp . $end
$end shift, and go to state 5
State 3
1 exp: num .
$default reduce using rule 1 (exp)
State 4
2 exp: id .
$default reduce using rule 2 (exp)
State 5
0 $accept: exp $end .
$default accept
]])
AT_CLEANUP
## -------------------- ##
## %expect not enough. ##
## -------------------- ##
AT_SETUP([%expect not enough])
AT_DATA([input.y],
[[%token NUM OP
%expect 0
%%
exp: exp OP exp | NUM;
]])
AT_BISON_CHECK([-o input.c input.y], 1, [],
[[input.y: error: shift/reduce conflicts: 1 found, 0 expected
]])
AT_CLEANUP
## --------------- ##
## %expect right. ##
## --------------- ##
AT_SETUP([%expect right])
AT_DATA([input.y],
[[%token NUM OP
%expect 1
%%
exp: exp OP exp | NUM;
]])
AT_BISON_CHECK([-o input.c input.y])
AT_CLEANUP
## ------------------ ##
## %expect too much. ##
## ------------------ ##
AT_SETUP([%expect too much])
AT_DATA([input.y],
[[%token NUM OP
%expect 2
%%
exp: exp OP exp | NUM;
]])
AT_BISON_CHECK([-o input.c input.y], 1, [],
[[input.y: error: shift/reduce conflicts: 1 found, 2 expected
]])
AT_CLEANUP
## ------------------------------- ##
## %expect with reduce conflicts. ##
## ------------------------------- ##
AT_SETUP([%expect with reduce conflicts])
AT_DATA([input.y],
[[%expect 0
%%
program: a 'a' | a a;
a: 'a';
]])
AT_BISON_CHECK([-o input.c input.y], 1, [],
[[input.y: error: reduce/reduce conflicts: 1 found, 0 expected
]])
AT_CLEANUP
## ------------------------- ##
## %prec with user strings. ##
## ------------------------- ##
AT_SETUP([%prec with user string])
AT_DATA([[input.y]],
[[%%
exp:
"foo" %prec "foo"
;
]])
AT_BISON_CHECK([-o input.c input.y])
AT_CLEANUP
## -------------------------------- ##
## %no-default-prec without %prec. ##
## -------------------------------- ##
AT_SETUP([%no-default-prec without %prec])
AT_DATA([[input.y]],
[[%left '+'
%left '*'
%%
%no-default-prec;
e: e '+' e
| e '*' e
| '0'
;
]])
AT_BISON_CHECK([-Wall -o input.c input.y], 0, [],
[[input.y: warning: 4 shift/reduce conflicts [-Wconflicts-sr]
input.y:1.1-5: warning: useless precedence and associativity for '+' [-Wprecedence]
input.y:2.1-5: warning: useless precedence and associativity for '*' [-Wprecedence]
]])
AT_CLEANUP
## ----------------------------- ##
## %no-default-prec with %prec. ##
## ----------------------------- ##
AT_SETUP([%no-default-prec with %prec])
AT_DATA([[input.y]],
[[%left '+'
%left '*'
%%
%no-default-prec;
e: e '+' e %prec '+'
| e '*' e %prec '*'
| '0'
;
]])
AT_BISON_CHECK([-o input.c input.y])
AT_CLEANUP
## --------------- ##
## %default-prec. ##
## --------------- ##
AT_SETUP([%default-prec])
AT_DATA([[input.y]],
[[%left '+'
%left '*'
%%
%default-prec;
e: e '+' e
| e '*' e
| '0'
;
]])
AT_BISON_CHECK([-o input.c input.y])
AT_CLEANUP
## ---------------------------------------------- ##
## Unreachable States After Conflict Resolution. ##
## ---------------------------------------------- ##
AT_SETUP([[Unreachable States After Conflict Resolution]])
# If conflict resolution makes states unreachable, remove those states, report
# rules that are then unused, and don't report conflicts in those states. Test
# what happens when a nonterminal becomes useless as a result of state removal
# since that causes lalr.o's goto map to be rewritten.
AT_DATA([[input.y]],
[[%output "input.c"
%left 'a'
%%
start: resolved_conflict 'a' reported_conflicts 'a' ;
/* S/R conflict resolved as reduce, so the state with item
* (resolved_conflict: 'a' . unreachable1) and all it transition successors are
* unreachable, and the associated production is useless. */
resolved_conflict:
'a' unreachable1
| %prec 'a'
;
/* S/R conflict that need not be reported since it is unreachable because of
* the previous conflict resolution. Nonterminal unreachable1 and all its
* productions are useless. */
unreachable1:
'a' unreachable2
|
;
/* Likewise for a R/R conflict and nonterminal unreachable2. */
unreachable2: | ;
/* Make sure remaining S/R and R/R conflicts are still reported correctly even
* when their states are renumbered due to state removal. */
reported_conflicts:
'a'
| 'a'
|
;
]])
AT_BISON_CHECK([[--report=all input.y]], 0, [],
[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
input.y: warning: 1 reduce/reduce conflict [-Wconflicts-rr]
input.y:12.5-20: warning: rule useless in parser due to conflicts [-Wother]
input.y:20.5-20: warning: rule useless in parser due to conflicts [-Wother]
input.y:21.4: warning: rule useless in parser due to conflicts [-Wother]
input.y:25.13: warning: rule useless in parser due to conflicts [-Wother]
input.y:25.16: warning: rule useless in parser due to conflicts [-Wother]
input.y:31.5-7: warning: rule useless in parser due to conflicts [-Wother]
input.y:32.4: warning: rule useless in parser due to conflicts [-Wother]
]])
AT_CHECK([[cat input.output]], 0,
[[Rules useless in parser due to conflicts
2 resolved_conflict: 'a' unreachable1
4 unreachable1: 'a' unreachable2
5 | %empty
6 unreachable2: %empty
7 | %empty
9 reported_conflicts: 'a'
10 | %empty
State 4 conflicts: 1 shift/reduce
State 5 conflicts: 1 reduce/reduce
Grammar
0 $accept: start $end
1 start: resolved_conflict 'a' reported_conflicts 'a'
2 resolved_conflict: 'a' unreachable1
3 | %empty
4 unreachable1: 'a' unreachable2
5 | %empty
6 unreachable2: %empty
7 | %empty
8 reported_conflicts: 'a'
9 | 'a'
10 | %empty
Terminals, with rules where they appear
$end (0) 0
'a' (97) 1 2 4 8 9
error (256)
Nonterminals, with rules where they appear
$accept (4)
on left: 0
start (5)
on left: 1, on right: 0
resolved_conflict (6)
on left: 2 3, on right: 1
unreachable1 (7)
on left: 4 5, on right: 2
unreachable2 (8)
on left: 6 7, on right: 4
reported_conflicts (9)
on left: 8 9 10, on right: 1
State 0
0 $accept: . start $end
1 start: . resolved_conflict 'a' reported_conflicts 'a'
2 resolved_conflict: . 'a' unreachable1
3 | . %empty ['a']
$default reduce using rule 3 (resolved_conflict)
start go to state 1
resolved_conflict go to state 2
Conflict between rule 3 and token 'a' resolved as reduce (%left 'a').
State 1
0 $accept: start . $end
$end shift, and go to state 3
State 2
1 start: resolved_conflict . 'a' reported_conflicts 'a'
'a' shift, and go to state 4
State 3
0 $accept: start $end .
$default accept
State 4
1 start: resolved_conflict 'a' . reported_conflicts 'a'
8 reported_conflicts: . 'a'
9 | . 'a'
10 | . %empty ['a']
'a' shift, and go to state 5
'a' [reduce using rule 10 (reported_conflicts)]
reported_conflicts go to state 6
State 5
8 reported_conflicts: 'a' . ['a']
9 | 'a' . ['a']
'a' reduce using rule 8 (reported_conflicts)
'a' [reduce using rule 9 (reported_conflicts)]
$default reduce using rule 8 (reported_conflicts)
State 6
1 start: resolved_conflict 'a' reported_conflicts . 'a'
'a' shift, and go to state 7
State 7
1 start: resolved_conflict 'a' reported_conflicts 'a' .
$default reduce using rule 1 (start)
]])
AT_DATA([[input-keep.y]],
[[%define lr.keep-unreachable-state
]])
AT_CHECK([[cat input.y >> input-keep.y]])
AT_BISON_CHECK([[input-keep.y]], 0, [],
[[input-keep.y: warning: 2 shift/reduce conflicts [-Wconflicts-sr]
input-keep.y: warning: 2 reduce/reduce conflicts [-Wconflicts-rr]
input-keep.y:22.4: warning: rule useless in parser due to conflicts [-Wother]
input-keep.y:26.16: warning: rule useless in parser due to conflicts [-Wother]
input-keep.y:32.5-7: warning: rule useless in parser due to conflicts [-Wother]
input-keep.y:33.4: warning: rule useless in parser due to conflicts [-Wother]
]])
AT_CLEANUP
## ------------------------------------------------------------ ##
## Solved conflicts report for multiple reductions in a state. ##
## ------------------------------------------------------------ ##
AT_SETUP([[Solved conflicts report for multiple reductions in a state]])
# Used to lose earlier solved conflict messages even within a single S/R/R.
AT_DATA([[input.y]],
[[%left 'a'
%right 'b'
%right 'c'
%right 'd'
%%
start:
'a'
| empty_a 'a'
| 'b'
| empty_b 'b'
| 'c'
| empty_c1 'c'
| empty_c2 'c'
| empty_c3 'c'
;
empty_a: %prec 'a' ;
empty_b: %prec 'b' ;
empty_c1: %prec 'c' ;
empty_c2: %prec 'c' ;
empty_c3: %prec 'd' ;
]])
AT_BISON_CHECK([[--report=all -o input.c input.y]], 0, [], [ignore])
AT_CHECK([[cat input.output | sed -n '/^State 0$/,/^State 1$/p']], 0,
[[State 0
0 $accept: . start $end
1 start: . 'a'
2 | . empty_a 'a'
3 | . 'b'
4 | . empty_b 'b'
5 | . 'c'
6 | . empty_c1 'c'
7 | . empty_c2 'c'
8 | . empty_c3 'c'
9 empty_a: . %empty ['a']
10 empty_b: . %empty []
11 empty_c1: . %empty []
12 empty_c2: . %empty []
13 empty_c3: . %empty ['c']
'b' shift, and go to state 1
'c' reduce using rule 13 (empty_c3)
$default reduce using rule 9 (empty_a)
start go to state 2
empty_a go to state 3
empty_b go to state 4
empty_c1 go to state 5
empty_c2 go to state 6
empty_c3 go to state 7
Conflict between rule 9 and token 'a' resolved as reduce (%left 'a').
Conflict between rule 10 and token 'b' resolved as shift (%right 'b').
Conflict between rule 11 and token 'c' resolved as shift (%right 'c').
Conflict between rule 12 and token 'c' resolved as shift (%right 'c').
Conflict between rule 13 and token 'c' resolved as reduce ('c' < 'd').
State 1
]])
AT_CLEANUP
## ------------------------------------------------------------ ##
## %nonassoc error actions for multiple reductions in a state. ##
## ------------------------------------------------------------ ##
# Used to abort when trying to resolve conflicts as %nonassoc error actions for
# multiple reductions in a state.
# For a %nonassoc error action token, used to print the first remaining
# reduction on that token without brackets.
AT_SETUP([[%nonassoc error actions for multiple reductions in a state]])
AT_DATA([[input.y]],
[[%nonassoc 'a' 'b' 'c'
%%
start:
'a'
| empty_a 'a'
| 'b'
| empty_b 'b'
| 'c'
| empty_c1 'c'
| empty_c2 'c'
| empty_c3 'c'
;
empty_a: %prec 'a' ;
empty_b: %prec 'b' ;
empty_c1: %prec 'c' ;
empty_c2: %prec 'c' ;
empty_c3: %prec 'c' ;
]])
AT_BISON_CHECK([[--report=all -o input.c input.y]], 0, [], [ignore])
AT_CHECK([[cat input.output | sed -n '/^State 0$/,/^State 1$/p']], 0,
[[State 0
0 $accept: . start $end
1 start: . 'a'
2 | . empty_a 'a'
3 | . 'b'
4 | . empty_b 'b'
5 | . 'c'
6 | . empty_c1 'c'
7 | . empty_c2 'c'
8 | . empty_c3 'c'
9 empty_a: . %empty []
10 empty_b: . %empty []
11 empty_c1: . %empty []
12 empty_c2: . %empty ['c']
13 empty_c3: . %empty ['c']
'a' error (nonassociative)
'b' error (nonassociative)
'c' error (nonassociative)
'c' [reduce using rule 12 (empty_c2)]
'c' [reduce using rule 13 (empty_c3)]
start go to state 1
empty_a go to state 2
empty_b go to state 3
empty_c1 go to state 4
empty_c2 go to state 5
empty_c3 go to state 6
Conflict between rule 9 and token 'a' resolved as an error (%nonassoc 'a').
Conflict between rule 10 and token 'b' resolved as an error (%nonassoc 'b').
Conflict between rule 11 and token 'c' resolved as an error (%nonassoc 'c').
State 1
]])
AT_CLEANUP
## -------------------- ##
## %expect-rr non GLR. ##
## -------------------- ##
AT_SETUP([[%expect-rr non GLR]])
AT_DATA([[1.y]],
[[%expect-rr 0
%%
exp: 'a'
]])
AT_BISON_CHECK([[1.y]], [[0]], [],
[[1.y: warning: %expect-rr applies only to GLR parsers [-Wother]
]])
AT_DATA([[2.y]],
[[%expect-rr 1
%%
exp: 'a' | 'a';
]])
AT_BISON_CHECK([[2.y]], [[0]], [],
[[2.y: warning: %expect-rr applies only to GLR parsers [-Wother]
2.y: warning: 1 reduce/reduce conflict [-Wconflicts-rr]
2.y:3.12-14: warning: rule useless in parser due to conflicts [-Wother]
]])
AT_CLEANUP
## ---------------------------------- ##
## -W versus %expect and %expect-rr. ##
## ---------------------------------- ##
AT_SETUP([[-W versus %expect and %expect-rr]])
AT_DATA([[sr-rr.y]],
[[%glr-parser
%%
start: 'a' | A 'a' | B 'a' ;
A: ;
B: ;
]])
AT_DATA([[sr.y]],
[[%glr-parser
%%
start: 'a' | A 'a' ;
A: ;
]])
AT_DATA([[rr.y]],
[[%glr-parser
%%
start: A | B ;
A: ;
B: ;
]])
AT_BISON_CHECK([[sr-rr.y]], [[0]], [[]],
[[sr-rr.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
sr-rr.y: warning: 1 reduce/reduce conflict [-Wconflicts-rr]
]])
AT_BISON_CHECK([[-Wno-conflicts-sr sr-rr.y]], [[0]], [[]],
[[sr-rr.y: warning: 1 reduce/reduce conflict [-Wconflicts-rr]
]])
AT_BISON_CHECK([[-Wno-conflicts-rr sr-rr.y]], [[0]], [[]],
[[sr-rr.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
]])
[
# This is piece of code is rather complex for a simple task: try every
# combinaison of (0 or 1 real SR) x (0 or 1 real RR) x (don't %expect
# or %expect 0, 1, or 2 SR) x (don't %expect-rr or %expect-rr 0, 1, or 2
# RR).
# Number and types of genuine conflicts in the grammar.
for gram in sr-rr sr rr; do
# Number of expected s/r conflicts.
for sr_exp_i in '' 0 1 2; do
# Number of expected r/r conflicts.
for rr_exp_i in '' 0 1 2; do
test -z "$sr_exp_i" && test -z "$rr_exp_i" && continue
# Build grammar file.
sr_exp=0
rr_exp=0
file=$gram
directives=
if test -n "$sr_exp_i"; then
sr_exp=$sr_exp_i
file=$file-expect-$sr_exp
directives="%expect $sr_exp"
fi
if test -n "$rr_exp_i"; then
rr_exp=$rr_exp_i
file=$file-expect-rr-$rr_exp
directives="$directives %expect-rr $rr_exp"
fi
file=$file.y
echo "$directives" > $file
cat $gram.y >> $file
# Number of found conflicts.
case $gram in
(sr) sr_count=1; rr_count=0;;
(rr) sr_count=0; rr_count=1;;
(sr-rr) sr_count=1; rr_count=1;;
esac
# Update number of expected conflicts: if %expect is given then
# %expect-rr defaults to 0, and vice-versa. Leave empty if
# nothing expected.
case $sr_exp_i:$rr_exp_i in
?:) rr_exp_i=0;;
:?) sr_exp_i=0;;
esac
# Run tests.
if test $sr_count -eq $sr_exp && test $rr_count -eq $rr_exp; then
]AT_BISON_CHECK([[-Wnone $file]])[
]AT_BISON_CHECK([[-Werror $file]])[
else
{
if test -z "$sr_exp_i" && test "$sr_count" -ne 0; then
echo "warning: $sr_count shift/reduce conflicts"
elif test "$sr_exp_i" -ne "$sr_count"; then
echo "error: shift/reduce conflicts: $sr_count found, $sr_exp_i expected"
fi
if test -z "$rr_exp_i" && test "$rr_count" -ne 0; then
echo "warning: $rr_count reduce/reduce conflicts"
elif test "$rr_exp_i" -ne "$rr_count"; then
echo "error: reduce/reduce conflicts: $rr_count found, $rr_exp_i expected"
fi
} | sed -e "s/^/$file: /" > experr
]AT_BISON_CHECK([[-Wnone $file]], [[1]], [[]], [[experr]])[
]AT_BISON_CHECK([[-Werror $file]], [[1]], [[]], [[experr]])[
fi
done
done
done]
AT_CLEANUP