# Bison Regressions. -*- Autotest -*-
-# Copyright 2001 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 2, or (at your option)
-# any later version.
+# Copyright (C) 2001-2013 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, write to the Free Software
-# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA
-# 02111-1307, USA.
+# along with this program. If not, see <http://www.gnu.org/licenses/>.
AT_BANNER([[Regression tests.]])
+
## ------------------ ##
-## Duplicate string. ##
+## Trivial grammars. ##
## ------------------ ##
+AT_SETUP([Trivial grammars])
-AT_SETUP([Duplicate string])
+AT_BISON_OPTION_PUSHDEFS
+AT_DATA_GRAMMAR([input.y],
+[[%{
+]AT_YYERROR_DECLARE_EXTERN[
+]AT_YYLEX_DECLARE_EXTERN[
+#define YYSTYPE int *
+%}
-AT_DATA([duplicate.y],
-[[/* `Bison -v' used to dump core when two tokens are defined with the same
- string, as LE and GE below. */
-
-%token NUM
-%token LE "<="
-%token GE "<="
+%error-verbose
%%
-exp: '(' exp ')' | NUM ;
-%%
+
+program: 'x';
]])
+AT_BISON_OPTION_POPDEFS
-AT_CHECK([bison -v duplicate.y -o duplicate.c], 0, ignore, ignore)
+AT_BISON_CHECK([-o input.c input.y])
+AT_COMPILE([input.o])
+AT_COMPILE([input.o], [-DYYDEBUG -c input.c])
AT_CLEANUP
-## ------------------------- ##
-## Unresolved SR Conflicts. ##
-## ------------------------- ##
-AT_SETUP([Unresolved SR Conflicts])
+## ----------------- ##
+## YYSTYPE typedef. ##
+## ----------------- ##
+
+AT_SETUP([YYSTYPE typedef])
+
+AT_BISON_OPTION_PUSHDEFS
+AT_DATA_GRAMMAR([input.y],
+[[%{
+]AT_YYERROR_DECLARE_EXTERN[
+]AT_YYLEX_DECLARE_EXTERN[
+typedef union { char const *val; } YYSTYPE;
+%}
+
+%type <val> program
-AT_DATA([input.y],
-[[%token NUM OP
%%
-exp: exp OP exp | NUM;
+
+program: { $$ = ""; };
]])
+AT_BISON_OPTION_POPDEFS
-AT_CHECK([bison input.y -o input.c -v], 0, [],
-[input.y contains 1 shift/reduce conflict.
-])
+AT_BISON_CHECK([-o input.c input.y])
+AT_COMPILE([input.o])
-# Check the contents of the report.
-AT_CHECK([cat input.output], [],
-[[State 4 contains 1 shift/reduce conflict.
+AT_CLEANUP
-Grammar
- Number, Line, Rule
- 1 3 exp -> exp OP exp
- 2 3 exp -> NUM
+## ------------------------------------- ##
+## Early token definitions with --yacc. ##
+## ------------------------------------- ##
-Terminals, with rules where they appear
+AT_SETUP([Early token definitions with --yacc])
-$ (-1)
-error (256)
-NUM (257) 2
-OP (258) 1
+# Found in GCJ: they expect the tokens to be defined before the user
+# prologue, so that they can use the token definitions in it.
+AT_BISON_OPTION_PUSHDEFS
+AT_DATA_GRAMMAR([input.y],
+[[%{
+]AT_YYERROR_DECLARE_EXTERN[
+]AT_YYLEX_DECLARE_EXTERN[
+%}
+
+%union
+{
+ int val;
+};
+%{
+#ifndef MY_TOKEN
+# error "MY_TOKEN not defined."
+#endif
+%}
+%token MY_TOKEN
+%%
+exp: MY_TOKEN;
+%%
+]])
+AT_BISON_OPTION_POPDEFS
+
+AT_BISON_CHECK([-y -o input.c input.y])
+AT_COMPILE([input.o])
+
+AT_CLEANUP
-Nonterminals, with rules where they appear
-exp (5)
- on left: 1 2, on right: 1
+## ---------------------------------------- ##
+## Early token definitions without --yacc. ##
+## ---------------------------------------- ##
-state 0
- NUM shift, and go to state 1
+AT_SETUP([Early token definitions without --yacc])
- exp go to state 2
+# Found in GCJ: they expect the tokens to be defined before the user
+# prologue, so that they can use the token definitions in it.
+AT_BISON_OPTION_PUSHDEFS
+AT_DATA_GRAMMAR([input.y],
+[[%{
+]AT_YYERROR_DECLARE_EXTERN[
+]AT_YYLEX_DECLARE_EXTERN[
+void print_my_token (void);
+%}
+%union
+{
+ int val;
+};
+%{
+#include <stdio.h>
+void
+print_my_token (void)
+{
+ enum yytokentype my_token = MY_TOKEN;
+ printf ("%d\n", my_token);
+}
+%}
+%token MY_TOKEN
+%%
+exp: MY_TOKEN;
+%%
+]])
+AT_BISON_OPTION_POPDEFS
+
+AT_BISON_CHECK([-o input.c input.y])
+AT_COMPILE([input.o])
-state 1
+AT_CLEANUP
+
+
+
+## ---------------- ##
+## Braces parsing. ##
+## ---------------- ##
- exp -> NUM . (rule 2)
- $default reduce using rule 2 (exp)
+AT_SETUP([Braces parsing])
+AT_BISON_OPTION_PUSHDEFS
+AT_DATA([input.y],
+[[/* Bison used to swallow the character after '}'. */
+%%
+exp: { tests = {{{{{{{{{{}}}}}}}}}}; };
+%%
+]])
+AT_BISON_OPTION_POPDEFS
-state 2
+AT_BISON_CHECK([-v -o input.c input.y])
- exp -> exp . OP exp (rule 1)
+AT_CHECK([grep 'tests = {{{{{{{{{{}}}}}}}}}};' input.c], 0, [ignore])
- $ shift, and go to state 5
- OP shift, and go to state 3
+AT_CLEANUP
+## ------------------ ##
+## Duplicate string. ##
+## ------------------ ##
-state 3
- exp -> exp OP . exp (rule 1)
+AT_SETUP([Duplicate string])
- NUM shift, and go to state 1
+AT_BISON_OPTION_PUSHDEFS
+AT_DATA([input.y],
+[[/* 'Bison -v' used to dump core when two tokens are defined with the same
+ string, as LE and GE below. */
- exp go to state 4
+%token NUM
+%token LE "<="
+%token GE "<="
+%%
+exp: '(' exp ')' | NUM ;
+%%
+]])
+AT_BISON_OPTION_POPDEFS
+AT_BISON_CHECK([-v -o input.c input.y], 0, [],
+[[input.y:6.8-14: warning: symbol "<=" used more than once as a literal string [-Wother]
+]])
-state 4
+AT_CLEANUP
- exp -> exp . OP exp (rule 1)
- exp -> exp OP exp . (rule 1)
- OP shift, and go to state 3
+## ------------------- ##
+## Rule Line Numbers. ##
+## ------------------- ##
- OP [reduce using rule 1 (exp)]
- $default reduce using rule 1 (exp)
+AT_SETUP([Rule Line Numbers])
+AT_KEYWORDS([report])
+AT_BISON_OPTION_PUSHDEFS
+AT_DATA([input.y],
+[[%%
+expr:
+'a'
-state 5
+{
- $ shift, and go to state 6
+}
+'b'
+{
-state 6
+}
- $default accept
+|
-]])
+{
-AT_CLEANUP
+}
-## --------------------- ##
-## Solved SR Conflicts. ##
-## --------------------- ##
+'c'
-AT_SETUP([Solved SR Conflicts])
+{
-AT_DATA([input.y],
-[[%token NUM OP
-%right OP
-%%
-exp: exp OP exp | NUM;
+};
]])
+AT_BISON_OPTION_POPDEFS
-AT_CHECK([bison input.y -o input.c -v], 0, [], [])
+AT_BISON_CHECK([-o input.c -v input.y])
# Check the contents of the report.
AT_CHECK([cat input.output], [],
-[[Conflict in state 4 between rule 1 and token OP resolved as shift.
+[[Grammar
+
+ 0 $accept: expr $end
+ 1 $@1: %empty
-Grammar
+ 2 expr: 'a' $@1 'b'
- Number, Line, Rule
- 1 4 exp -> exp OP exp
- 2 4 exp -> NUM
+ 3 $@2: %empty
+
+ 4 expr: $@2 'c'
Terminals, with rules where they appear
-$ (-1)
+$end (0) 0
+'a' (97) 2
+'b' (98) 2
+'c' (99) 4
error (256)
-NUM (257) 2
-OP (258) 1
Nonterminals, with rules where they appear
-exp (5)
- on left: 1 2, on right: 1
+$accept (6)
+ on left: 0
+expr (7)
+ on left: 2 4, on right: 0
+$@1 (8)
+ on left: 1, on right: 2
+$@2 (9)
+ on left: 3, on right: 4
+
+State 0
-state 0
+ 0 $accept: . expr $end
- NUM shift, and go to state 1
+ 'a' shift, and go to state 1
- exp go to state 2
+ $default reduce using rule 3 ($@2)
+ expr go to state 2
+ $@2 go to state 3
-state 1
+State 1
- exp -> NUM . (rule 2)
+ 2 expr: 'a' . $@1 'b'
- $default reduce using rule 2 (exp)
+ $default reduce using rule 1 ($@1)
+ $@1 go to state 4
-state 2
+State 2
- exp -> exp . OP exp (rule 1)
+ 0 $accept: expr . $end
- $ shift, and go to state 5
- OP shift, and go to state 3
+ $end shift, and go to state 5
+State 3
-state 3
+ 4 expr: $@2 . 'c'
- exp -> exp OP . exp (rule 1)
+ 'c' shift, and go to state 6
- NUM shift, and go to state 1
- exp go to state 4
+State 4
+ 2 expr: 'a' $@1 . 'b'
+ 'b' shift, and go to state 7
-state 4
- exp -> exp . OP exp (rule 1)
- exp -> exp OP exp . (rule 1)
+State 5
- OP shift, and go to state 3
+ 0 $accept: expr $end .
- $default reduce using rule 1 (exp)
+ $default accept
+State 6
-state 5
+ 4 expr: $@2 'c' .
- $ shift, and go to state 6
+ $default reduce using rule 4 (expr)
+State 7
-state 6
+ 2 expr: 'a' $@1 'b' .
- $default accept
+ $default reduce using rule 2 (expr)
+]])
+AT_CLEANUP
+
+
+## ---------------------- ##
+## Mixing %token styles. ##
+## ---------------------- ##
+
+
+AT_SETUP([Mixing %token styles])
+
+# Taken from the documentation.
+AT_DATA([input.y],
+[[%token <operator> OR "||"
+%token <operator> LE 134 "<="
+%left OR "<="
+%%
+exp: %empty;
+%%
+]])
+
+AT_BISON_CHECK([-v -Wall -o input.c input.y], 0, [],
+[[input.y:1.29-32: warning: useless precedence and associativity for "||" [-Wprecedence]
+input.y:2.29-32: warning: useless precedence and associativity for "<=" [-Wprecedence]
]])
AT_CLEANUP
+## ---------------- ##
+## Invalid inputs. ##
+## ---------------- ##
-## ------------------- ##
-## Rule Line Numbers. ##
-## ------------------- ##
-AT_SETUP([Rule Line Numbers])
+AT_SETUP([Invalid inputs])
AT_DATA([input.y],
[[%%
-expr:
-'a'
+?
+default: 'a' }
+%&
+%a-does-not-exist
+%-
+%{
+]])
-{
+AT_BISON_CHECK([input.y], [1], [],
+[[input.y:2.1: error: invalid character: '?'
+input.y:3.14: error: invalid character: '}'
+input.y:4.1: error: invalid character: '%'
+input.y:4.2: error: invalid character: '&'
+input.y:5.1-17: error: invalid directive: '%a-does-not-exist'
+input.y:6.1: error: invalid character: '%'
+input.y:6.2: error: invalid character: '-'
+input.y:7.1-8.0: error: missing '%}' at end of file
+input.y:7.1-8.0: error: syntax error, unexpected %{...%}
+]])
-}
+AT_CLEANUP
-'b'
-{
+AT_SETUP([Invalid inputs with {}])
-}
+AT_DATA([input.y],
+[[
+%destructor
+%initial-action
+%lex-param
+%parse-param
+%printer
+%union
+]])
-|
+AT_BISON_CHECK([input.y], [1], [],
+[[input.y:3.1-15: error: syntax error, unexpected %initial-action, expecting {...}
+]])
+AT_CLEANUP
-{
-}
+## ------------------- ##
+## Token definitions. ##
+## ------------------- ##
-'c'
-{
+AT_SETUP([Token definitions])
+
+AT_BISON_OPTION_PUSHDEFS
+# Bison managed, when fed with '%token 'f' "f"' to #define 'f'!
+AT_DATA_GRAMMAR([input.y],
+[%{
+]AT_YYERROR_DECLARE[
+]AT_YYLEX_DECLARE[
+%}
+[%error-verbose
+%token MYEOF 0 "end of file"
+%token 'a' "a"
+%token B_TOKEN "b"
+%token C_TOKEN 'c'
+%token 'd' D_TOKEN
+%token SPECIAL "\\\'\?\"\a\b\f\n\r\t\v\001\201\x001\x000081??!"
+%token SPECIAL "\\\'\?\"\a\b\f\n\r\t\v\001\201\x001\x000081??!"
+%%
+exp: "a" "\\\'\?\"\a\b\f\n\r\t\v\001\201\x001\x000081??!";
+%%
+]AT_YYERROR_DEFINE[
+]AT_YYLEX_DEFINE([{ SPECIAL }])[
+]AT_MAIN_DEFINE[
+]])
+AT_BISON_OPTION_POPDEFS
+
+# Checking the warning message guarantees that the trigraph "??!" isn't
+# unnecessarily escaped here even though it would need to be if encoded in a
+# C-string literal. Also notice that unnecessary escaping, such as "\?", from
+# the user specification is eliminated.
+AT_BISON_CHECK([-o input.c input.y], [[0]], [[]],
+[[input.y:22.8-14: warning: symbol SPECIAL redeclared [-Wother]
+input.y:22.8-63: warning: symbol "\\'?\"\a\b\f\n\r\t\v\001\201\001\201??!" used more than once as a literal string [-Wother]
+]])
+AT_BISON_CHECK([-fcaret -o input.c input.y], [[0]], [[]],
+[[input.y:22.8-14: warning: symbol SPECIAL redeclared [-Wother]
+ %token SPECIAL "\\\'\?\"\a\b\f\n\r\t\v\001\201\x001\x000081??!"
+ ^^^^^^^
+input.y:22.8-63: warning: symbol "\\'?\"\a\b\f\n\r\t\v\001\201\001\201??!" used more than once as a literal string [-Wother]
+ %token SPECIAL "\\\'\?\"\a\b\f\n\r\t\v\001\201\x001\x000081??!"
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+]])
+AT_COMPILE([input])
+
+# Checking the error message here guarantees that yytname, which does contain
+# C-string literals, does have the trigraph escaped correctly. Thus, the
+# symbol name reported by the parser is exactly the same as that reported by
+# Bison itself.
+AT_DATA([experr],
+[[syntax error, unexpected "\\'?\"\a\b\f\n\r\t\v\001\201\001\201??!", expecting a
+]])
+AT_PARSER_CHECK([./input], 1, [], [experr])
+AT_CLEANUP
-}
+
+
+## -------------------- ##
+## Characters Escapes. ##
+## -------------------- ##
+
+
+AT_SETUP([Characters Escapes])
+
+AT_BISON_OPTION_PUSHDEFS
+AT_DATA_GRAMMAR([input.y],
+[%{
+]AT_YYERROR_DECLARE_EXTERN[
+]AT_YYLEX_DECLARE_EXTERN[
+%}
+[%%
+exp:
+ '\'' "\'"
+| '\"' "\""
+| '"' "'" /* Pacify font-lock-mode: ". */
+;
]])
-AT_CHECK([bison input.y -o input.c -v], 0, [], [])
+AT_BISON_OPTION_POPDEFS
-# Check the contents of the report.
-AT_CHECK([cat input.output], [],
+AT_BISON_CHECK([-o input.c input.y])
+AT_COMPILE([input.o])
+AT_CLEANUP
+
+
+
+## -------------- ##
+## Web2c Report. ##
+## -------------- ##
+
+# The generation of the reduction was once wrong in Bison, and made it
+# miss some reductions. In the following test case, the reduction on
+# 'undef_id_tok' in state 1 was missing. This is stripped down from
+# the actual web2c.y.
+
+AT_SETUP([Web2c Report])
+
+AT_KEYWORDS([report])
+
+AT_DATA([input.y],
+[[%token undef_id_tok const_id_tok
+
+%start CONST_DEC_PART
+\f
+%%
+CONST_DEC_PART:
+ CONST_DEC_LIST
+ ;
+
+CONST_DEC_LIST:
+ CONST_DEC
+ | CONST_DEC_LIST CONST_DEC
+ ;
+
+CONST_DEC:
+ { } undef_id_tok '=' const_id_tok ';'
+ ;
+%%
+]])
+
+AT_BISON_CHECK([-v input.y])
+AT_CHECK([cat input.output], 0,
[[Grammar
- Number, Line, Rule
- 1 2 @1 -> /* empty */
- 2 2 expr -> 'a' @1 'b'
- 3 15 @2 -> /* empty */
- 4 15 expr -> @2 'c'
+ 0 $accept: CONST_DEC_PART $end
+
+ 1 CONST_DEC_PART: CONST_DEC_LIST
+
+ 2 CONST_DEC_LIST: CONST_DEC
+ 3 | CONST_DEC_LIST CONST_DEC
+
+ 4 $@1: %empty
+
+ 5 CONST_DEC: $@1 undef_id_tok '=' const_id_tok ';'
Terminals, with rules where they appear
-$ (-1)
-'a' (97) 2
-'b' (98) 2
-'c' (99) 4
+$end (0) 0
+';' (59) 5
+'=' (61) 5
error (256)
+undef_id_tok (258) 5
+const_id_tok (259) 5
Nonterminals, with rules where they appear
-expr (6)
- on left: 2 4
-@1 (7)
- on left: 1, on right: 2
-@2 (8)
- on left: 3, on right: 4
+$accept (7)
+ on left: 0
+CONST_DEC_PART (8)
+ on left: 1, on right: 0
+CONST_DEC_LIST (9)
+ on left: 2 3, on right: 1 3
+CONST_DEC (10)
+ on left: 5, on right: 2 3
+$@1 (11)
+ on left: 4, on right: 5
+
+
+State 0
+
+ 0 $accept: . CONST_DEC_PART $end
+
+ $default reduce using rule 4 ($@1)
+ CONST_DEC_PART go to state 1
+ CONST_DEC_LIST go to state 2
+ CONST_DEC go to state 3
+ $@1 go to state 4
-state 0
- 'a' shift, and go to state 1
+State 1
- $default reduce using rule 3 (@2)
+ 0 $accept: CONST_DEC_PART . $end
- expr go to state 6
- @2 go to state 2
+ $end shift, and go to state 5
+State 2
-state 1
+ 1 CONST_DEC_PART: CONST_DEC_LIST .
+ 3 CONST_DEC_LIST: CONST_DEC_LIST . CONST_DEC
- expr -> 'a' . @1 'b' (rule 2)
+ undef_id_tok reduce using rule 4 ($@1)
+ $default reduce using rule 1 (CONST_DEC_PART)
- $default reduce using rule 1 (@1)
+ CONST_DEC go to state 6
+ $@1 go to state 4
- @1 go to state 3
+State 3
+ 2 CONST_DEC_LIST: CONST_DEC .
-state 2
+ $default reduce using rule 2 (CONST_DEC_LIST)
- expr -> @2 . 'c' (rule 4)
- 'c' shift, and go to state 4
+State 4
+ 5 CONST_DEC: $@1 . undef_id_tok '=' const_id_tok ';'
+ undef_id_tok shift, and go to state 7
-state 3
- expr -> 'a' @1 . 'b' (rule 2)
+State 5
- 'b' shift, and go to state 5
+ 0 $accept: CONST_DEC_PART $end .
+ $default accept
-state 4
+State 6
- expr -> @2 'c' . (rule 4)
+ 3 CONST_DEC_LIST: CONST_DEC_LIST CONST_DEC .
- $default reduce using rule 4 (expr)
+ $default reduce using rule 3 (CONST_DEC_LIST)
+State 7
-state 5
+ 5 CONST_DEC: $@1 undef_id_tok . '=' const_id_tok ';'
- expr -> 'a' @1 'b' . (rule 2)
+ '=' shift, and go to state 8
- $default reduce using rule 2 (expr)
+State 8
+ 5 CONST_DEC: $@1 undef_id_tok '=' . const_id_tok ';'
-state 6
+ const_id_tok shift, and go to state 9
- $ shift, and go to state 7
+State 9
+ 5 CONST_DEC: $@1 undef_id_tok '=' const_id_tok . ';'
-state 7
+ ';' shift, and go to state 10
- $default accept
+State 10
+ 5 CONST_DEC: $@1 undef_id_tok '=' const_id_tok ';' .
+
+ $default reduce using rule 5 (CONST_DEC)
]])
AT_CLEANUP
+## --------------- ##
+## Web2c Actions. ##
+## --------------- ##
-## -------------------- ##
-## %expect not enough. ##
-## -------------------- ##
+# The generation of the mapping 'state -> action' was once wrong in
+# extremely specific situations. web2c.y exhibits this situation.
+# Below is a stripped version of the grammar. It looks like one can
+# simplify it further, but just don't: it is tuned to exhibit a bug,
+# which disapears when applying sane grammar transformations.
+#
+# It used to be wrong on yydefact only:
+#
+# static const yytype_uint8 yydefact[] =
+# {
+# - 2, 0, 1, 0, 0, 2, 3, 2, 5, 4,
+# + 2, 0, 1, 0, 0, 0, 3, 2, 5, 4,
+# 0, 0
+# };
+#
+# but let's check all the tables.
+
+
+AT_SETUP([Web2c Actions])
-AT_SETUP([%expect not enough])
+AT_KEYWORDS([report])
AT_DATA([input.y],
-[[%token NUM OP
-%expect 0
+[[%%
+statement: struct_stat;
+struct_stat: %empty | if else;
+if: "if" "const" "then" statement;
+else: "else" statement;
%%
-exp: exp OP exp | NUM;
]])
-AT_CHECK([bison input.y -o input.c], 1, [],
-[input.y contains 1 shift/reduce conflict.
-expected 0 shift/reduce conflicts
-])
+AT_BISON_CHECK([-v -o input.c input.y])
+
+# Check only the tables.
+[sed -n 's/ *$//;/^static const.*\[\] =/,/^}/p' input.c >tables.c]
+
+AT_CHECK([[cat tables.c]], 0,
+[[static const yytype_uint8 yytranslate[] =
+{
+ 0, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 1, 2, 3, 4,
+ 5, 6
+};
+static const yytype_uint8 yyrline[] =
+{
+ 0, 2, 2, 3, 3, 4, 5
+};
+static const char *const yytname[] =
+{
+ "$end", "error", "$undefined", "\"if\"", "\"const\"", "\"then\"",
+ "\"else\"", "$accept", "statement", "struct_stat", "if", "else", YY_NULL
+};
+static const yytype_uint16 yytoknum[] =
+{
+ 0, 256, 257, 258, 259, 260, 261
+};
+static const yytype_int8 yypact[] =
+{
+ -2, -1, 4, -8, 0, 2, -8, -2, -8, -2,
+ -8, -8
+};
+static const yytype_uint8 yydefact[] =
+{
+ 3, 0, 0, 2, 0, 0, 1, 3, 4, 3,
+ 6, 5
+};
+static const yytype_int8 yypgoto[] =
+{
+ -8, -7, -8, -8, -8
+};
+static const yytype_int8 yydefgoto[] =
+{
+ -1, 2, 3, 4, 8
+};
+static const yytype_uint8 yytable[] =
+{
+ 10, 1, 11, 5, 6, 0, 7, 9
+};
+static const yytype_int8 yycheck[] =
+{
+ 7, 3, 9, 4, 0, -1, 6, 5
+};
+static const yytype_uint8 yystos[] =
+{
+ 0, 3, 8, 9, 10, 4, 0, 6, 11, 5,
+ 8, 8
+};
+static const yytype_uint8 yyr1[] =
+{
+ 0, 7, 8, 9, 9, 10, 11
+};
+static const yytype_uint8 yyr2[] =
+{
+ 0, 2, 1, 0, 2, 4, 2
+};
+]])
+
AT_CLEANUP
-## --------------- ##
-## %expect right. ##
-## --------------- ##
+## ------------------------- ##
+## yycheck Bound Violation. ##
+## ------------------------- ##
-AT_SETUP([%expect right])
-AT_DATA([input.y],
-[[%token NUM OP
-%expect 1
+# _AT_DATA_DANCER_Y(BISON-OPTIONS)
+# --------------------------------
+# The following grammar, taken from Andrew Suffield's GPL'd implementation
+# of DGMTP, the Dancer Generic Message Transport Protocol, used to violate
+# yycheck's bounds where issuing a verbose error message. Keep this test
+# so that possible bound checking compilers could check all the skeletons.
+m4_define([_AT_DATA_DANCER_Y],
+[AT_DATA_GRAMMAR([dancer.y],
+[[%code provides
+{
+ ]AT_YYERROR_DECLARE[
+ ]AT_YYLEX_DECLARE[
+}
+$1
+%token ARROW INVALID NUMBER STRING DATA
+%verbose
+%error-verbose
+/* Grammar follows */
+%%
+line: header body
+ ;
+
+header: '<' from ARROW to '>' type ':'
+ | '<' ARROW to '>' type ':'
+ | ARROW to type ':'
+ | type ':'
+ | '<' '>'
+ ;
+
+from: DATA
+ | STRING
+ | INVALID
+ ;
+
+to: DATA
+ | STRING
+ | INVALID
+ ;
+
+type: DATA
+ | STRING
+ | INVALID
+ ;
+
+body: %empty
+ | body member
+ ;
+
+member: STRING
+ | DATA
+ | '+' NUMBER
+ | '-' NUMBER
+ | NUMBER
+ | INVALID
+ ;
%%
-exp: exp OP exp | NUM;
+]AT_YYERROR_DEFINE[
+]AT_YYLEX_DEFINE([":"])[
+]AT_MAIN_DEFINE[
]])
+])# _AT_DATA_DANCER_Y
+
+
+# AT_CHECK_DANCER(BISON-OPTIONS)
+# ------------------------------
+# Generate the grammar, compile it, run it.
+m4_define([AT_CHECK_DANCER],
+[AT_SETUP([Dancer $1])
+AT_BISON_OPTION_PUSHDEFS([$1])
+_AT_DATA_DANCER_Y([$1])
+AT_FULL_COMPILE([dancer])
+AT_PARSER_CHECK([./dancer], 1, [],
+[syntax error, unexpected ':'
+])
+AT_BISON_OPTION_POPDEFS
+AT_CLEANUP
+])
-AT_CHECK([bison input.y -o input.c], 0)
+AT_CHECK_DANCER()
+AT_CHECK_DANCER([%glr-parser])
+AT_CHECK_DANCER([%skeleton "lalr1.cc"])
+
+
+## ------------------------------------------ ##
+## Diagnostic that expects two alternatives. ##
+## ------------------------------------------ ##
+
+
+# _AT_DATA_EXPECT2_Y(BISON-OPTIONS)
+# --------------------------------
+m4_define([_AT_DATA_EXPECT2_Y],
+[AT_DATA_GRAMMAR([expect2.y],
+[%{
+static int yylex (AT_LALR1_CC_IF([int *], [void]));
+AT_LALR1_CC_IF([[#include <cstdlib>]],
+[[
+]AT_YYERROR_DECLARE])[
+%}
+$1
+%error-verbose
+%token A 1000
+%token B
+
+%%
+program: %empty
+ | program e ';'
+ | program error ';';
+
+e: e '+' t | t;
+t: A | B;
+
+%%
+]AT_YYERROR_DEFINE[
+]AT_LALR1_CC_IF(
+[int
+yyparse ()
+{
+ yy::parser parser;
+ return parser.parse ();
+}
+])[
+
+#include <assert.h>
+static int
+yylex (]AT_LALR1_CC_IF([int *lval], [void])[)
+{
+ static int const tokens[] =
+ {
+ 1000, '+', '+', -1
+ };
+ static size_t toknum;
+ ]AT_LALR1_CC_IF([*lval = 0; /* Pacify GCC. */])[
+ assert (toknum < sizeof tokens / sizeof *tokens);
+ return tokens[toknum++];
+}
+]AT_MAIN_DEFINE[
+]])
+])# _AT_DATA_EXPECT2_Y
+
+
+# AT_CHECK_EXPECT2(BISON-OPTIONS)
+# -------------------------------
+# Generate the grammar, compile it, run it.
+m4_define([AT_CHECK_EXPECT2],
+[AT_SETUP([Expecting two tokens $1])
+AT_BISON_OPTION_PUSHDEFS([$1])
+_AT_DATA_EXPECT2_Y([$1])
+AT_FULL_COMPILE([expect2])
+AT_PARSER_CHECK([./expect2], 1, [],
+[syntax error, unexpected '+', expecting A or B
+])
+AT_BISON_OPTION_POPDEFS
AT_CLEANUP
+])
+AT_CHECK_EXPECT2()
+AT_CHECK_EXPECT2([%glr-parser])
+AT_CHECK_EXPECT2([%skeleton "lalr1.cc"])
-## ------------------ ##
-## %expect too much. ##
-## ------------------ ##
-AT_SETUP([%expect too much])
-AT_DATA([input.y],
-[[%token NUM OP
-%expect 2
+## --------------------------------------------- ##
+## Braced code in declaration in rules section. ##
+## --------------------------------------------- ##
+
+AT_SETUP([Braced code in declaration in rules section])
+
+# Bison once mistook braced code in a declaration in the rules section to be a
+# rule action.
+AT_BISON_OPTION_PUSHDEFS([%debug])
+AT_DATA_GRAMMAR([input.y],
+[[%{
+]AT_YYERROR_DECLARE[
+]AT_YYLEX_DECLARE[
+%}
+%debug
+%error-verbose
+%%
+
+start:
+ {
+ printf ("Bison would once convert this action to a midrule because of the"
+ " subsequent braced code.\n");
+ }
+ ;
+
+%destructor { fprintf (stderr, "DESTRUCTOR\n"); } 'a';
+%printer { fprintf (yyoutput, "PRINTER"); } 'a';
+
%%
-exp: exp OP exp | NUM;
+]AT_YYERROR_DEFINE[
+]AT_YYLEX_DEFINE(["a"])[
+]AT_MAIN_DEFINE[
+]])
+AT_BISON_OPTION_POPDEFS
+
+AT_BISON_CHECK([-o input.c input.y])
+AT_COMPILE([input])
+AT_PARSER_CHECK([./input --debug], 1,
+[[Bison would once convert this action to a midrule because of the subsequent braced code.
+]],
+[[Starting parse
+Entering state 0
+Reducing stack by rule 1 (line 20):
+-> $$ = nterm start ()
+Stack now 0
+Entering state 1
+Reading a token: Next token is token 'a' (PRINTER)
+syntax error, unexpected 'a', expecting $end
+Error: popping nterm start ()
+Stack now 0
+Cleanup: discarding lookahead token 'a' (PRINTER)
+DESTRUCTOR
+Stack now 0
]])
-AT_CHECK([bison input.y -o input.c], 1, [],
-[input.y contains 1 shift/reduce conflict.
-expected 2 shift/reduce conflicts
-])
AT_CLEANUP
-## ---------------------- ##
-## Mixing %token styles. ##
-## ---------------------- ##
+## --------------------------------- ##
+## String alias declared after use. ##
+## --------------------------------- ##
-AT_SETUP([Mixing %token styles])
+AT_SETUP([String alias declared after use])
+
+# Bison once incorrectly asserted that the symbol number for either a token or
+# its alias was the highest symbol number so far at the point of the alias
+# declaration. That was true unless the declaration appeared after their first
+# uses and other tokens appeared in between.
-# Taken from the documentation.
AT_DATA([input.y],
-[[%token <operator> OR "||"
-%token <operator> LE 134 "<="
-%left OR "<="
+[[%%
+start: 'a' "A" 'b';
+%token 'a' "A";
+]])
+
+AT_BISON_CHECK([-o input.c input.y])
+
+AT_CLEANUP
+
+
+
+## -------------------------------- ##
+## Extra lookahead sets in report. ##
+## -------------------------------- ##
+
+AT_SETUP([[Extra lookahead sets in report]])
+
+# Bison prints each reduction's lookahead set only next to the associated
+# state's one item that (1) is associated with the same rule as the reduction
+# and (2) has its dot at the end of its RHS. Previously, Bison also
+# erroneously printed the lookahead set next to all of the state's other items
+# associated with the same rule. This bug affected only the '.output' file and
+# not the generated parser source code.
+
+AT_DATA([[input.y]],
+[[%%
+start: a | 'a' a 'a' ;
+a: 'a' ;
+]])
+
+AT_BISON_CHECK([[--report=all input.y]])
+AT_CHECK([[sed -n '/^State 1$/,/^State 2$/p' input.output]], [[0]],
+[[State 1
+
+ 2 start: 'a' . a 'a'
+ 3 a: . 'a'
+ 3 | 'a' . [$end]
+
+ 'a' shift, and go to state 4
+
+ $default reduce using rule 3 (a)
+
+ a go to state 5
+
+
+State 2
+]])
+
+AT_CLEANUP
+
+
+
+## ---------------------------------------- ##
+## Token number in precedence declaration. ##
+## ---------------------------------------- ##
+
+AT_SETUP([[Token number in precedence declaration]])
+
+# POSIX says token numbers can be declared in %left, %right, and %nonassoc, but
+# we lost this in Bison 1.50.
+AT_BISON_OPTION_PUSHDEFS
+AT_DATA_GRAMMAR([input.y],
+[[%code {
+ ]AT_YYERROR_DECLARE[
+ ]AT_YYLEX_DECLARE[
+}
+
+%error-verbose
+%right END 0
+%left TK1 1 TK2 2 "tok alias" 3
+
%%
-exp: ;
+
+start:
+ TK1 sr_conflict "tok alias"
+ | start %prec END
+ ;
+sr_conflict:
+ TK2
+ | TK2 "tok alias"
+ ;
+
%%
+
+]AT_YYERROR_DEFINE[
+]AT_YYLEX_DEFINE([{ 1, 2, 3, 0 }])[
+]AT_MAIN_DEFINE[
]])
+AT_BISON_OPTION_POPDEFS
-AT_CHECK([bison -v input.y -o input.c], 0, ignore, ignore)
+AT_BISON_CHECK([[-Wall -o input.c input.y]], [[0]],,
+[[input.y:24.5-19: warning: rule useless in parser due to conflicts [-Wother]
+input.y:28.5-19: warning: rule useless in parser due to conflicts [-Wother]
+input.y:18.7-9: warning: useless precedence and associativity for TK1 [-Wprecedence]
+]])
+AT_COMPILE([[input]])
+AT_PARSER_CHECK([[./input]])
AT_CLEANUP
-## ---------------------- ##
-## %union and --defines. ##
-## ---------------------- ##
+## --------------------------- ##
+## parse-gram.y: LALR = IELR. ##
+## --------------------------- ##
+
+# If parse-gram.y's LALR and IELR parser tables ever begin to differ, we
+# need to fix parse-gram.y or start using IELR.
+
+AT_SETUP([[parse-gram.y: LALR = IELR]])
+
+# Avoid tests/bison's dark magic by processing a local copy of the
+# grammar. Avoid differences in synclines by telling bison that the
+# output files have the same name.
+[cp $abs_top_srcdir/src/parse-gram.y input.y]
+AT_BISON_CHECK([[-o input.c -Dlr.type=lalr input.y]])
+[mv input.c lalr.c]
+AT_CAPTURE_FILE([lalr.c])
+AT_BISON_CHECK([[-o input.c -Dlr.type=ielr input.y]])
+[mv input.c ielr.c]
+AT_CAPTURE_FILE([ielr.c])
+AT_CHECK([[diff lalr.c ielr.c]], [[0]])
+
+AT_CLEANUP
+
+
+
+## -------------------------------------------- ##
+## parse.error=verbose and YYSTACK_USE_ALLOCA. ##
+## -------------------------------------------- ##
+
+AT_SETUP([[parse.error=verbose and YYSTACK_USE_ALLOCA]])
+
+AT_BISON_OPTION_PUSHDEFS
+AT_DATA_GRAMMAR([input.y],
+[[%code {
+ ]AT_YYERROR_DECLARE[
+ ]AT_YYLEX_DECLARE[
+ #define YYSTACK_USE_ALLOCA 1
+}
+%define parse.error verbose
-AT_SETUP([%union and --defines])
+%%
+
+start: check syntax_error syntax_error ;
-AT_DATA([union.y],
-[%union
+check:
{
- int integer;
- char *string ;
+ if (128 < sizeof yymsgbuf)
+ {
+ fprintf (stderr,
+ "The initial size of yymsgbuf in yyparse has increased\n"
+ "since this test group was last updated. As a result,\n"
+ "this test group may no longer manage to induce a\n"
+ "reallocation of the syntax error message buffer.\n"
+ "This test group must be adjusted to produce a longer\n"
+ "error message.\n");
+ YYABORT;
+ }
}
+;
+
+// Induce a syntax error message whose total length is more than
+// sizeof yymsgbuf in yyparse. Each token here is 64 bytes.
+syntax_error:
+ "123456789112345678921234567893123456789412345678951234567896123A"
+| "123456789112345678921234567893123456789412345678951234567896123B"
+| error 'a' 'b' 'c'
+;
+
%%
-exp: {};
-])
-AT_CHECK([bison --defines union.y])
+]AT_YYERROR_DEFINE[
+ /* Induce two syntax error messages (which requires full error
+ recovery by shifting 3 tokens) in order to detect any loss of the
+ reallocated buffer. */
+]AT_YYLEX_DEFINE(["abc"])[
+]AT_MAIN_DEFINE[
+]])
+AT_BISON_OPTION_POPDEFS
+
+AT_BISON_CHECK([[-o input.c input.y]])
+AT_COMPILE([[input]])
+AT_PARSER_CHECK([[./input]], [[1]], [],
+[[syntax error, unexpected 'a', expecting 123456789112345678921234567893123456789412345678951234567896123A or 123456789112345678921234567893123456789412345678951234567896123B
+syntax error, unexpected $end, expecting 123456789112345678921234567893123456789412345678951234567896123A or 123456789112345678921234567893123456789412345678951234567896123B
+]])
AT_CLEANUP
-## --------------------------------------- ##
-## Duplicate '/' in C comments in %union ##
-## --------------------------------------- ##
+## ------------------------------ ##
+## parse.error=verbose overflow. ##
+## ------------------------------ ##
+
+# Imagine the case where YYSTACK_ALLOC_MAXIMUM = YYSIZE_MAXIMUM and an
+# invocation of yysyntax_error has caused yymsg_alloc to grow to exactly
+# YYSTACK_ALLOC_MAXIMUM (perhaps because the normal doubling of size had
+# to be clipped to YYSTACK_ALLOC_MAXIMUM). In an old version of yacc.c,
+# a subsequent invocation of yysyntax_error that overflows during its
+# size calculation would return YYSIZE_MAXIMUM to yyparse. Then,
+# yyparse would invoke yyerror using the old contents of yymsg.
+
+AT_SETUP([[parse.error=verbose overflow]])
+
+AT_BISON_OPTION_PUSHDEFS
+AT_DATA_GRAMMAR([input.y],
+[[%code {
+ ]AT_YYERROR_DECLARE[
+ ]AT_YYLEX_DECLARE[
-AT_SETUP([%union and C comments])
+ /* This prevents this test case from having to induce error messages
+ large enough to overflow size_t. */
+ #define YYSIZE_T unsigned char
+
+ /* Bring in malloc and set EXIT_SUCCESS so yacc.c doesn't try to
+ provide a malloc prototype using our YYSIZE_T. */
+ #include <stdlib.h>
+ #ifndef EXIT_SUCCESS
+ # define EXIT_SUCCESS 0
+ #endif
+
+ /* Max depth is usually much smaller than YYSTACK_ALLOC_MAXIMUM, and
+ we don't want gcc to warn everywhere this constant would be too big
+ to make sense for our YYSIZE_T. */
+ #define YYMAXDEPTH 100
+}
-AT_DATA([union-comment.y],
-[%union
+%define parse.error verbose
+
+%%
+
+start: syntax_error1 check syntax_error2 ;
+
+// Induce a syntax error message whose total length causes yymsg in
+// yyparse to be reallocated to size YYSTACK_ALLOC_MAXIMUM, which
+// should be 255. Each token here is 64 bytes.
+syntax_error1:
+ "123456789112345678921234567893123456789412345678951234567896123A"
+| "123456789112345678921234567893123456789412345678951234567896123B"
+| "123456789112345678921234567893123456789412345678951234567896123C"
+| error 'a' 'b' 'c'
+;
+
+check:
{
- /* The int. */ int integer;
- /* The string. */ char *string ;
+ if (yymsg_alloc != YYSTACK_ALLOC_MAXIMUM
+ || YYSTACK_ALLOC_MAXIMUM != YYSIZE_MAXIMUM
+ || YYSIZE_MAXIMUM != 255)
+ {
+ fprintf (stderr,
+ "The assumptions of this test group are no longer\n"
+ "valid, so it may no longer catch the error it was\n"
+ "designed to catch. Specifically, the following\n"
+ "values should all be 255:\n\n");
+ fprintf (stderr, " yymsg_alloc = %d\n", yymsg_alloc);
+ fprintf (stderr, " YYSTACK_ALLOC_MAXIMUM = %d\n",
+ YYSTACK_ALLOC_MAXIMUM);
+ fprintf (stderr, " YYSIZE_MAXIMUM = %d\n", YYSIZE_MAXIMUM);
+ YYABORT;
+ }
}
+;
+
+// Now overflow.
+syntax_error2:
+ "123456789112345678921234567893123456789412345678951234567896123A"
+| "123456789112345678921234567893123456789412345678951234567896123B"
+| "123456789112345678921234567893123456789412345678951234567896123C"
+| "123456789112345678921234567893123456789412345678951234567896123D"
+| "123456789112345678921234567893123456789412345678951234567896123E"
+;
+
%%
-exp: {};
-])
-AT_CHECK([bison union-comment.y])
-AT_CHECK([fgrep '//*' union-comment.tab.c], [1], [])
+]AT_YYERROR_DEFINE[
+ /* Induce two syntax error messages (which requires full error
+ recovery by shifting 3 tokens). */
+]AT_YYLEX_DEFINE(["abc"])[
+int
+main (void)
+{
+ /* Push parsers throw away the message buffer between tokens, so skip
+ this test under maintainer-push-check. */
+ if (YYPUSH)
+ return 77;
+ return yyparse ();
+}
+]])
+
+AT_BISON_CHECK([[-o input.c input.y]])
+
+# gcc warns about tautologies and fallacies involving comparisons for
+# unsigned char. However, it doesn't produce these same warnings for
+# size_t and many other types when the warnings would seem to make just
+# as much sense. We ignore the warnings.
+[CFLAGS="$NO_WERROR_CFLAGS"]
+AT_COMPILE([[input]])
+AT_PARSER_CHECK([[./input]], [[2]], [],
+[[syntax error, unexpected 'a', expecting 123456789112345678921234567893123456789412345678951234567896123A or 123456789112345678921234567893123456789412345678951234567896123B or 123456789112345678921234567893123456789412345678951234567896123C
+syntax error
+memory exhausted
+]])
+AT_BISON_OPTION_POPDEFS
AT_CLEANUP
-## ----------------- ##
-## Invalid input 1. ##
-## ----------------- ##
+## ------------------------ ##
+## LAC: Exploratory stack. ##
+## ------------------------ ##
-AT_SETUP([Invalid input: 1])
+AT_SETUP([[LAC: Exploratory stack]])
-AT_DATA([input.y],
-[[%%
-?
-]])
+m4_pushdef([AT_LAC_CHECK], [
-AT_CHECK([bison input.y], [1], [],
-[[input.y:2: invalid input: `?'
-input.y:3: fatal error: no rules in the input grammar
-]])
+AT_BISON_OPTION_PUSHDEFS([%debug $1])
-AT_CLEANUP
+AT_DATA_GRAMMAR([input.y],
+[[%code {
+ ]AT_YYERROR_DECLARE[
+ int yylex (]AT_PURE_IF([[YYSTYPE *]], [[void]])[);
+}
+%debug
+]$1[
+%define parse.error verbose
+%token 'c'
+%%
-## ----------------- ##
-## Invalid input 2. ##
-## ----------------- ##
+// default reductions in inconsistent states
+// v v v v v v v v v v v v v v
+S: A B A A B A A A A B A A A A A A A B C C A A A A A A A A A A A A B ;
+// ^ ^ ^
+// LAC reallocs
+A: 'a' | /*empty*/ { printf ("inconsistent default reduction\n"); } ;
+B: 'b' ;
+C: /*empty*/ { printf ("consistent default reduction\n"); } ;
-AT_SETUP([Invalid input: 2])
+%%
+]AT_YYERROR_DEFINE[
+int
+yylex (]AT_PURE_IF([[YYSTYPE *v]], [[void]])[)
+{
+ static char const *input = "bbbbc";]AT_PURE_IF([[
+ *v = 0;]])[
+ return *input++;
+}
-AT_DATA([input.y],
-[[%%
-default: 'a' }
+]AT_MAIN_DEFINE[
]])
-AT_CHECK([bison input.y], [1], [],
-[[input.y:2: invalid input: `}'
+AT_BISON_CHECK([[-Dparse.lac=full -Dparse.lac.es-capacity-initial=1 \
+ -Dparse.lac.memory-trace=full \
+ -o input.c input.y]], [[0]], [],
+[[input.y: warning: 21 shift/reduce conflicts [-Wconflicts-sr]
]])
+AT_COMPILE([[input]])
+AT_PARSER_CHECK([[./input --debug > stdout.txt 2> stderr.txt]], [[1]])
+
+# Make sure syntax error doesn't forget that 'a' is expected. It would
+# be forgotten without lookahead correction.
+AT_CHECK([[grep 'syntax error,' stderr.txt]], [[0]],
+[[syntax error, unexpected 'c', expecting 'a' or 'b'
+]])
+
+# Check number of default reductions in inconsistent states to be sure
+# syntax error is detected before unnecessary reductions are performed.
+AT_CHECK([[$PERL -0777 -ne 'print s/inconsistent default reduction//g;' \
+ < stdout.txt || exit 77]], [[0]], [[14]])
+
+# Check number of default reductions in consistent states to be sure
+# it is performed before the syntax error is detected.
+AT_CHECK([[$PERL -0777 -ne 'print s/\bconsistent default reduction//g;' \
+ < stdout.txt || exit 77]], [[0]], [[2]])
+
+# Check number of reallocs to be sure reallocated memory isn't somehow
+# lost between LAC invocations.
+AT_CHECK([[$PERL -0777 -ne 'print s/\(realloc//g;' < stderr.txt \
+ || exit 77]], [[0]], [[3]])
+
+AT_BISON_OPTION_POPDEFS
+])
+
+AT_LAC_CHECK([[%define api.push-pull pull]])
+AT_LAC_CHECK([[%define api.push-pull pull %define api.pure]])
+AT_LAC_CHECK([[%define api.push-pull both]])
+AT_LAC_CHECK([[%define api.push-pull both %define api.pure]])
+
+m4_popdef([AT_LAC_CHECK])
AT_CLEANUP
-## -------------------- ##
-## Invalid %directive. ##
-## -------------------- ##
+## ------------------------ ##
+## LAC: Memory exhaustion. ##
+## ------------------------ ##
+AT_SETUP([[LAC: Memory exhaustion]])
-AT_SETUP([Invalid %directive])
+m4_pushdef([AT_LAC_CHECK],
+[AT_BISON_OPTION_PUSHDEFS([%debug])
+AT_DATA_GRAMMAR([input.y],
+[[%code {
+ ]AT_YYERROR_DECLARE[
+ ]AT_YYLEX_DECLARE[
+ #define YYMAXDEPTH 8
+}
+%debug
+%error-verbose
-AT_DATA([input.y],
-[[%invalid
+%%
+
+S: A A A A A A A A A ;
+A: /*empty*/ | 'a' ;
+
+%%
+]AT_YYERROR_DEFINE[
+]AT_YYLEX_DEFINE(["$1"])[
+]AT_MAIN_DEFINE[
]])
-AT_CHECK([bison input.y], [1], [],
-[[input.y:1: unrecognized: %invalid
-input.y:1: Skipping to next %
-input.y:2: fatal error: no input grammar
+AT_BISON_CHECK([[-Dparse.lac=full -Dparse.lac.es-capacity-initial=1 \
+ -o input.c input.y]], [[0]], [],
+[[input.y: warning: 8 shift/reduce conflicts [-Wconflicts-sr]
+]])
+AT_COMPILE([[input]])
+AT_BISON_OPTION_POPDEFS
+])
+
+# Check for memory exhaustion during parsing.
+AT_LAC_CHECK([])
+AT_PARSER_CHECK([[./input --debug]], [[2]], [],
+[[Starting parse
+Entering state 0
+Reading a token: Now at end of input.
+LAC: initial context established for $end
+LAC: checking lookahead $end: R2 G3 R2 G5 R2 G6 R2 G7 R2 G8 R2 G9 R2 G10 R2 G11 R2 (max size exceeded)
+memory exhausted
+Cleanup: discarding lookahead token $end ()
+Stack now 0
]])
+# Induce an immediate syntax error with an undefined token, and check
+# for memory exhaustion while building syntax error message.
+AT_LAC_CHECK([z], [[0]])
+AT_PARSER_CHECK([[./input --debug]], [[2]], [],
+[[Starting parse
+Entering state 0
+Reading a token: Next token is token $undefined ()
+LAC: initial context established for $undefined
+LAC: checking lookahead $undefined: Always Err
+Constructing syntax error message
+LAC: checking lookahead $end: R2 G3 R2 G5 R2 G6 R2 G7 R2 G8 R2 G9 R2 G10 R2 G11 R2 (max size exceeded)
+syntax error
+memory exhausted
+Cleanup: discarding lookahead token $undefined ()
+Stack now 0
+]])
+
+m4_popdef([AT_LAC_CHECK])
+
AT_CLEANUP
+## ---------------------- ##
+## Lex and parse params. ##
+## ---------------------- ##
-## --------------------- ##
-## Invalid CPP headers. ##
-## --------------------- ##
+# AT_TEST(SKELETON)
+# -----------------
+# Check that the identifier of the params is properly fetched
+# even when there are trailing blanks.
+
+m4_pushdef([AT_TEST],
+[AT_SETUP([[Lex and parse params: $1]])
+
+## FIXME: Improve parsing of parse-param.
+AT_BISON_OPTION_PUSHDEFS([%locations %skeleton "$1" %parse-param { int x } %parse-param { int y }])
+
+AT_DATA_GRAMMAR([input.y],
+[[%locations
+%skeleton "$1"
+%union { int ival; }
+%parse-param { int x }
+// Spaces, tabs, and new lines.
+%parse-param { @&t@
+@tb@ int y@tb@ @&t@
+ @&t@
+ @&t@
+}
-# AT_TEST_CPP_GUARD_H([INPUT-FILE-BASE)
-# -------------------------------------
-m4_define([AT_TEST_CPP_GUARD_H],
-[AT_SETUP([Invalid CPP guards: $1])
+%{
+ ]AT_YYERROR_DECLARE[
+ ]AT_YYLEX_DECLARE[
+%}
-# Possibly create inner directories.
-dirname=`AS_DIRNAME([$1])`
-AS_MKDIR_P([$dirname])
+%%
+exp: 'a' { fprintf (stdout, "x: %d, y: %d\n", x, y); };
+%%
+]AT_YYERROR_DEFINE[
+]AT_YYLEX_DEFINE(["a"])[
-AT_DATA([$1.y],
-[%%
-dummy:
-])
+]AT_SKEL_CC_IF(
+[int
+yyparse (int x, int y)
+{
+ yy::parser parser(x, y);
+ return parser.parse ();
+}
+])[
-AT_CHECK([bison --defines=$1.h $1.y])
+int
+main (void)
+{
+ return yyparse(1, 2);
+}
+]])
-# CPP should be happy with it.
-AT_CHECK([$CC -E $1.h], 0, [ignore])
+AT_FULL_COMPILE([input])
+AT_PARSER_CHECK([./input], 0, [[x: 1, y: 2
+]])
+AT_BISON_OPTION_POPDEFS
AT_CLEANUP
])
-AT_TEST_CPP_GUARD_H([input/input])
-AT_TEST_CPP_GUARD_H([9foo])
+## FIXME: test Java, and iterate over skeletons.
+AT_TEST([yacc.c])
+AT_TEST([glr.c])
+AT_TEST([lalr1.cc])
+AT_TEST([glr.cc])
+m4_popdef([AT_TEST])
-## ---------------- ##
-## Broken Closure. ##
-## ---------------- ##
-# TC was once broken during a massive `simplification' of the code.
-# It resulted in bison dumping core on the following grammar (the
-# computation of FIRSTS uses TC). It managed to produce a pretty
-# exotic closure:
-#
-# TC: Input
-#
-# 01234567
-# +--------+
-# 0| 1 |
-# 1| 1 |
-# 2| 1 |
-# 3| 1 |
-# 4| 1 |
-# 5| 1 |
-# 6| 1|
-# 7| |
-# +--------+
-#
-# TC: Output
-#
-# 01234567
-# +--------+
-# 0| 1 |
-# 1| 111 |
-# 2| 111 |
-# 3| 1111 |
-# 4| 111 1 |
-# 5| 111 1 |
-# 6| 111 1|
-# 7| 111 |
-# +--------+
-#
-# instead of that below.
+## ----------------------- ##
+## stdio.h is not needed. ##
+## ----------------------- ##
-AT_SETUP([Broken Closure])
+# At some point, by accident, yy_location_print_ was using fprintf and
+# FILE which are from stdio.h, which we do not require.
+AT_SETUP([[stdio.h is not needed]])
-AT_DATA([input.y],
-[[%%
-a: b
-b: c
-c: d
-d: e
-e: f
-f: g
-g: h
-h: 'h'
-]])
-
-AT_CHECK([bison --trace input.y 2>&1 |
- sed -n '/^TC: Output BEGIN/,/^TC: Output END/p'],
- [0],
-[[TC: Output BEGIN
- @&t@
- 01234567
- +--------+
- 0| 1111111|
- 1| 111111|
- 2| 11111|
- 3| 1111|
- 4| 111|
- 5| 11|
- 6| 1|
- 7| |
- +--------+
-TC: Output END
+AT_BISON_OPTION_PUSHDEFS
+
+AT_DATA_GRAMMAR([input.y],
+[[%locations
+%code
+{
+ static int yylex (void) { return 0; }
+ static void yyerror (const char* msg) { (void) msg; }
+}
+%%
+exp: {}
+%%
+]AT_MAIN_DEFINE[
]])
+AT_FULL_COMPILE([input])
+
+AT_BISON_OPTION_POPDEFS
+
AT_CLEANUP
projects / bison.git / blobdiff