# Bison Regressions. -*- Autotest -*-
-# Copyright 2001 Free Software Foundation, Inc.
+
+# Copyright (C) 2001, 2002, 2003, 2004, 2005, 2006 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
# 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.
+# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+# 02110-1301, USA.
AT_BANNER([[Regression tests.]])
+
## ------------------ ##
-## Duplicate string. ##
+## Trivial grammars. ##
## ------------------ ##
+AT_SETUP([Trivial grammars])
-AT_SETUP([Duplicate string])
+AT_DATA_GRAMMAR([input.y],
+[[%{
+void yyerror (char const *);
+int yylex (void);
+#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. */
+%error-verbose
-%token NUM
-%token LE "<="
-%token GE "<="
-
-%%
-exp: '(' exp ')' | NUM ;
%%
+
+program: 'x';
]])
-AT_CHECK([bison -v duplicate.y -o duplicate.c], 0, ignore, ignore)
+AT_CHECK([bison -o input.c input.y])
+AT_COMPILE([input.o], [-c input.c])
+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_DATA_GRAMMAR([input.y],
+[[%{
+void yyerror (char const *);
+int yylex (void);
+typedef union { char const *val; } YYSTYPE;
+%}
+
+%type <val> program
-AT_DATA([input.y],
-[[%token NUM OP
%%
-exp: exp OP exp | NUM;
+
+program: { $$ = ""; };
]])
-AT_CHECK([bison input.y -o input.c -v], 0, [],
-[input.y contains 1 shift/reduce conflict.
-])
+AT_CHECK([bison -o input.c input.y])
+AT_COMPILE([input.o], [-c input.c])
-# 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_DATA_GRAMMAR([input.y],
+[[%{
+void yyerror (const char *s);
+int yylex (void);
+%}
-Nonterminals, with rules where they appear
+%union
+{
+ int val;
+};
+%{
+#ifndef MY_TOKEN
+# error "MY_TOKEN not defined."
+#endif
+%}
+%token MY_TOKEN
+%%
+exp: MY_TOKEN;
+%%
+]])
-exp (5)
- on left: 1 2, on right: 1
+AT_CHECK([bison -y -o input.c input.y])
+AT_COMPILE([input.o], [-c input.c])
+AT_CLEANUP
-state 0
- NUM shift, and go to state 1
- exp go to state 2
+## ---------------------------------------- ##
+## Early token definitions without --yacc. ##
+## ---------------------------------------- ##
+AT_SETUP([Early token definitions without --yacc])
-state 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.
- exp -> NUM . (rule 2)
+AT_DATA_GRAMMAR([input.y],
+[[%{
+#include <stdio.h>
+void yyerror (const char *s);
+int yylex (void);
+void print_my_token (void);
+%}
- $default reduce using rule 2 (exp)
+%union
+{
+ int val;
+};
+%{
+void
+print_my_token (void)
+{
+ enum yytokentype my_token = MY_TOKEN;
+ printf ("%d\n", my_token);
+}
+%}
+%token MY_TOKEN
+%%
+exp: MY_TOKEN;
+%%
+]])
+AT_CHECK([bison -o input.c input.y])
+AT_COMPILE([input.o], [-c input.c])
+AT_CLEANUP
-state 2
- exp -> exp . OP exp (rule 1)
- $ go to state 5
- OP shift, and go to state 3
+## ---------------- ##
+## Braces parsing. ##
+## ---------------- ##
+AT_SETUP([Braces parsing])
-state 3
+AT_DATA([input.y],
+[[/* Bison used to swallow the character after `}'. */
- exp -> exp OP . exp (rule 1)
+%%
+exp: { tests = {{{{{{{{{{}}}}}}}}}}; };
+%%
+]])
- NUM shift, and go to state 1
+AT_CHECK([bison -v -o input.c input.y])
- exp go to state 4
+AT_CHECK([grep 'tests = {{{{{{{{{{}}}}}}}}}};' input.c], 0, [ignore])
+AT_CLEANUP
-state 4
+## ------------------ ##
+## Duplicate string. ##
+## ------------------ ##
- exp -> exp . OP exp (rule 1)
- exp -> exp OP exp . (rule 1)
- OP shift, and go to state 3
+AT_SETUP([Duplicate string])
- OP [reduce using rule 1 (exp)]
- $default reduce using rule 1 (exp)
+AT_DATA([input.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 "<="
+%%
+exp: '(' exp ')' | NUM ;
+%%
+]])
-state 5
+AT_CHECK([bison -v -o input.c input.y], 0, [],
+[[input.y:6.8-14: warning: symbol `"<="' used more than once as a literal string
+]])
- $ go to state 6
+AT_CLEANUP
+## ------------------- ##
+## Rule Line Numbers. ##
+## ------------------- ##
-state 6
+AT_SETUP([Rule Line Numbers])
- $default accept
+AT_KEYWORDS([report])
+AT_DATA([input.y],
+[[%%
+expr:
+'a'
-]])
+{
-AT_CLEANUP
+}
+'b'
-## --------------------- ##
-## Solved SR Conflicts. ##
-## --------------------- ##
+{
-AT_SETUP([Solved SR Conflicts])
+}
-AT_DATA([input.y],
-[[%token NUM OP
-%right OP
-%%
-exp: exp OP exp | NUM;
+|
+
+
+{
+
+
+}
+
+'c'
+
+{
+
+};
]])
-AT_CHECK([bison input.y -o input.c -v], 0, [], [])
+AT_CHECK([bison -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
- NUM shift, and go to state 1
+ 0 $accept: . expr $end
- exp go to state 2
+ 'a' shift, and go to state 1
+ $default reduce using rule 3 (@2)
+
+ expr go to state 2
+ @2 go to state 3
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
- exp -> exp . OP exp (rule 1)
-
- $ go to state 5
- OP shift, and go to state 3
+ 0 $accept: expr . $end
+ $end shift, and go to state 5
state 3
- exp -> exp OP . exp (rule 1)
+ 4 expr: @2 . 'c'
- NUM shift, and go to state 1
+ 'c' shift, and go to state 6
- exp go to state 4
+state 4
+ 2 expr: 'a' @1 . 'b'
-state 4
+ 'b' shift, and go to state 7
- exp -> exp . OP exp (rule 1)
- exp -> exp OP exp . (rule 1)
- OP shift, and go to state 3
+state 5
- $default reduce using rule 1 (exp)
+ 0 $accept: expr $end .
+ $default accept
-state 5
+state 6
- $ go to state 6
+ 4 expr: @2 'c' .
+ $default reduce using rule 4 (expr)
+
+
+state 7
+
+ 2 expr: 'a' @1 'b' .
+
+ $default reduce using rule 2 (expr)
+]])
+
+AT_CLEANUP
-state 6
- $default accept
+## ---------------------- ##
+## 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: ;
+%%
]])
+AT_CHECK([bison -v -o input.c input.y])
+
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_CHECK([bison input.y], [1], [],
+[[input.y:2.1: invalid character: `?'
+input.y:3.14: invalid character: `}'
+input.y:4.1: invalid character: `%'
+input.y:4.2: invalid character: `&'
+input.y:5.1-17: invalid directive: `%a-does-not-exist'
+input.y:6.1: invalid character: `%'
+input.y:6.2: invalid character: `-'
+input.y:7.1-8.0: missing `%}' at end of file
+input.y:7.1-8.0: 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_CHECK([bison input.y], [1], [],
+[[input.y:3.1-15: syntax error, unexpected %initial-action, expecting {...}
+]])
+AT_CLEANUP
-{
-}
+## ------------------- ##
+## Token definitions. ##
+## ------------------- ##
-'c'
+AT_SETUP([Token definitions])
+
+# Bison managed, when fed with `%token 'f' "f"' to #define 'f'!
+AT_DATA_GRAMMAR([input.y],
+[%{
+#include <stdio.h>
+void yyerror (const char *s);
+int yylex (void);
+%}
+[%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??!"
+%%
+exp: "a" "\\\'\?\"\a\b\f\n\r\t\v\001\201\x001\x000081??!";
+%%
+void
+yyerror (char const *s)
+{
+ fprintf (stderr, "%s\n", s);
+}
+
+int
+yylex (void)
{
+ return SPECIAL;
+}
+int
+main (void)
+{
+ return yyparse ();
}
]])
-AT_CHECK([bison input.y -o input.c -v], 0, [], [])
+AT_CHECK([bison -o input.c input.y])
+AT_COMPILE([input])
+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
-# Check the contents of the report.
-AT_CHECK([cat input.output], [],
+
+
+## -------------------- ##
+## Characters Escapes. ##
+## -------------------- ##
+
+
+AT_SETUP([Characters Escapes])
+
+AT_DATA_GRAMMAR([input.y],
+[%{
+void yyerror (const char *s);
+int yylex (void);
+%}
+[%token QUOTES "\""
+%token TICK "'"
+%%
+exp:
+ '\'' "\'"
+| '\"' "\""
+| '"' "'"
+;
+]])
+# Pacify font-lock-mode: "
+
+AT_CHECK([bison -o input.c input.y])
+AT_COMPILE([input.o], [-c input.c])
+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_CHECK([bison -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
- 'a' shift, and go to state 1
-
- $default reduce using rule 3 (@2)
+ 0 $accept: . CONST_DEC_PART $end
- expr go to state 6
- @2 go to state 2
+ $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 1
- expr -> 'a' . @1 'b' (rule 2)
-
- $default reduce using rule 1 (@1)
-
- @1 go to state 3
+ 0 $accept: CONST_DEC_PART . $end
+ $end shift, and go to state 5
state 2
- expr -> @2 . 'c' (rule 4)
+ 1 CONST_DEC_PART: CONST_DEC_LIST .
+ 3 CONST_DEC_LIST: CONST_DEC_LIST . CONST_DEC
- 'c' shift, and go to state 4
+ undef_id_tok reduce using rule 4 (@1)
+ $default reduce using rule 1 (CONST_DEC_PART)
+ CONST_DEC go to state 6
+ @1 go to state 4
state 3
- expr -> 'a' @1 . 'b' (rule 2)
-
- 'b' shift, and go to state 5
+ 2 CONST_DEC_LIST: CONST_DEC .
+ $default reduce using rule 2 (CONST_DEC_LIST)
state 4
- expr -> @2 'c' . (rule 4)
-
- $default reduce using rule 4 (expr)
+ 5 CONST_DEC: @1 . undef_id_tok '=' const_id_tok ';'
+ undef_id_tok shift, and go to state 7
state 5
- expr -> 'a' @1 'b' . (rule 2)
-
- $default reduce using rule 2 (expr)
+ 0 $accept: CONST_DEC_PART $end .
+ $default accept
state 6
- $ go to state 7
+ 3 CONST_DEC_LIST: CONST_DEC_LIST CONST_DEC .
+ $default reduce using rule 3 (CONST_DEC_LIST)
state 7
- $ go to state 8
+ 5 CONST_DEC: @1 undef_id_tok . '=' const_id_tok ';'
+ '=' shift, and go to state 8
state 8
- $default accept
+ 5 CONST_DEC: @1 undef_id_tok '=' . const_id_tok ';'
+ const_id_tok shift, and go to state 9
-]])
-AT_CLEANUP
+state 9
+ 5 CONST_DEC: @1 undef_id_tok '=' const_id_tok . ';'
+ ';' shift, and go to state 10
-## -------------------- ##
-## %expect not enough. ##
-## -------------------- ##
-AT_SETUP([%expect not enough])
+state 10
-AT_DATA([input.y],
-[[%token NUM OP
-%expect 0
-%%
-exp: exp OP exp | NUM;
+ 5 CONST_DEC: @1 undef_id_tok '=' const_id_tok ';' .
+
+ $default reduce using rule 5 (CONST_DEC)
]])
-AT_CHECK([bison input.y -o input.c], 1, [],
-[input.y contains 1 shift/reduce conflict.
-expected 0 shift/reduce conflicts
-])
AT_CLEANUP
## --------------- ##
-## %expect right. ##
+## Web2c Actions. ##
## --------------- ##
-AT_SETUP([%expect right])
-
-AT_DATA([input.y],
-[[%token NUM OP
-%expect 1
-%%
-exp: exp OP exp | NUM;
-]])
-
-AT_CHECK([bison input.y -o input.c], 0)
-AT_CLEANUP
+# The generation of the mapping `state -> action' was once wrong in
+# extremely specific situations. web2c.y exhibits this situation.
+# Below is a stripped version of the grammar. It looks like one can
+# simplify it further, but just don't: it is tuned to exhibit a bug,
+# which disapears when applying sane grammar transformations.
+#
+# It used to be wrong on yydefact only:
+#
+# static const yytype_uint8 yydefact[] =
+# {
+# - 2, 0, 1, 0, 0, 2, 3, 2, 5, 4,
+# + 2, 0, 1, 0, 0, 0, 3, 2, 5, 4,
+# 0, 0
+# };
+#
+# but let's check all the tables.
-## ------------------ ##
-## %expect too much. ##
-## ------------------ ##
+AT_SETUP([Web2c Actions])
-AT_SETUP([%expect too much])
+AT_KEYWORDS([report])
AT_DATA([input.y],
-[[%token NUM OP
-%expect 2
+[[%%
+statement: struct_stat;
+struct_stat: /* empty. */ | if else;
+if: "if" "const" "then" statement;
+else: "else" statement;
+%token IF "if";
+%token CONST "const";
+%token THEN "then";
+%token ELSE "else";
%%
-exp: exp OP exp | NUM;
]])
-AT_CHECK([bison input.y -o input.c], 1, [],
-[input.y contains 1 shift/reduce conflict.
-expected 2 shift/reduce conflicts
-])
-AT_CLEANUP
-
+AT_CHECK([bison -v -o input.c input.y])
-## ---------------------- ##
-## Mixing %token styles. ##
-## ---------------------- ##
-
-
-AT_SETUP([Mixing %token styles])
+# Check only the tables. We don't use --no-parser, because it is
+# still to be implemented in the experimental branch of Bison.
+[sed -n 's/ *$//;/^static const.*\[\] =/,/^}/p' input.c >tables.c]
-# Taken from the documentation.
-AT_DATA([input.y],
-[[%token <operator> OR "||"
-%token <operator> LE 134 "<="
-%left OR "<="
-%%
-exp: ;
-%%
+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 yyprhs[] =
+{
+ 0, 0, 3, 5, 6, 9, 14
+};
+static const yytype_int8 yyrhs[] =
+{
+ 8, 0, -1, 9, -1, -1, 10, 11, -1, 3,
+ 4, 5, 8, -1, 6, 8, -1
+};
+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", 0
+};
+static const yytype_uint16 yytoknum[] =
+{
+ 0, 256, 257, 258, 259, 260, 261
+};
+static const yytype_uint8 yyr1[] =
+{
+ 0, 7, 8, 9, 9, 10, 11
+};
+static const yytype_uint8 yyr2[] =
+{
+ 0, 2, 1, 0, 2, 4, 2
+};
+static const yytype_uint8 yydefact[] =
+{
+ 3, 0, 0, 2, 0, 0, 1, 3, 4, 3,
+ 6, 5
+};
+static const yytype_int8 yydefgoto[] =
+{
+ -1, 2, 3, 4, 8
+};
+static const yytype_int8 yypact[] =
+{
+ -2, -1, 4, -8, 0, 2, -8, -2, -8, -2,
+ -8, -8
+};
+static const yytype_int8 yypgoto[] =
+{
+ -8, -7, -8, -8, -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
+};
]])
-AT_CHECK([bison -v input.y -o input.c], 0, ignore, ignore)
-
AT_CLEANUP
-
-## ---------------------- ##
-## %union and --defines. ##
-## ---------------------- ##
+## ------------------------- ##
+## yycheck Bound Violation. ##
+## ------------------------- ##
-AT_SETUP([%union and --defines])
+# _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],
+[%{
+static int yylex (AT_LALR1_CC_IF([int *], [void]));
+AT_LALR1_CC_IF([],
+[#include <stdio.h>
+static void yyerror (const char *);])
+%}
+$1
+%token ARROW INVALID NUMBER STRING DATA
+%defines
+%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
+ ;
+%%
+AT_LALR1_CC_IF(
+[/* A C++ error reporting function. */
+void
+yy::parser::error (const location&, const std::string& m)
+{
+ std::cerr << m << std::endl;
+}
-AT_DATA([union.y],
-[%union
+int
+yyparse ()
{
- int integer;
- char *string ;
+ yy::parser parser;
+ parser.set_debug_level (!!YYDEBUG);
+ return parser.parse ();
}
-%%
-exp: {};
+],
+[static void
+yyerror (const char *s)
+{
+ fprintf (stderr, "%s\n", s);
+}])
+
+static int
+yylex (AT_LALR1_CC_IF([int *lval], [void]))
+[{
+ static int toknum = 0;
+ static int tokens[] =
+ {
+ ':', -1
+ };
+ ]AT_LALR1_CC_IF([*lval = 0; /* Pacify GCC. */])[
+ return tokens[toknum++];
+}]
+
+int
+main (void)
+{
+ return yyparse ();
+}
+])
+])# _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_CHECK([bison -o dancer.c dancer.y])
+AT_LALR1_CC_IF(
+ [AT_CHECK([bison -o dancer.cc dancer.y])
+ AT_COMPILE_CXX([dancer])],
+ [AT_CHECK([bison -o dancer.c dancer.y])
+ AT_COMPILE([dancer])])
+AT_PARSER_CHECK([./dancer], 1, [],
+[syntax error, unexpected ':'
+])
+AT_BISON_OPTION_POPDEFS
+AT_CLEANUP
])
-AT_CHECK([bison --defines union.y])
+AT_CHECK_DANCER()
+AT_CHECK_DANCER([%glr-parser])
+AT_CHECK_DANCER([%skeleton "lalr1.cc"])
-AT_CLEANUP
+## ------------------------------------------ ##
+## Diagnostic that expects two alternatives. ##
+## ------------------------------------------ ##
-## --------------------------------------- ##
-## Duplicate '/' in C comments in %union ##
-## --------------------------------------- ##
+# _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 <stdio.h>
+static void yyerror (const char *);])
+%}
+$1
+%defines
+%error-verbose
+%token A 1000
+%token B
-AT_SETUP([%union and C comments])
+%%
+program: /* empty */
+ | program e ';'
+ | program error ';';
+
+e: e '+' t | t;
+t: A | B;
-AT_DATA([union-comment.y],
-[%union
+%%
+AT_LALR1_CC_IF(
+[/* A C++ error reporting function. */
+void
+yy::parser::error (const location&, const std::string& m)
{
- /* The int. */ int integer;
- /* The string. */ char *string ;
+ std::cerr << m << std::endl;
}
-%%
-exp: {};
-])
-
-AT_CHECK([bison union-comment.y])
-AT_CHECK([fgrep '//*' union-comment.tab.c], [1], [])
+int
+yyparse ()
+{
+ yy::parser parser;
+ return parser.parse ();
+}
+],
+[static void
+yyerror (const char *s)
+{
+ fprintf (stderr, "%s\n", s);
+}])
+
+static int
+yylex (AT_LALR1_CC_IF([int *lval], [void]))
+[{
+ static int toknum = 0;
+ static int tokens[] =
+ {
+ 1000, '+', '+', -1
+ };
+ ]AT_LALR1_CC_IF([*lval = 0; /* Pacify GCC. */])[
+ return tokens[toknum++];
+}]
+
+int
+main (void)
+{
+ return yyparse ();
+}
+])
+])# _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_CHECK([bison -o expect2.c expect2.y])
+AT_LALR1_CC_IF(
+ [AT_CHECK([bison -o expect2.cc expect2.y])
+ AT_COMPILE_CXX([expect2])],
+ [AT_CHECK([bison -o expect2.c expect2.y])
+ AT_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"])
-## ----------------- ##
-## Invalid input 1. ##
-## ----------------- ##
-
-
-AT_SETUP([Invalid input: 1])
-
-AT_DATA([input.y],
-[[%%
-?
-]])
-
-AT_CHECK([bison input.y], [1], [],
-[[input.y:2: invalid input: `?'
-input.y:3: fatal error: no rules in the input grammar
-]])
-
-AT_CLEANUP
-## ----------------- ##
-## Invalid input 2. ##
-## ----------------- ##
+## --------------------------------------------- ##
+## Braced code in declaration in rules section. ##
+## --------------------------------------------- ##
+AT_SETUP([Braced code in declaration in rules section])
-AT_SETUP([Invalid input: 2])
+# Bison once mistook braced code in a declaration in the rules section to be a
+# rule action.
-AT_DATA([input.y],
-[[%%
-default: 'a' }
-]])
+AT_DATA_GRAMMAR([input.y],
+[[%{
+#include <stdio.h>
+static void yyerror (char const *msg);
+static int yylex (void);
+%}
-AT_CHECK([bison input.y], [1], [],
-[[input.y:2: invalid input: `}'
-]])
+%error-verbose
-AT_CLEANUP
+%%
+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';
-## -------------------- ##
-## Invalid %directive. ##
-## -------------------- ##
+%%
+static void
+yyerror (char const *msg)
+{
+ fprintf (stderr, "%s\n", msg);
+}
-AT_SETUP([Invalid %directive])
+static int
+yylex (void)
+{
+ return 'a';
+}
-AT_DATA([input.y],
-[[%invalid
+int
+main (void)
+{
+ yydebug = 1;
+ return !yyparse ();
+}
]])
-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_CHECK([bison -t -o input.c input.y])
+AT_COMPILE([input])
+AT_PARSER_CHECK([./input], 0,
+[[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 22):
+-> $$ = 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_CLEANUP
-## --------------------- ##
-## Invalid CPP headers. ##
-## --------------------- ##
+## --------------------------------- ##
+## String alias declared after use. ##
+## --------------------------------- ##
-# AT_TEST_CPP_GUARD_H([INPUT-FILE-BASE)
-# -------------------------------------
-m4_define([AT_TEST_CPP_GUARD_H],
-[AT_SETUP([Invalid CPP guards: $1])
+AT_SETUP([String alias declared after use])
-# Possibly create inner directories.
-dirname=`AS_DIRNAME([$1])`
-AS_MKDIR_P([$dirname])
+# 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.
-AT_DATA([$1.y],
-[%%
-dummy:
-])
-
-AT_CHECK([bison --defines=$1.h $1.y])
+AT_DATA([input.y],
+[[%%
+start: 'a' "A" 'b';
+%token 'a' "A";
+]])
-# CPP should be happy with it.
-AT_CHECK([$CC -E $1.h], 0, [ignore])
+AT_CHECK([bison -t -o input.c input.y])
AT_CLEANUP
-])
-AT_TEST_CPP_GUARD_H([input/input])
-AT_TEST_CPP_GUARD_H([9foo])
-## ---------------- ##
-## Broken Closure. ##
-## ---------------- ##
+## --------------------------- ##
+## Undeclared string literal. ##
+## --------------------------- ##
-# 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.
+AT_SETUP([Undeclared string literal])
-AT_SETUP([Broken Closure])
+# Bison once allowed a string literal to be used in the grammar without any
+# declaration assigning it as an alias of another token.
AT_DATA([input.y],
[[%%
-a: b
-b: c
-c: d
-d: e
-e: f
-f: g
-g: h
-h: 'h'
+start: "abc";
]])
-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_CHECK([bison -t -o input.c input.y], [1], [],
+[[input.y:2.8-12: "abc" undeclared
]])
AT_CLEANUP
projects / bison.git / blobdiff