# Bison Regressions. -*- Autotest -*-
-# Copyright 2001 Free Software Foundation, Inc.
+# Copyright (C) 2001, 2002, 2003 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
AT_BANNER([[Regression tests.]])
+
+## ------------------------- ##
+## Early token definitions. ##
+## ------------------------- ##
+
+
+AT_SETUP([Early token definitions])
+
+# 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);
+%}
+
+%union
+{
+ int val;
+};
+%{
+#ifndef MY_TOKEN
+# error "MY_TOKEN not defined."
+#endif
+%}
+%token MY_TOKEN
+%%
+exp: MY_TOKEN;
+%%
+]])
+
+AT_CHECK([bison -o input.c input.y])
+AT_COMPILE([input.o], [-c input.c])
+
+AT_CLEANUP
+
+
+
+## ---------------- ##
+## Braces parsing. ##
+## ---------------- ##
+
+
+AT_SETUP([Braces parsing])
+
+AT_DATA([input.y],
+[[/* Bison used to swallow the character after `}'. */
+
+%%
+exp: { tests = {{{{{{{{{{}}}}}}}}}}; };
+%%
+]])
+
+AT_CHECK([bison -v -o input.c input.y])
+
+AT_CHECK([grep 'tests = {{{{{{{{{{}}}}}}}}}};' input.c], 0, [ignore])
+
+AT_CLEANUP
+
+
## ------------------ ##
## Duplicate string. ##
## ------------------ ##
AT_SETUP([Duplicate string])
-AT_DATA([duplicate.y],
+AT_DATA([input.y],
[[/* `Bison -v' used to dump core when two tokens are defined with the same
string, as LE and GE below. */
%%
]])
-AT_CHECK([bison -v duplicate.y -o duplicate.c], 0, ignore, ignore)
+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
+]])
+
+AT_CLEANUP
-AT_CLEANUP([duplicate.*])
+## ------------------- ##
+## Rule Line Numbers. ##
+## ------------------- ##
-## ------------------------- ##
-## Unresolved SR Conflicts. ##
-## ------------------------- ##
+AT_SETUP([Rule Line Numbers])
-AT_SETUP([Unresolved SR Conflicts])
+AT_KEYWORDS([report])
AT_DATA([input.y],
-[[%token NUM OP
-%%
-exp: exp OP exp | NUM;
+[[%%
+expr:
+'a'
+
+{
+
+}
+
+'b'
+
+{
+
+}
+
+|
+
+
+{
+
+
+}
+
+'c'
+
+{
+
+};
]])
-AT_CHECK([bison input.y -o input.c -v], 0, [],
-[input.y contains 1 shift/reduce conflict.
-])
+AT_CHECK([bison -o input.c -v input.y])
# Check the contents of the report.
AT_CHECK([cat input.output], [],
-[[State 4 contains 1 shift/reduce conflict.
+[[Grammar
+
+ 0 $accept: expr $end
+
+ 1 @1: /* empty */
-Grammar
+ 2 expr: 'a' @1 'b'
+
+ 3 @2: /* empty */
+
+ 4 expr: @2 'c'
-rule 1 exp -> exp OP exp
-rule 2 exp -> NUM
Terminals, with rules where they appear
-$ (-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
+
+ '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
- 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
- OP [reduce using rule 1 (exp)]
- $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 6
+state 7
- $default accept
+ 2 expr: 'a' @1 'b' .
+
+ $default reduce using rule 2 (expr)
]])
-AT_CLEANUP(input.c input.output)
+AT_CLEANUP
+
-## --------------------- ##
-## Solved SR Conflicts. ##
-## --------------------- ##
+## ---------------------- ##
+## Mixing %token styles. ##
+## ---------------------- ##
-AT_SETUP([Solved SR Conflicts])
+AT_SETUP([Mixing %token styles])
+
+# Taken from the documentation.
AT_DATA([input.y],
-[[%token NUM OP
-%right OP
+[[%token <operator> OR "||"
+%token <operator> LE 134 "<="
+%left OR "<="
+%%
+exp: ;
%%
-exp: exp OP exp | NUM;
]])
-AT_CHECK([bison input.y -o input.c -v], 0, [], [])
+AT_CHECK([bison -v -o input.c input.y])
-# Check the contents of the report.
-AT_CHECK([cat input.output], [],
-[[Conflict in state 4 between rule 1 and token OP resolved as shift.
+AT_CLEANUP
-Grammar
-rule 1 exp -> exp OP exp
-rule 2 exp -> NUM
-Terminals, with rules where they appear
+## ---------------- ##
+## Invalid inputs. ##
+## ---------------- ##
-$ (-1)
-error (256)
-NUM (257) 2
-OP (258) 1
-Nonterminals, with rules where they appear
+AT_SETUP([Invalid inputs])
-exp (5)
- on left: 1 2, on right: 1
+AT_DATA([input.y],
+[[%%
+?
+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
+]])
-state 0
+AT_CLEANUP
- NUM shift, and go to state 1
- exp go to state 2
+## ------------------- ##
+## Token definitions. ##
+## ------------------- ##
-state 1
+AT_SETUP([Token definitions])
- exp -> NUM . (rule 2)
+# Bison managed, when fed with `%token 'f' "f"' to #define 'f'!
+AT_DATA_GRAMMAR([input.y],
+[%{
+void yyerror (const char *s);
+int yylex (void);
+%}
+[%token MYEOF 0 "end of file"
+%token 'a' "a"
+%token b "b"
+%token c 'c'
+%token 'd' d
+%%
+exp: "a";
+]])
- $default reduce using rule 2 (exp)
+AT_CHECK([bison -o input.c input.y])
+AT_COMPILE([input.o], [-c input.c])
+AT_CLEANUP
-state 2
+## -------------------- ##
+## Characters Escapes. ##
+## -------------------- ##
+
+
+AT_SETUP([Characters Escapes])
+
+AT_DATA_GRAMMAR([input.y],
+[%{
+void yyerror (const char *s);
+int yylex (void);
+%}
+[%%
+exp:
+ '\'' "\'"
+| '\"' "\""
+| '"' "'"
+;
+]])
+# Pacify font-lock-mode: "
- exp -> exp . OP exp (rule 1)
+AT_CHECK([bison -o input.c input.y])
+AT_COMPILE([input.o], [-c input.c])
+AT_CLEANUP
- $ go to state 5
- OP shift, and go to state 3
+## -------------- ##
+## Web2c Report. ##
+## -------------- ##
-state 3
+# 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.
- exp -> exp OP . exp (rule 1)
+AT_SETUP([Web2c Report])
- NUM shift, and go to state 1
+AT_KEYWORDS([report])
- exp go to state 4
+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
-state 4
+ 0 $accept: CONST_DEC_PART $end
- exp -> exp . OP exp (rule 1)
- exp -> exp OP exp . (rule 1)
+ 1 CONST_DEC_PART: CONST_DEC_LIST
- OP shift, and go to state 3
+ 2 CONST_DEC_LIST: CONST_DEC
+ 3 | CONST_DEC_LIST CONST_DEC
- $default reduce using rule 1 (exp)
+ 4 @1: /* empty */
+ 5 CONST_DEC: @1 undef_id_tok '=' const_id_tok ';'
-state 5
+Terminals, with rules where they appear
- $ go to state 6
+$end (0) 0
+';' (59) 5
+'=' (61) 5
+error (256)
+undef_id_tok (258) 5
+const_id_tok (259) 5
+Nonterminals, with rules where they appear
-state 6
+$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
- $default accept
-]])
-AT_CLEANUP(input.c input.output)
+state 0
+ 0 $accept: . CONST_DEC_PART $end
-## ---------------------- ##
-## Mixing %token styles. ##
-## ---------------------- ##
+ $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
-AT_SETUP([Mixing %token styles])
-# Taken from the documentation.
-AT_DATA([input.y],
-[[%token <operator> OR "||"
-%token <operator> LE 134 "<="
-%left OR "<="
-%%
-exp: ;
-%%
-]])
+state 1
-AT_CHECK([bison -v input.y -o input.c], 0, ignore, ignore)
+ 0 $accept: CONST_DEC_PART . $end
-AT_CLEANUP([input.*])
+ $end shift, and go to state 5
+state 2
-## ---------------------- ##
-## %union and --defines. ##
-## ---------------------- ##
+ 1 CONST_DEC_PART: CONST_DEC_LIST .
+ 3 CONST_DEC_LIST: CONST_DEC_LIST . CONST_DEC
+ undef_id_tok reduce using rule 4 (@1)
+ $default reduce using rule 1 (CONST_DEC_PART)
-AT_SETUP([%union and --defines])
+ CONST_DEC go to state 6
+ @1 go to state 4
-AT_DATA([union.y],
-[%union
-{
- int integer;
- char *string ;
-}
-%%
-exp: {};
-])
-AT_CHECK([bison --defines union.y])
+state 3
-AT_CLEANUP([union.*])
+ 2 CONST_DEC_LIST: CONST_DEC .
+ $default reduce using rule 2 (CONST_DEC_LIST)
-## --------------------------------------- ##
-## Duplicate '/' in C comments in %union ##
-## --------------------------------------- ##
+state 4
-AT_SETUP([%union and C comments])
+ 5 CONST_DEC: @1 . undef_id_tok '=' const_id_tok ';'
-AT_DATA([union-comment.y],
-[%union
-{
- /* The int. */ int integer;
- /* The string. */ char *string ;
-}
-%%
-exp: {};
-])
+ undef_id_tok shift, and go to state 7
-AT_CHECK([bison union-comment.y])
-AT_CHECK([fgrep '//*' union-comment.tab.c], [1], [])
-AT_CLEANUP([union-comment.*])
+state 5
+ 0 $accept: CONST_DEC_PART $end .
-## ----------------- ##
-## Invalid input 1. ##
-## ----------------- ##
+ $default accept
-AT_SETUP([Invalid input: 1])
+state 6
-AT_DATA([input.y],
-[[%%
-?
-]])
+ 3 CONST_DEC_LIST: CONST_DEC_LIST CONST_DEC .
+
+ $default reduce using rule 3 (CONST_DEC_LIST)
+
+
+state 7
+
+ 5 CONST_DEC: @1 undef_id_tok . '=' const_id_tok ';'
+
+ '=' shift, and go to state 8
-AT_CHECK([bison input.y], [1], [],
-[input.y:2: invalid input: `?'
-input.y:3: fatal error: no rules in the input grammar
-])
+
+state 8
+
+ 5 CONST_DEC: @1 undef_id_tok '=' . const_id_tok ';'
+
+ const_id_tok shift, and go to state 9
+
+
+state 9
+
+ 5 CONST_DEC: @1 undef_id_tok '=' const_id_tok . ';'
+
+ ';' shift, and go to state 10
+
+
+state 10
+
+ 5 CONST_DEC: @1 undef_id_tok '=' const_id_tok ';' .
+
+ $default reduce using rule 5 (CONST_DEC)
+]])
AT_CLEANUP
-## ----------------- ##
-## Invalid input 2. ##
-## ----------------- ##
+## --------------- ##
+## Web2c Actions. ##
+## --------------- ##
+
+# The generation of the mapping `state -> action' was once wrong in
+# extremely specific situations. web2c.y exhibits this situation.
+# Below is a stripped version of the grammar. It looks like one can
+# simplify it further, but just don't: it is tuned to exhibit a bug,
+# which disapears when applying sane grammar transformations.
+#
+# It used to be wrong on yydefact only:
+#
+# static const short yydefact[] =
+# {
+# - 2, 0, 1, 0, 0, 2, 3, 2, 5, 4,
+# + 2, 0, 1, 0, 0, 0, 3, 2, 5, 4,
+# 0, 0
+# };
+#
+# but let's check all the tables.
+
+AT_SETUP([Web2c Actions])
-AT_SETUP([Invalid input: 2])
+AT_KEYWORDS([report])
AT_DATA([input.y],
[[%%
-default: 'a' }
+statement: struct_stat;
+struct_stat: /* empty. */ | if else;
+if: "if" "const" "then" statement;
+else: "else" statement;
+%%
]])
-AT_CHECK([bison input.y], [1], [],
-[input.y:2: invalid input: `}'
-])
-
-AT_CLEANUP
+AT_CHECK([bison -v -o input.c input.y])
+# Check only the tables. We don't use --no-parser, because it is
+# still to be implemented in the experimental branch of Bison.
+[sed -n 's/ *$//;/^static const.*\[\] =/,/^}/p' input.c >tables.c]
+AT_CHECK([[cat tables.c]], 0,
+[[static const unsigned char 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 unsigned char yyprhs[] =
+{
+ 0, 0, 3, 5, 6, 9, 14
+};
+static const yysigned_char yyrhs[] =
+{
+ 8, 0, -1, 9, -1, -1, 10, 11, -1, 3,
+ 4, 5, 8, -1, 6, 8, -1
+};
+static const unsigned char 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 unsigned short yytoknum[] =
+{
+ 0, 256, 257, 258, 259, 260, 261
+};
+static const unsigned char yyr1[] =
+{
+ 0, 7, 8, 9, 9, 10, 11
+};
+static const unsigned char yyr2[] =
+{
+ 0, 2, 1, 0, 2, 4, 2
+};
+static const unsigned char yydefact[] =
+{
+ 3, 0, 0, 2, 0, 0, 1, 3, 4, 3,
+ 6, 5
+};
+static const yysigned_char yydefgoto[] =
+{
+ -1, 2, 3, 4, 8
+};
+static const yysigned_char yypact[] =
+{
+ -2, -1, 4, -8, 0, 2, -8, -2, -8, -2,
+ -8, -8
+};
+static const yysigned_char yypgoto[] =
+{
+ -8, -7, -8, -8, -8
+};
+static const unsigned char yytable[] =
+{
+ 10, 1, 11, 5, 6, 0, 7, 9
+};
+static const yysigned_char yycheck[] =
+{
+ 7, 3, 9, 4, 0, -1, 6, 5
+};
+static const unsigned char yystos[] =
+{
+ 0, 3, 8, 9, 10, 4, 0, 6, 11, 5,
+ 8, 8
+};
+]])
-## --------------------- ##
-## Invalid CPP headers. ##
-## --------------------- ##
+AT_CLEANUP
-# AT_TEST_CPP_GUARD_H([INPUT-FILE-BASE)
-# -------------------------------------
-m4_define([AT_TEST_CPP_GUARD_H],
-[AT_SETUP([Invalid CPP guards: $1])
-# possibly create and nuke inner directories.
-m4_bmatch([$1], [[/]],
-[dirname=`AS_DIRNAME([$1])`
-AS_MKDIR_P([$dirname])
-AT_CLEANUP_FILES([$dirname])])
+## ------------------------- ##
+## yycheck Bound Violation. ##
+## ------------------------- ##
-AT_DATA([$1.y],
-[%%
-dummy:
-])
-AT_CHECK([bison --defines=$1.h $1.y])
+# _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],
+[%{
+AT_LALR1_CC_IF(
+[static int yylex (int *lval);],
+[#include <stdio.h>
+static void yyerror (const char *s);
+static int yylex ();])
+%}
+$1
+%token ARROW INVALID NUMBER STRING DATA
+%defines
+%verbose
+%error-verbose
+/* Grammar follows */
+%%
+line: header body
+ ;
+
+header: '<' from ARROW to '>' type ':'
+ | '<' ARROW to '>' type ':'
+ | ARROW to type ':'
+ | type ':'
+ | '<' '>'
+ ;
+
+from: DATA
+ | STRING
+ | INVALID
+ ;
+
+to: DATA
+ | STRING
+ | INVALID
+ ;
+
+type: DATA
+ | STRING
+ | INVALID
+ ;
+
+body: /* empty */
+ | body member
+ ;
+
+member: STRING
+ | DATA
+ | '+' NUMBER
+ | '-' NUMBER
+ | NUMBER
+ | INVALID
+ ;
+%%
+AT_LALR1_CC_IF(
+[/* Currently, print_ is required in C++. */
+void
+yy::Parser::print_ ()
+{
+}
-# CPP should be happy with it.
-AT_CHECK([$CC -E $1.h], 0, [ignore])
+/* A C++ error reporting function. */
+void
+yy::Parser::error_ ()
+{
+ std::cerr << message << std::endl;
+}
-AT_CLEANUP($1.*)
+int
+yyparse (void)
+{
+ yy::Parser parser (!!YYDEBUG);
+ return parser.parse ();
+}
+],
+[static void
+yyerror (const char *s)
+{
+ fprintf (stderr, "%s\n", s);
+}])
+
+static int
+yylex (AT_LALR1_CC_IF([int *lval]))
+[{
+ static int toknum = 0;
+ int tokens[] =
+ {
+ ':', -1
+ };
+ 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_COMPILE_CXX([dancer])],
+ [AT_COMPILE([dancer])])
+AT_PARSER_CHECK([./dancer], 1, [],
+[syntax error, unexpected ':'
+])
+AT_BISON_OPTION_POPDEFS
+AT_CLEANUP
])
-AT_TEST_CPP_GUARD_H([input/input])
-AT_TEST_CPP_GUARD_H([9foo])
+AT_CHECK_DANCER()
+AT_CHECK_DANCER([%glr-parser])
+AT_CHECK_DANCER([%skeleton "lalr1.cc"])
projects / bison.git / blobdiff