X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/e0c40012bd7c94c88a0bb9886e37de9622729d9a..abd189e8dc6ca848f038da12e4110d6192374b82:/tests/regression.at?ds=inline diff --git a/tests/regression.at b/tests/regression.at index 68f02adc..c07b7442 100644 --- a/tests/regression.at +++ b/tests/regression.at @@ -1,23 +1,181 @@ # 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) 2001-2011 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 . AT_BANNER([[Regression tests.]]) + +## ------------------ ## +## Trivial grammars. ## +## ------------------ ## + +AT_SETUP([Trivial grammars]) + +AT_DATA_GRAMMAR([input.y], +[[%{ +void yyerror (char const *); +int yylex (void); +#define YYSTYPE int * +%} + +%error-verbose + +%% + +program: 'x'; +]]) + +AT_BISON_CHECK([-o input.c input.y]) +AT_COMPILE([input.o], [-c input.c]) +AT_COMPILE([input.o], [-DYYDEBUG -c input.c]) + +AT_CLEANUP + + + +## ----------------- ## +## 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 program + +%% + +program: { $$ = ""; }; +]]) + +AT_BISON_CHECK([-o input.c input.y]) +AT_COMPILE([input.o], [-c input.c]) + +AT_CLEANUP + + + +## ------------------------------------- ## +## Early token definitions with --yacc. ## +## ------------------------------------- ## + + +AT_SETUP([Early token definitions with --yacc]) + +# 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_BISON_CHECK([-y -o input.c input.y]) +AT_COMPILE([input.o], [-c input.c]) + +AT_CLEANUP + + + +## ---------------------------------------- ## +## Early token definitions without --yacc. ## +## ---------------------------------------- ## + + +AT_SETUP([Early token definitions without --yacc]) + +# 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], +[[%{ +#include +void yyerror (const char *s); +int yylex (void); +void print_my_token (void); +%} + +%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_BISON_CHECK([-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_BISON_CHECK([-v -o input.c input.y]) + +AT_CHECK([grep 'tests = {{{{{{{{{{}}}}}}}}}};' input.c], 0, [ignore]) + +AT_CLEANUP + + ## ------------------ ## ## Duplicate string. ## ## ------------------ ## @@ -25,7 +183,7 @@ AT_BANNER([[Regression tests.]]) 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. */ @@ -38,421 +196,1478 @@ exp: '(' exp ')' | NUM ; %% ]]) -AT_CHECK([bison -v duplicate.y -o duplicate.c], 0, ignore, ignore) +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 +]]) AT_CLEANUP -## ------------------------- ## -## Unresolved SR Conflicts. ## -## ------------------------- ## +## ------------------- ## +## Rule Line Numbers. ## +## ------------------- ## -AT_SETUP([Unresolved SR Conflicts]) +AT_SETUP([Rule Line Numbers]) + +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_BISON_CHECK([-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: $@2 'c' - Number, Line, Rule - 1 3 exp -> exp OP exp - 2 3 exp -> NUM Terminals, with rules where they appear -$ (-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 -## --------------------- ## -## Solved SR Conflicts. ## -## --------------------- ## -AT_SETUP([Solved SR Conflicts]) +## ---------------------- ## +## Mixing %token styles. ## +## ---------------------- ## + + +AT_SETUP([Mixing %token styles]) +# Taken from the documentation. AT_DATA([input.y], -[[%token NUM OP -%right OP +[[%token OR "||" +%token LE 134 "<=" +%left OR "<=" +%% +exp: ; %% -exp: exp OP exp | NUM; ]]) -AT_CHECK([bison input.y -o input.c -v], 0, [], []) +AT_BISON_CHECK([-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 + + + +## ---------------- ## +## Invalid inputs. ## +## ---------------- ## + + +AT_SETUP([Invalid inputs]) + +AT_DATA([input.y], +[[%% +? +default: 'a' } +%& +%a-does-not-exist +%- +%{ +]]) + +AT_BISON_CHECK([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 + + +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: syntax error, unexpected %initial-action, expecting {...} +]]) + +AT_CLEANUP + + + +## ------------------- ## +## Token definitions. ## +## ------------------- ## + + +AT_SETUP([Token definitions]) + +# Bison managed, when fed with `%token 'f' "f"' to #define 'f'! +AT_DATA_GRAMMAR([input.y], +[%{ +#include +#include +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??!" +%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) +{ + static int called; + if (called++) + abort (); + return SPECIAL; +} + +int +main (void) +{ + return yyparse (); +} +]]) + +# 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 +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 +]]) +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_DATA_GRAMMAR([input.y], +[%{ +void yyerror (const char *s); +int yylex (void); +%} +[%% +exp: + '\'' "\'" +| '\"' "\"" +| '"' "'" +; +]]) +# Pacify font-lock-mode: " + +AT_BISON_CHECK([-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 + +%% +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 + + 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 */ -Grammar + 5 CONST_DEC: $@1 undef_id_tok '=' const_id_tok ';' - Number, Line, Rule - 1 4 exp -> exp OP exp - 2 4 exp -> NUM Terminals, with rules where they appear -$ (-1) +$end (0) 0 +';' (59) 5 +'=' (61) 5 error (256) -NUM (257) 2 -OP (258) 1 +undef_id_tok (258) 5 +const_id_tok (259) 5 + Nonterminals, with rules where they appear -exp (5) - on left: 1 2, on right: 1 +$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 - NUM shift, and go to state 1 + 0 $accept: . CONST_DEC_PART $end - exp 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 - exp -> NUM . (rule 2) - - $default reduce using rule 2 (exp) + 0 $accept: CONST_DEC_PART . $end + $end shift, and go to state 5 state 2 - exp -> exp . OP exp (rule 1) + 1 CONST_DEC_PART: CONST_DEC_LIST . + 3 CONST_DEC_LIST: CONST_DEC_LIST . CONST_DEC - $ go to state 5 - OP shift, and go to state 3 + 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 - exp -> exp OP . exp (rule 1) + 2 CONST_DEC_LIST: CONST_DEC . - NUM shift, and go to state 1 + $default reduce using rule 2 (CONST_DEC_LIST) - exp go to state 4 +state 4 + 5 CONST_DEC: $@1 . undef_id_tok '=' const_id_tok ';' -state 4 + undef_id_tok 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: CONST_DEC_PART $end . + $default accept -state 5 +state 6 - $ go to state 6 + 3 CONST_DEC_LIST: CONST_DEC_LIST CONST_DEC . + $default reduce using rule 3 (CONST_DEC_LIST) -state 6 +state 7 - $default accept -]]) + 5 CONST_DEC: $@1 undef_id_tok . '=' const_id_tok ';' -AT_CLEANUP + '=' shift, and go to state 8 +state 8 -## -------------------- ## -## %expect not enough. ## -## -------------------- ## + 5 CONST_DEC: $@1 undef_id_tok '=' . const_id_tok ';' -AT_SETUP([%expect not enough]) + const_id_tok shift, and go to state 9 -AT_DATA([input.y], -[[%token NUM OP -%expect 0 -%% -exp: exp OP exp | NUM; + +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 ';' . + + $default reduce using rule 5 (CONST_DEC) ]]) -AT_CHECK([bison input.y -o input.c], 1, [], -[input.y contains 1 shift/reduce conflict. -expected 0 shift/reduce conflicts -]) AT_CLEANUP ## --------------- ## -## %expect right. ## +## Web2c Actions. ## ## --------------- ## -AT_SETUP([%expect right]) +# The generation of the mapping `state -> action' was once wrong in +# extremely specific situations. web2c.y exhibits this situation. +# Below is a stripped version of the grammar. It looks like one can +# simplify it further, but just don't: it is tuned to exhibit a bug, +# which disapears when applying sane grammar transformations. +# +# It used to be wrong on yydefact only: +# +# static const yytype_uint8 yydefact[] = +# { +# - 2, 0, 1, 0, 0, 2, 3, 2, 5, 4, +# + 2, 0, 1, 0, 0, 0, 3, 2, 5, 4, +# 0, 0 +# }; +# +# but let's check all the tables. + + +AT_SETUP([Web2c Actions]) + +AT_KEYWORDS([report]) AT_DATA([input.y], -[[%token NUM OP -%expect 1 +[[%% +statement: struct_stat; +struct_stat: /* empty. */ | if else; +if: "if" "const" "then" statement; +else: "else" statement; %% -exp: exp OP exp | NUM; ]]) -AT_CHECK([bison input.y -o input.c], 0) +AT_BISON_CHECK([-v -o input.c input.y]) + +# Check only the tables. +[sed -n 's/ *$//;/^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 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_CLEANUP -## ------------------ ## -## %expect too much. ## -## ------------------ ## +## ------------------------- ## +## yycheck Bound Violation. ## +## ------------------------- ## -AT_SETUP([%expect too much]) -AT_DATA([input.y], -[[%token NUM OP -%expect 2 +# _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 +#include +static void yyerror (const char *);]) +%} +$1 +%token ARROW INVALID NUMBER STRING DATA +%defines +%verbose +%error-verbose +/* Grammar follows */ %% -exp: exp OP exp | NUM; -]]) +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; +} + +int +yyparse () +{ + yy::parser parser; +#if YYDEBUG + parser.set_debug_level (YYDEBUG); +#endif + 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 const tokens[] = + { + ':', -1 + }; + static size_t toknum; + ]AT_LALR1_CC_IF([*lval = 0; /* Pacify GCC. */])[ + if (! (toknum < sizeof tokens / sizeof *tokens)) + abort (); + 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_BISON_CHECK([-o dancer.c dancer.y]) +AT_FULL_COMPILE([dancer]) +AT_PARSER_CHECK([./dancer], 1, [], +[syntax error, unexpected ':' +]) +AT_BISON_OPTION_POPDEFS +AT_CLEANUP +]) + +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 +#include +static void yyerror (const char *);]) +%} +$1 +%defines +%error-verbose +%token A 1000 +%token B + +%% +program: /* empty */ + | program e ';' + | program error ';'; + +e: e '+' t | t; +t: A | B; -AT_CHECK([bison input.y -o input.c], 1, [], -[input.y contains 1 shift/reduce conflict. -expected 2 shift/reduce conflicts +%% +AT_LALR1_CC_IF( +[/* A C++ error reporting function. */ +void +yy::parser::error (const location&, const std::string& m) +{ + std::cerr << m << std::endl; +} + +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 const tokens[] = + { + 1000, '+', '+', -1 + }; + static size_t toknum; + ]AT_LALR1_CC_IF([*lval = 0; /* Pacify GCC. */])[ + if (! (toknum < sizeof tokens / sizeof *tokens)) + abort (); + 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_BISON_CHECK([-o expect2.c expect2.y]) +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"]) -## ---------------------- ## -## Mixing %token styles. ## -## ---------------------- ## -AT_SETUP([Mixing %token styles]) +## --------------------------------------------- ## +## 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_DATA_GRAMMAR([input.y], +[[%{ +#include +static void yyerror (char const *msg); +static int yylex (void); +%} + +%error-verbose -# Taken from the documentation. -AT_DATA([input.y], -[[%token OR "||" -%token LE 134 "<=" -%left OR "<=" %% -exp: ; + +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'; + %% + +static void +yyerror (char const *msg) +{ + fprintf (stderr, "%s\n", msg); +} + +static int +yylex (void) +{ + return 'a'; +} + +int +main (void) +{ + yydebug = 1; + return !yyparse (); +} ]]) -AT_CHECK([bison -v input.y -o input.c], 0, ignore, ignore) +AT_BISON_CHECK([-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 20): +-> $$ = nterm start () +Stack now 0 +Entering state 1 +Reading a token: Next token is token 'a' (PRINTER) +syntax error, unexpected 'a', expecting $end +Error: popping nterm start () +Stack now 0 +Cleanup: discarding lookahead token 'a' (PRINTER) +DESTRUCTOR +Stack now 0 +]]) AT_CLEANUP -## ---------------------- ## -## %union and --defines. ## -## ---------------------- ## +## --------------------------------- ## +## String alias declared after use. ## +## --------------------------------- ## + +AT_SETUP([String alias declared after use]) + +# Bison once incorrectly asserted that the symbol number for either a token or +# its alias was the highest symbol number so far at the point of the alias +# declaration. That was true unless the declaration appeared after their first +# uses and other tokens appeared in between. + +AT_DATA([input.y], +[[%% +start: 'a' "A" 'b'; +%token 'a' "A"; +]]) + +AT_BISON_CHECK([-t -o input.c input.y]) + +AT_CLEANUP + + + +## -------------------------------- ## +## Extra lookahead sets in report. ## +## -------------------------------- ## +AT_SETUP([[Extra lookahead sets in report]]) -AT_SETUP([%union and --defines]) +# Bison prints each reduction's lookahead set only next to the associated +# state's one item that (1) is associated with the same rule as the reduction +# and (2) has its dot at the end of its RHS. Previously, Bison also +# erroneously printed the lookahead set next to all of the state's other items +# associated with the same rule. This bug affected only the `.output' file and +# not the generated parser source code. -AT_DATA([union.y], -[%union +AT_DATA([[input.y]], +[[%% +start: a | 'a' a 'a' ; +a: 'a' ; +]]) + +AT_BISON_CHECK([[--report=all input.y]]) +AT_CHECK([[sed -n '/^state 1$/,/^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_DATA_GRAMMAR([input.y], +[[%{ + #include + void yyerror (char const *); + int yylex (void); +%} + +%error-verbose +%right END 0 +%left TK1 1 TK2 2 "tok alias" 3 + +%% + +start: + TK1 sr_conflict "tok alias" + | start %prec END + ; +sr_conflict: + TK2 + | TK2 "tok alias" + ; + +%% + +void +yyerror (char const *msg) { - int integer; - char *string ; + fprintf (stderr, "%s\n", msg); } -%% -exp: {}; -]) -AT_CHECK([bison --defines union.y]) +int +yylex (void) +{ + static int const input[] = { 1, 2, 3, 0 }; + static int const *inputp = input; + return *inputp++; +} + +int +main (void) +{ + return yyparse (); +} +]]) + +AT_BISON_CHECK([[-o input.c input.y]], [[0]],, +[[input.y:23.5-19: warning: rule useless in parser due to conflicts: start: start +input.y:27.5-19: warning: rule useless in parser due to conflicts: sr_conflict: TK2 "tok alias" +]]) +AT_COMPILE([[input]]) +AT_PARSER_CHECK([[./input]]) AT_CLEANUP + +## --------------------------- ## +## 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 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_BISON_CHECK([[-o input.c -Dlr.type=ielr input.y]]) +[mv input.c ielr.c] +AT_CHECK([[diff -u lalr.c ielr.c]]) + +AT_CLEANUP + + + ## --------------------------------------- ## -## Duplicate '/' in C comments in %union ## +## %error-verbose and YYSTACK_USE_ALLOCA. ## ## --------------------------------------- ## +AT_SETUP([[%error-verbose and YYSTACK_USE_ALLOCA]]) -AT_SETUP([%union and C comments]) - -AT_DATA([union-comment.y], -[%union -{ - /* The int. */ int integer; - /* The string. */ char *string ; +AT_DATA_GRAMMAR([input.y], +[[%code { + #include + void yyerror (char const *); + int yylex (void); + #define YYSTACK_USE_ALLOCA 1 } + +%error-verbose + %% -exp: {}; -]) -AT_CHECK([bison union-comment.y]) -AT_CHECK([fgrep '//*' union-comment.tab.c], [1], []) +start: check syntax_error syntax_error ; -AT_CLEANUP +check: +{ + 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' +; -## ----------------- ## -## Invalid input 1. ## -## ----------------- ## +%% +void +yyerror (char const *msg) +{ + fprintf (stderr, "%s\n", msg); +} -AT_SETUP([Invalid input: 1]) +int +yylex (void) +{ + /* Induce two syntax error messages (which requires full error + recovery by shifting 3 tokens) in order to detect any loss of the + reallocated buffer. */ + static char const *input = "abc"; + return *input++; +} -AT_DATA([input.y], -[[%% -? +int +main (void) +{ + return yyparse (); +} ]]) -AT_CHECK([bison input.y], [1], [], -[[input.y:2: invalid input: `?' -input.y:3: fatal error: no rules in the input grammar +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 -## ----------------- ## -## Invalid input 2. ## -## ----------------- ## +## ------------------------- ## +## %error-verbose overflow. ## +## ------------------------- ## -AT_SETUP([Invalid input: 2]) +# 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([[%error-verbose overflow]]) + +AT_DATA_GRAMMAR([input.y], +[[%code { + #include + void yyerror (char const *); + int yylex (void); + + /* 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 + #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([input.y], -[[%% -default: 'a' } +%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: +{ + 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" +; + +%% + +void +yyerror (char const *msg) +{ + fprintf (stderr, "%s\n", msg); +} + +int +yylex (void) +{ + /* Induce two syntax error messages (which requires full error + recovery by shifting 3 tokens). */ + static char const *input = "abc"; + return *input++; +} + +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_CHECK([bison input.y], [1], [], -[[input.y:2: invalid input: `}' +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_CLEANUP -## -------------------- ## -## Invalid %directive. ## -## -------------------- ## +## ------------------------ ## +## LAC: Exploratory stack. ## +## ------------------------ ## +AT_SETUP([[LAC: Exploratory stack]]) -AT_SETUP([Invalid %directive]) +m4_pushdef([AT_LAC_CHECK], [ -AT_DATA([input.y], -[[%invalid +AT_BISON_OPTION_PUSHDEFS([$1]) + +AT_DATA_GRAMMAR([input.y], +[[%code { + #include + void yyerror (char const *); + int yylex (]AT_PURE_IF([[YYSTYPE *]], [[void]])[); +} + +]$1[ +%error-verbose +%token 'c' + +%% + +// 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"); } ; + +%% + +void +yyerror (char const *msg) +{ + fprintf (stderr, "%s\n", msg); +} + +int +yylex (]AT_PURE_IF([[YYSTYPE *v]], [[void]])[) +{ + static char const *input = "bbbbc";]AT_PURE_IF([[ + *v = 0;]])[ + return *input++; +} + +int +main (void) +{ + yydebug = 1; + return yyparse (); +} +]]) + +AT_BISON_CHECK([[-Dparse.lac=full -Dparse.lac.es-capacity-initial=1 \ + -Dparse.lac.memory-trace=full \ + -t -o input.c input.y]], [[0]], [], +[[input.y: conflicts: 21 shift/reduce ]]) +AT_COMPILE([[input]]) +AT_PARSER_CHECK([[./input > stdout.txt 2> stderr.txt]], [[1]]) -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 +# 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 CPP headers. ## -## --------------------- ## +## ------------------------ ## +## LAC: Memory exhaustion. ## +## ------------------------ ## -# AT_TEST_CPP_GUARD_H([INPUT-FILE-BASE) -# ------------------------------------- -m4_define([AT_TEST_CPP_GUARD_H], -[AT_SETUP([Invalid CPP guards: $1]) +AT_SETUP([[LAC: Memory exhaustion]]) -# Possibly create inner directories. -dirname=`AS_DIRNAME([$1])` -AS_MKDIR_P([$dirname]) +m4_pushdef([AT_LAC_CHECK], [ -AT_DATA([$1.y], -[%% -dummy: -]) +AT_DATA_GRAMMAR([input.y], +[[%code { + #include + void yyerror (char const *); + int yylex (void); + #define YYMAXDEPTH 8 +} -AT_CHECK([bison --defines=$1.h $1.y]) +%error-verbose -# CPP should be happy with it. -AT_CHECK([$CC -E $1.h], 0, [ignore]) +%% + +S: A A A A A A A A A ; +A: /*empty*/ | 'a' ; + +%% + +void +yyerror (char const *msg) +{ + fprintf (stderr, "%s\n", msg); +} + +int +yylex (void) +{ + static char const *input = "]$1["; + return *input++; +} + +int +main (void) +{ + yydebug = 1; + return yyparse (); +} +]]) + +AT_BISON_CHECK([[-Dparse.lac=full -Dparse.lac.es-capacity-initial=1 \ + -t -o input.c input.y]], [[0]], [], +[[input.y: conflicts: 8 shift/reduce +]]) +AT_COMPILE([[input]]) -AT_CLEANUP ]) -AT_TEST_CPP_GUARD_H([input/input]) -AT_TEST_CPP_GUARD_H([9foo]) +# Check for memory exhaustion during parsing. +AT_LAC_CHECK([[]]) +AT_PARSER_CHECK([[./input]], [[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]], [[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