changelog

GNU的词法/语法分析器bison源码

	* tests/calc.at (AT_CHECK_CALC): Adjust to lack of quotes
	in diagnostics.
	* tests/input.at (Torturing the Scanner): Likewise.
	* tests/regression.at (Token definitions, Web2c Actions): Likewise.

2005-04-16  Paul Eggert  <[e-mail]>

	* tests/torture.at (AT_INCREASE_DATA_SIZE): Skip the test if
	the data size is known to be too small and we can't increase it.
	This works around an HP-UX 11.00 glitch reported by Andrew Benham.

2005-04-15  Paul Eggert  <[e-mail]>

	* src/parse-gram.y: Include quotearg.h.
	(string_as_id): Quote $1 before using it as a key, since the
	lexer no longer quotes it for us.
	(string_content): Don't strip quotes, since lexer no longer
	quotes it for us.
	* src/scan-gram.l: Include quotearg.h.
	("\""): Omit quote.
	("'"<SC_ESCAPED_CHARACTER>): Quote symbol before using it as
	a key, since the rest of the lexer doesn't quote it.
	* src/symtab.c (symbol_get): Don't quote symbol; caller does it now.
	* tests/regression.at (Token definitions): Check for backslashes
	in token strings.

	* data/yacc.c (YYSTACK_ALLOC_MAXIMUM): New macro.
	(YYSIZE_T): Define to unsigned long int when using an older compiler.
	(yyparse): Revamp code to generate long syntax error message, to
	make it easier to translate, and to avoid problems with arithmetic
	overflow.  Change "virtual memory" to "memory" in diagnostic, since
	we don't know whether the memory is virtual.

2005-04-13  Paul Eggert  <[e-mail]>

	* NEWS: Bison-generated C parsers now use the _ macro to
	translate strings.
	* data/yacc.c (_) [!defined _]: New macro.
	All English strings wrapped inside this macro.
	* doc/bison.texinfo (Bison Parser): Document _.
	* po/POTFILES.in: Include src/parse-gram.c, since it now
	includes translateable strings that parse-gram.y doesn't.

2005-04-12  Paul Eggert  <[e-mail]>

	* src/symtab.c (symbol_make_alias): Call symbol_type_set,
	reverting the 2004-10-11 change to this function.
	(symbol_check_alias_consistency): Don't call symbol_type_set
	if the type name is already correct.
	* tests/input.at (Typed symbol aliases): New test, from Tim Van Holder.

2005-03-25  Paul Eggert  <[e-mail]>

	* tests/regression.at (Token definitions): Don't use a token named
	c, as that generates a "#define c ..." that runs afoul of buggy
	stdlib.h that uses the identifier c as a member of struct
	drand48_data.  Problem reported by Horst Wente.

2005-03-21  Paul Eggert  <[e-mail]>

	* bootstrap: Change translation URL from
	http://www2.iro.umontreal.ca/~gnutra/po/maint/bison/ to
	http://www.iro.umontreal.ca/translation/maint/bison/ to avoid
	redirection glitches.  Problem reported by [e-mail].

2005-03-20  Paul Eggert  <[e-mail]>

	* tests/local.at (AT_COMPILE, AT_COMPILE_CXX): Don't put options
	after operands; POSIX says this isn't portable for the c99 command.

2005-03-18  Paul Eggert  <[e-mail]>

	* tests/glr-regression.at (glr-regr2a.y): Try to dump core
	immediately if a data overrun has occurred; this may help us track
	down what may be a spurious failure on MacOS.

2005-03-17  Paul Eggert  <[e-mail]>

	Respond to problems reported by [e-mail].

	* bootstrap: Use "trap - 0" rather than the unportable "trap 0".

	* src/vcg.h: Comment fix.
	* src/vcg_defaults.h: Parenthesize macro bodies to make them safe.
	(G_CMAX): Change to -1 instead of INT_MAX.

	* data/yacc.c (yyparse): Omit spaces before #line.

2005-03-15  Paul Eggert  <[e-mail]>

	* src/tables.c (state_number_to_vector_number): Put it inside an
	"#if 0", since it's not currently used.  Problem reported by
	Roland McGrath.

2005-03-06  Paul Eggert  <[e-mail]>

	* src/output.c (escaped_output): Renamed from
	escaped_file_name_output, since we now use it for symbol tags as
	well.  All uses changed.
	(symbol_destructors_output, symbol_printers_output):
	Escape symbol tags too.
	Problem reported by Matyas Forstner in
	<http://lists.gnu.org/archive/html/bug-bison/2005-03/msg00009.html>.

	* src/muscle_tab.c (muscle_code_grow): Don't quote numbers; it's
	not needed.
	* src/output.c (user_actions_output, token_definitions_output,
	symbol_destructors_output, symbol_printers_output): Likewise.
	* src/reader.c (prologue_augment): Likewise.
	* src/scan-gram.l (handle_action_dollar, handle_action_at): Likewise.

	* src/vcg.c (output_edge): Don't quote linestyle arg.
	Problem reported by [e-mail].

2005-02-28  Paul Eggert  <[e-mail]>

	* doc/bison.texinfo (Semantic Tokens): Fix scoping problem in
	example, reported by Derek M Jones.  Also, make the example even
	more outrageous, to better illustrate how bad the problem is.

2005-02-24  Paul Eggert  <[e-mail]>

	* doc/bison.texinfo (Mfcalc Symtab): Correct the prototype for
	putsym.  Typo reported by Sebastian Piping.

2005-02-23  Paul Eggert  <[e-mail]>

	* doc/bison.texinfo (Language and Grammar): some -> same
	(Epilogue): int he -> in the
	Typos reported by Sebastian Piping via Justin Pence.

2005-02-07  Paul Eggert  <[e-mail]>

	* tests/glr-regression.at (Improper handling of embedded actions
	and dollar(-N) in GLR parsers): Renamed from "Improper handling of
	embedded actions and $-N in GLR parsers", work around an Autoconf bug
	with dollar signs in test names.
	* tests/input.at (Invalid dollar-n): Renamed from "Invalid \$n",
	for a similar reason.

2005-01-28  Paul Eggert  <[e-mail]>

	* src/vcg.c (output_graph): G_VIEW -> normal_view in case someone
	wants to redefine G_VIEW.

2005-01-27  Paul Eggert  <[e-mail]>

	* src/vcg.c (get_view_str): Remove case for normal_view.
	Problem reported by [e-mail].

2005-01-24  Paul Eggert  <[e-mail]>

	* configure.ac (O0CFLAGS, O0CXXFLAGS): Fix quoting bug.
	Problem reported by [e-mail].

	* doc/bison.texinfo: Change @dircategory from "GNU programming
	tools" to "Software development".  Requested by Richard Stallman
	via Karl Berry.

2005-01-23  Paul Eggert  <[e-mail]>

	* tests/c++.at (AT_CHECK_DOXYGEN): Don't use options after operands.
	Problem reported by [e-mail].

2005-01-21  Paul Eggert  <[e-mail]>

	* data/yacc.c (YYCOPY, yystpcpy, yyparse): Remove "register"
	keyword; it's not needed with modern compilers, and it doesn't
	affect correctness with older compilers.  Suggested by
	[e-mail].

2005-01-17  Paul Eggert  <[e-mail]>

	* data/glr.c (yyuserAction): Add "default: break;" case to pacify
	gcc -Wswitch-default.
	* data/lalr1.cc (yy::]b4_parser_class_name[::parse): Likewise.
	* data/yacc.c (yyparse): Likewise.

2005-01-12  Paul Eggert  <[e-mail]>

	* src/system.h (OUTPUT_EXT, TAB_EXT): Define only if not defined
	already.  Let config.h define any nonstandard values.

2005-01-10  Paul Eggert  <[e-mail]>

	* tests/calc.at (_AT_DATA_CALC_Y): Use alarm (100), not alarm (10),
	for the benefit of slower hosts.  Problem reported by
	Nelson H. F. Beebe.

2005-01-07  Paul Eggert  <[e-mail]>

	* data/yacc.c (yyparse): Pacify non-GCC compilers about yyerrorlab
	being defined and not used.
	* data/lalr1.cc (yyparse): Likewise.
	Use "if (false)" rather than "if (0)".

2005-01-05  Paul Eggert  <[e-mail]>

	* TODO: Mention that we should allow NUL bytes in tokens.

2005-01-02  Paul Eggert  <[e-mail]>

	* src/scan-skel.l (<<EOF>>): Don't close standard output.
	Problem reported by Hans Aberg.

2005-01-01  Paul Eggert  <[e-mail]>

	* src/getargs.c (version): Happy new year; update overall
	program copyright date from 2004 to 2005.

	* src/scan-skel.l ("@output ".*\n): Don't close standard output.
	Problem reported by Hans Aberg.
	* tests/output.at (AT_CHECK_OUTPUT): New arg SHELLIO.
	(Output file names.): Add a test for the case when standard output
	is closed.

2004-12-26  Paul Eggert  <[e-mail]>

	* doc/bison.texinfo (@copying): Update FDL version number to 1.2,
	to fix an oversight in the Bison 2.0 manual.

2004-12-25  Paul Eggert  <[e-mail]>

	* NEWS: Version 2.0.  Reformat the existing news items since
	1.875, so that related items are grouped together.
	* configure.ac (AC_INIT): Bump version to 2.0.
	* src/parse-gram.c, src/parse-gram.h: Regenerate with 2.0.

	* tests/torture.at (Exploding the Stack Size with Alloca): Set
	YYSTACK_USE_ALLOCA to 1 if __GNUC__ or alloca are defined;
	otherwise, we're not testing alloca.  Unfortunately there's no
	simple way to consult HAVE_ALLOCA here.

	* data/lalr1.cc (yydestruct_): Pacify unused variable warning
	for yymsg, too.

	* src/LR0.c (new_itemsets): Use memset rather than zeroing by
	hand.  This avoids a warning about comparing int to size_t when
	GCC warnings are enabled.

2004-12-22  Paul Eggert  <[e-mail]>

	* NEWS: Bison-generated parsers no longer default to using the
	alloca function (when available) to extend the parser stack, due
	to widespread problems in unchecked stack-overflow detection.
	* data/glr.c (YYMAXDEPTH): Remove undef when zero.  It's the user's
	responsibility to set it to a positive value.  This lets the user
	specify a value that is not a preprocessor constant.
	* data/yacc.c (YYMAXDEPTH): Likewise.
	(YYSTACK_ALLOC): Define only if YYSTACK_USE_ALLOCA is nonzero.
	* doc/bison.texinfo (Stack Overflow): YYMAXDEPTH no longer needs
	to be a compile-time constant.  However, explain the constraints on it.
	Also, explain the constraints on YYINITDEPTH.
	(Table of Symbols): Explain that alloca is no longer the default.
	Explain the user's responsibility if they define YYSTACK_USE_ALLOCA
	to 1.

	* doc/bison.texinfo (Location Default Action): Mention that n must
	be zero when k is zero.  Suggested by Frank Heckenbach.

2004-12-22  Akim Demaille  <[e-mail]>

	* data/lalr1.cc (parser::token_number_type, parser::rhs_number_type)
	(parser::state_type, parser::semantic_type, parser::location_type):
	Private, not public.
	(parser::parse): Return ints, not bool.
	Returning a bool introduces a problem: 0 corresponds to false, and
	it seems weird to return false on success.  Returning true changes
	the conventions for yyparse.
	Alternatively we could return void and send an exception.
	There is no clear consensus (yet?).
	(state_stack, semantic_stack, location_stack): Rename as...
	(state_stack_type, semantic_stack_type, location_stack_type): these.
	Private, not public.
	* tests/c++.at: New.
	* tests/testsuite.at, tests/Makefile.am: Adjust.

2004-12-21  Akim Demaille  <[e-mail]>

	* data/lalr1.cc (parser::parse): Return a bool instead of an int.

2004-12-21  Akim Demaille  <[e-mail]>

	Don't impose std::string for filenames.

	* data/lalr1.cc (b4_filename_type): New.
	(position::filename): Use it.
	(parser.hh): Move the inclusion of stack.hh and location.hh below
	the user code, so that needed headers for the filename type can be
	included first.
	Forward declare them before the user code.
	* tests/Makefile.am (check-local, installcheck-local): Pass
	TESTSUITEFLAGS to the TESTSUITE.

2004-12-20  Akim Demaille  <[e-mail]>

	Use more STL like names: my_class instead of MyClass.

	* data/lalr1.cc (LocationStack, LocationType, RhsNumberType)
	(SemanticStack, SemanticType, StateStack, StateType)
	(TokenNumberType, Stack, Slice, Traits, Parser::location)
	(Parser::value): Rename as...
	(location_stack, location_type, rhs_number_type, semantic_stack)
	(semantic_type, state_stack, state_type, token_number_type, stack)
	(slice, traits, parser::yylloc, parser::yylval): these.

	* tests/calc.at, tests/regression.at, tests/actions.at: Adjust.

2004-12-19  Paul Eggert  <[e-mail]>

	* data/glr.c (YYLLOC_DEFAULT): Use GNU spacing conventions.
	* data/yacc.c (YYLLOC_DEFAULT): Likewise.

2004-12-17  Paul Eggert  <[e-mail]>

	Remove uses of 'short int' and 'unsigned short int'.  This raises
	some arbitrary limits.  It uses more memory but nowadays that's
	not much of an issue.

	This change does not affect the generated parsers; that's a different
	task, as some users will want to conserve memory there.