4 Changes in version 1.75a, 2002-10-24:
6 * Bison can no longer be built by a K&R C compiler; it requires C89 or
7 later to be built. This change originally took place a few versions
8 ago, but nobody noticed until we recently asked someone to try
9 building Bison with a K&R C compiler.
12 This new directive is preferred over YYERROR_VERBOSE.
14 * %lex-param, %parse-param
15 These new directives are preferred over PARSE_PARAM and LEX_PARAM.
16 In addition, they provide a means for yyerror to remain pure, and
17 to access to the current location.
19 Changes in version 1.75, 2002-10-14:
21 * Bison should now work on 64-bit hosts.
23 * Indonesian translation thanks to Tedi Heriyanto.
26 Fix spurious parse errors.
29 Some people redefine yyerror to steal yyparse' private variables.
30 Reenable this trick until an official feature replaces it.
33 In agreement with POSIX and with other Yaccs, leaving a default
34 action is valid when $$ is untyped, and $1 typed:
38 but the converse remains an error:
42 * Values of mid-rule actions
45 foo: { ... } { $$ = $1; } ...
47 was incorrectly rejected: $1 is defined in the second mid-rule
48 action, and is equal to the $$ of the first mid-rule action.
50 Changes in version 1.50, 2002-10-04:
55 causes Bison to produce a Generalized LR (GLR) parser, capable of handling
56 almost any context-free grammar, ambiguous or not. The new declarations
57 %dprec and %merge on grammar rules allow parse-time resolution of
58 ambiguities. Contributed by Paul Hilfinger.
60 Unfortunately Bison 1.50 does not work properly on 64-bit hosts
61 like the Alpha, so please stick to 32-bit hosts for now.
64 When not in Yacc compatibility mode, when the output file was not
65 specified, running `bison foo/bar.y' created `foo/bar.c'. It
69 The undefined token was systematically mapped to 2 which prevented
70 the use of 2 by the user. This is no longer the case.
72 * Unknown token numbers
73 If yylex returned an out of range value, yyparse could die. This is
77 According to POSIX, the error token must be 256.
78 Bison extends this requirement by making it a preference: *if* the
79 user specified that one of her tokens is numbered 256, then error
80 will be mapped onto another number.
82 * Verbose error messages
83 They no longer report `..., expecting error or...' for states where
84 error recovery is possible.
87 Defaults to `$end' instead of `$'.
89 * Error recovery now conforms to documentation and to POSIX
90 When a Bison-generated parser encounters a syntax error, it now pops
91 the stack until it finds a state that allows shifting the error
92 token. Formerly, it popped the stack until it found a state that
93 allowed some non-error action other than a default reduction on the
94 error token. The new behavior has long been the documented behavior,
95 and has long been required by POSIX. For more details, please see
96 <http://mail.gnu.org/pipermail/bug-bison/2002-May/001452.html>.
99 Popped tokens and nonterminals are now reported.
102 Larger grammars are now supported (larger token numbers, larger grammar
103 size (= sum of the LHS and RHS lengths), larger LALR tables).
104 Formerly, many of these numbers ran afoul of 16-bit limits;
105 now these limits are 32 bits on most hosts.
107 * Explicit initial rule
108 Bison used to play hacks with the initial rule, which the user does
109 not write. It is now explicit, and visible in the reports and
113 Before, Bison reported the useless rules, but, although not used,
114 included them in the parsers. They are now actually removed.
116 * Useless rules, useless nonterminals
117 They are now reported, as a warning, with their locations.
119 * Rules never reduced
120 Rules that can never be reduced because of conflicts are now
123 * Incorrect `Token not used'
126 %token useless useful
128 exp: '0' %prec useful;
130 where a token was used to set the precedence of the last rule,
131 bison reported both `useful' and `useless' as useless tokens.
133 * Revert the C++ namespace changes introduced in 1.31
134 as they caused too many portability hassles.
137 By an accident of design, the default computation of @$ was
138 performed after another default computation was performed: @$ = @1.
139 The latter is now removed: YYLLOC_DEFAULT is fully responsible of
140 the computation of @$.
143 The token end of file may be specified by the user, in which case,
144 the user symbol is used in the reports, the graphs, and the verbose
145 error messages instead of `$end', which remains being the default.
149 %token YYEOF 0 "end of file"
152 This old option, which has been broken for ages, is removed.
155 Brazilian Portuguese, thanks to Alexandre Folle de Menezes.
156 Croatian, thanks to Denis Lackovic.
158 * Incorrect token definitions
159 When given `%token 'a' "A"', Bison used to output `#define 'a' 65'.
161 * Token definitions as enums
162 Tokens are output both as the traditional #define's, and, provided
163 the compiler supports ANSI C or is a C++ compiler, as enums.
164 This lets debuggers display names instead of integers.
167 In addition to --verbose, bison supports --report=THINGS, which
168 produces additional information:
170 complete the core item sets with their closure
172 explicitly associate lookaheads to items
174 describe shift/reduce conflicts solving.
175 Bison used to systematically output this information on top of
176 the report. Solved conflicts are now attached to their states.
179 Previous versions don't complain when there is a type clash on
180 the default action if the rule has a mid-rule action, such as in:
188 * GNU M4 is now required when using Bison.
190 Changes in version 1.35, 2002-03-25:
193 Some projects use Bison's C parser with C++ compilers, and define
194 YYSTYPE as a class. The recent adjustment of C parsers for data
195 alignment and 64 bit architectures made this impossible.
197 Because for the time being no real solution for C++ parser
198 generation exists, kludges were implemented in the parser to
199 maintain this use. In the future, when Bison has C++ parsers, this
200 kludge will be disabled.
202 This kludge also addresses some C++ problems when the stack was
205 Changes in version 1.34, 2002-03-12:
207 * File name clashes are detected
208 $ bison foo.y -d -o foo.x
209 fatal error: header and parser would both be named `foo.x'
211 * A missing `;' at the end of a rule triggers a warning
212 In accordance with POSIX, and in agreement with other
213 Yacc implementations, Bison will mandate this semicolon in the near
214 future. This eases the implementation of a Bison parser of Bison
215 grammars by making this grammar LALR(1) instead of LR(2). To
216 facilitate the transition, this release introduces a warning.
218 * Revert the C++ namespace changes introduced in 1.31, as they caused too
219 many portability hassles.
221 * DJGPP support added.
223 * Fix test suite portability problems.
225 Changes in version 1.33, 2002-02-07:
228 Groff could not be compiled for the definition of size_t was lacking
229 under some conditions.
234 Changes in version 1.32, 2002-01-23:
236 * Fix Yacc output file names
240 * Italian, Dutch translations
242 Changes in version 1.31, 2002-01-14:
246 * GNU Gettext and %expect
247 GNU Gettext asserts 10 s/r conflicts, but there are 7. Now that
248 Bison dies on incorrect %expectations, we fear there will be
249 too many bug reports for Gettext, so _for the time being_, %expect
250 does not trigger an error when the input file is named `plural.y'.
252 * Use of alloca in parsers
253 If YYSTACK_USE_ALLOCA is defined to 0, then the parsers will use
254 malloc exclusively. Since 1.29, but was not NEWS'ed.
256 alloca is used only when compiled with GCC, to avoid portability
259 * When the generated parser lacks debugging code, YYDEBUG is now 0
260 (as POSIX requires) instead of being undefined.
263 Bison has always permitted actions such as { $$ = $1 }: it adds the
264 ending semicolon. Now if in Yacc compatibility mode, the semicolon
265 is no longer output: one has to write { $$ = $1; }.
267 * Better C++ compliance
268 The output parsers try to respect C++ namespaces.
269 [This turned out to be a failed experiment, and it was reverted later.]
272 Fixed bugs when reporting useless nonterminals.
275 The parsers work properly on 64 bit hosts.
278 Some calls to strerror resulted in scrambled or missing error messages.
281 When the number of shift/reduce conflicts is correct, don't issue
284 * The verbose report includes the rule line numbers.
286 * Rule line numbers are fixed in traces.
288 * Swedish translation
291 Verbose parse error messages from the parsers are better looking.
292 Before: parse error: unexpected `'/'', expecting `"number"' or `'-'' or `'(''
293 Now: parse error: unexpected '/', expecting "number" or '-' or '('
295 * Fixed parser memory leaks.
296 When the generated parser was using malloc to extend its stacks, the
297 previous allocations were not freed.
299 * Fixed verbose output file.
300 Some newlines were missing.
301 Some conflicts in state descriptions were missing.
303 * Fixed conflict report.
304 Option -v was needed to get the result.
308 Mismatches are errors, not warnings.
310 * Fixed incorrect processing of some invalid input.
312 * Fixed CPP guards: 9foo.h uses BISON_9FOO_H instead of 9FOO_H.
314 * Fixed some typos in the documentation.
316 * %token MY_EOF 0 is supported.
317 Before, MY_EOF was silently renumbered as 257.
319 * doc/refcard.tex is updated.
321 * %output, %file-prefix, %name-prefix.
325 New, aliasing `--output-file'.
327 Changes in version 1.30, 2001-10-26:
329 * `--defines' and `--graph' have now an optionnal argument which is the
330 output file name. `-d' and `-g' do not change, they do not take any
333 * `%source_extension' and `%header_extension' are removed, failed
338 Changes in version 1.29, 2001-09-07:
340 * The output file does not define const, as this caused problems when used
341 with common autoconfiguration schemes. If you still use ancient compilers
342 that lack const, compile with the equivalent of the C compiler option
343 `-Dconst='. autoconf's AC_C_CONST macro provides one way to do this.
345 * Added `-g' and `--graph'.
347 * The Bison manual is now distributed under the terms of the GNU FDL.
349 * The input and the output files has automatically a similar extension.
351 * Russian translation added.
353 * NLS support updated; should hopefully be less troublesome.
355 * Added the old Bison reference card.
357 * Added `--locations' and `%locations'.
359 * Added `-S' and `--skeleton'.
361 * `%raw', `-r', `--raw' is disabled.
363 * Special characters are escaped when output. This solves the problems
364 of the #line lines with path names including backslashes.
367 `%yacc', `%fixed_output_files', `%defines', `%no_parser', `%verbose',
368 `%debug', `%source_extension' and `%header_extension'.
371 Automatic location tracking.
373 Changes in version 1.28, 1999-07-06:
375 * Should compile better now with K&R compilers.
379 * Fixed a problem with escaping the double quote character.
381 * There is now a FAQ.
383 Changes in version 1.27:
385 * The make rule which prevented bison.simple from being created on
386 some systems has been fixed.
388 Changes in version 1.26:
390 * Bison now uses automake.
392 * New mailing lists: <bug-bison@gnu.org> and <help-bison@gnu.org>.
394 * Token numbers now start at 257 as previously documented, not 258.
396 * Bison honors the TMPDIR environment variable.
398 * A couple of buffer overruns have been fixed.
400 * Problems when closing files should now be reported.
402 * Generated parsers should now work even on operating systems which do
403 not provide alloca().
405 Changes in version 1.25, 1995-10-16:
407 * Errors in the input grammar are not fatal; Bison keeps reading
408 the grammar file, and reports all the errors found in it.
410 * Tokens can now be specified as multiple-character strings: for
411 example, you could use "<=" for a token which looks like <=, instead
412 of chosing a name like LESSEQ.
414 * The %token_table declaration says to write a table of tokens (names
415 and numbers) into the parser file. The yylex function can use this
416 table to recognize multiple-character string tokens, or for other
419 * The %no_lines declaration says not to generate any #line preprocessor
420 directives in the parser file.
422 * The %raw declaration says to use internal Bison token numbers, not
423 Yacc-compatible token numbers, when token names are defined as macros.
425 * The --no-parser option produces the parser tables without including
426 the parser engine; a project can now use its own parser engine.
427 The actions go into a separate file called NAME.act, in the form of
428 a switch statement body.
430 Changes in version 1.23:
432 The user can define YYPARSE_PARAM as the name of an argument to be
433 passed into yyparse. The argument should have type void *. It should
434 actually point to an object. Grammar actions can access the variable
435 by casting it to the proper pointer type.
437 Line numbers in output file corrected.
439 Changes in version 1.22:
443 Changes in version 1.20:
445 Output file does not redefine const for C++.
453 Copyright (C) 2001, 2002 Free Software Foundation, Inc.
455 This file is part of GNU Autoconf.
457 GNU Autoconf is free software; you can redistribute it and/or modify
458 it under the terms of the GNU General Public License as published by
459 the Free Software Foundation; either version 2, or (at your option)
462 GNU Autoconf is distributed in the hope that it will be useful,
463 but WITHOUT ANY WARRANTY; without even the implied warranty of
464 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
465 GNU General Public License for more details.
467 You should have received a copy of the GNU General Public License
468 along with autoconf; see the file COPYING. If not, write to
469 the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
470 Boston, MA 02111-1307, USA.