Theophile Ranquet [Mon, 8 Oct 2012 15:53:44 +0000 (15:53 +0000)]
graphs: style: use left justification for states
The label text of nodes is centered "by default" (by the use of '\n' as
a line feed). This gives bad readability to the grammar rules shown in
state nodes, a left justification is much nicer. This is done by using '\l'
as the line feed.
In order to allow \l in the DOT file, changes to the quoting system seem
necessary.
* src/print_graph.c (print_core): Escape tokens here, instead of...
* src/graphviz.c (output_node): Here...
(escape): Using this, new.
Signed-off-by: Akim Demaille <akim@lrde.epita.fr>
Theophile Ranquet [Wed, 10 Oct 2012 17:14:02 +0000 (17:14 +0000)]
graphs: style: prefix rules and change shapes
* src/graphviz.c (start_graph): Use box rather than ellipsis.
* src/print_graph.c (print_core): Prefix rules with their number.
Signed-off-by: Akim Demaille <akim@lrde.epita.fr>
Theophile Ranquet [Wed, 10 Oct 2012 17:14:01 +0000 (17:14 +0000)]
obstack: import obstack_finish0 from master
* src/system.h (obstack_finish0): New.
Signed-off-by: Akim Demaille <akim@lrde.epita.fr>
Akim Demaille [Tue, 9 Oct 2012 10:09:59 +0000 (12:09 +0200)]
c++: api.location.type
This feature was introduced in
95a2de5695670ae0df98cb3c42141cad549f0204
(which is part of 2.5), but not documented. Give it a proper name, and
make it public.
* data/c++.m4, data/lalr1.cc, data/glr.cc, data/java.m4: Use
api.location.type instead of location_type.
* src/muscle-tab.c (muscle_percent_variable_update): Map the latter to
the former.
* tests/local.at: Adjust.
* tests/calc.at: Use api.location.type.
Leave tests/java.at with location_type, at least for the time being,
to cover both names.
* doc/bison.texi: Document api.location.type.
(User Defined Location Type): New.
* NEWS: Update.
Akim Demaille [Tue, 9 Oct 2012 10:13:55 +0000 (12:13 +0200)]
muscles: a function for backward compatibility
Based on commit
171ad99d6421935a278656be6dc7161591835d00 from master.
* src/muscle-tab.c (muscle_percent_variable_update): New.
(muscle_percent_define_insert): Use it.
Define the variables with their initial value.
Akim Demaille [Tue, 9 Oct 2012 10:12:52 +0000 (12:12 +0200)]
maint: more macros
* src/output.c (ARRAY_CARDINALITY): Move to...
* src/system.h: here.
(STREQ, STRNEQ): new.
Akim Demaille [Mon, 8 Oct 2012 11:46:50 +0000 (13:46 +0200)]
NEWS: warnings with clang
* NEWS: here.
Akim Demaille [Mon, 8 Oct 2012 07:17:20 +0000 (09:17 +0200)]
warnings: avoid warnings from clang
Fix the following warning
parse-gram.c:2078:14: error: equality comparison with extraneous parentheses
[-Werror,-Wparentheses-equality]
if (((yyn) == (-91)))
~~~~~~^~~~~~~~
parse-gram.c:2078:14: note: remove extraneous parentheses around the
comparison to silence this warning
if (((yyn) == (-91)))
~ ^ ~
parse-gram.c:2078:14: note: use '=' to turn this equality comparison into
an assignment
if (((yyn) == (-91)))
^~
=
1 error generated.
and the following one:
input.cc:740:1: error: function declared 'noreturn' should not return
[-Werror,-Winvalid-noreturn]
static void yyMemoryExhausted (yyGLRStack* yystackp)
__attribute__ ((__noreturn__));
static void
yyMemoryExhausted (yyGLRStack* yystackp)
{
YYLONGJMP (yystackp->yyexception_buffer, 2);
}
^
1 warning and 1 error generated.
This is Apple clang version 3.1 (tags/Apple/clang-318.0.61).
* data/c.m4 (b4_table_value_equals): Use (!!(A == B)) instead of (A == B)
to avoid this warning.
Any reasonable compiler should generate the same code.
* src/uniqstr.h (UNIQSTR_EQ): Likewise.
* data/glr.c (LONGJMP): abort after longjmp to pacify clang.
Akim Demaille [Mon, 8 Oct 2012 07:02:09 +0000 (09:02 +0200)]
tests: no longer disable -O compiler options
Tests are running without -O since
f377f69fec28013c79db4efe12bbb9d48987fb2c because some warnings (about
yylval not being initialized) show only when GCC is given -O2. The
previous patch fixes the warnings. Run the test suite with compiler
options unmodified.
* tests/atlocal.in (O0CFLAGS, O0CXXFLAGS): Remove, use CFLAGS and
CXXFLAGS.
Paul Eggert [Fri, 5 Oct 2012 17:55:11 +0000 (10:55 -0700)]
yacc.c: initialize yylval in pure-parser mode
See http://lists.gnu.org/archive/html/bison-patches/2012-08/msg00024.html
(spreading over September and October).
* data/yacc.c (YY_IGNORE_MAYBE_UNINITIALIZED_BEGIN)
(YY_IGNORE_MAYBE_UNINITIALIZED_END, YYLVAL_INITIALIZE):
New macros. Use them to suppress an unwanted GCC diagnostic.
Akim Demaille [Mon, 8 Oct 2012 07:12:10 +0000 (09:12 +0200)]
skeletons: style changes
* data/yacc.c, data/glr.c: Prefer Title case for (CPP) macro arguments.
Akim Demaille [Mon, 1 Oct 2012 09:41:26 +0000 (11:41 +0200)]
lalr1.cc: document exception safety
* NEWS: here.
* doc/bison.texi (Destructor Decl, C++ Parser Interface): and there.
Akim Demaille [Tue, 25 Sep 2012 12:18:04 +0000 (14:18 +0200)]
lalr1.cc: check exception safety of error handling
* tests/c++.at (Exception safety): Don't use swap here, it
is useless.
Cover more test cases: yyerror, YYERROR, YYABORT, and
error recovery.
(Object): Instead of just keeping a counter of instances, keep
a list of them.
Akim Demaille [Tue, 25 Sep 2012 09:41:22 +0000 (11:41 +0200)]
lalr1.cc: check (and fix) %printer exception safety
* tests/c++.at (Exception safety): Let the parser support the --debug
option.
On 'p', throw an exception from the %printer.
* data/lalr1.cc (yyparse): Do not display the values we discard, as it
uses %printer, which might have thrown the exception.
Akim Demaille [Tue, 25 Sep 2012 09:17:55 +0000 (11:17 +0200)]
lalr1.cc: check (and fix) %initial-action exception safety
* data/lalr1.cc: Check size > 1, rather than size != 1, when cleaning
the stack, as at the beginning, size is 0.
* tests/c++.at (Exception safety): Check exception safety in
%initial-action.
Akim Demaille [Thu, 20 Sep 2012 14:59:29 +0000 (16:59 +0200)]
lalr1.cc: fix exception safety
lalr1.cc does not reclaim its memory when ended by an exception.
Reported by Oleksii Taran:
http://lists.gnu.org/archive/html/help-bison/2012-09/msg00000.html
* data/lalr1.cc (yyparse): Protect the whole yyparse by a try-catch
block that cleans the stack and the lookahead.
Akim Demaille [Thu, 20 Sep 2012 09:43:19 +0000 (11:43 +0200)]
lalr1.cc: check exception safety.
* tests/c++.at (Exception safety): New.
Akim Demaille [Thu, 20 Sep 2012 15:04:50 +0000 (17:04 +0200)]
lalr1.cc: indentation fixes.
* data/lalr1.cc (yyparse): here.
Untabify a block of code.
Akim Demaille [Tue, 25 Sep 2012 12:14:57 +0000 (14:14 +0200)]
lalr1.cc: don't leave macros define to nothing
* data/lalr1.cc (YY_SYMBOL_PRINT, YY_REDUCE_PRINT, YY_STACK_PRINT):
Define to something so that, for instance, "if (foo) YY_SYMBOL_PRINT"
is valid even when !YYDEBUG.
Akim Demaille [Thu, 20 Sep 2012 09:42:06 +0000 (11:42 +0200)]
tests: minor improvements
* tests/c++.at: Space changes.
Use AT_YYERROR_DEFINE.
* tests/local.at (AT_YYERROR_DEFINE): Issue errors on unknown languages.
Akim Demaille [Fri, 5 Oct 2012 07:24:59 +0000 (09:24 +0200)]
tests: use $PERL instead of perl
* tests/atlocal.in (PERL): New.
Sort.
* tests/calc.at, tests/input.at, tests/local.at, tests/regression.at,
* tests/skeletons.at, tests/synclines.at, tests/torture.at: here.
Akim Demaille [Sun, 8 Apr 2012 06:58:43 +0000 (08:58 +0200)]
build: look for Perl in configure.
Bison uses "/usr/bin/perl" or "perl" in several places, and it does
not appear to be a problem. But, at least to make it simpler to
change PERL on the make command line, check for perl in configure.
* configure.ac (PERL): New.
* doc/Doxyfile.in, doc/Makefile.am, tests/bison.in: Use it.
Akim Demaille [Fri, 5 Oct 2012 07:20:30 +0000 (09:20 +0200)]
tests: fix sed portability issues
Reported by Didier Godefroy,
<http://lists.gnu.org/archive/html/bug-bison/2012-10/msg00005.html>.
* tests/calc.at (AT_CHECK_SPACES): Use Perl.
Akim Demaille [Fri, 5 Oct 2012 06:54:15 +0000 (08:54 +0200)]
tests: diff -u is not portable
Reported by Didier Godefroy
<http://lists.gnu.org/archive/html/bug-bison/2012-10/msg00006.html>.
* tests/existing.at (AT_LALR1_DIFF_CHECK): Skip if diff -u does not
work.
Akim Demaille [Thu, 4 Oct 2012 09:45:09 +0000 (11:45 +0200)]
maint: word changes
* README-hacking (Typical errors): Improve wording.
Akim Demaille [Thu, 4 Oct 2012 07:07:42 +0000 (09:07 +0200)]
lalr1.cc: fix test suite portability
Reported by Rob Vermaas' Hydra build farm on x86_64-darwin 10.2.0 with
G++ 4.6.3.
* tests/headers.at (Several parsers): Include AT_DATA_SOURCE_PROLOGUE
in the files to compile.
* data/location.cc: Do not include twice string and iostream (once
by position.hh, and then by location.hh).
* README-hacking (Typical errors): Some hints for other maintainers.
Theophile Ranquet [Wed, 3 Oct 2012 15:26:56 +0000 (15:26 +0000)]
maint: fix an erroneous include
This fixes test 130 (Several parsers).
* data/location.cc: Include <iostream> rather than <iosfwd> since
we really need << on strings for instance.
* NEWS: Document this.
Signed-off-by: Akim Demaille <akim@lrde.epita.fr>
Akim Demaille [Wed, 3 Oct 2012 07:17:58 +0000 (09:17 +0200)]
tests: check that headers are self contained
Reported by Alexandre Duret-Lutz.
* tests/headers.at (Several parsers): here.
Akim Demaille [Wed, 3 Oct 2012 06:43:10 +0000 (08:43 +0200)]
doc: add missing documentation for --report
* doc/bison.texi (Bison Options): Document --report's "solved", "all",
and "none".
Akim Demaille [Thu, 27 Sep 2012 07:43:49 +0000 (09:43 +0200)]
headers: move CPP guards into YY_*_INCLUDED to avoid collisions
See <http://lists.gnu.org/archive/html/bug-bison/2012-09/msg00016.html>.
* data/c.m4 (b4_cpp_guard): Prepend YY_ and append _INCLUDED.
* tests/headers.at: Adjust.
* NEWS, doc/bison.texi: Document.
Akim Demaille [Thu, 27 Sep 2012 07:42:57 +0000 (09:42 +0200)]
minor changes.
* NEWS: Word changes.
* doc/bison.texi: Spell check.
Fix minor issues.
* tests/headers.at: Comment and formatting changes.
Akim Demaille [Fri, 28 Sep 2012 13:04:57 +0000 (15:04 +0200)]
gnulib: update
Theophile Ranquet [Fri, 28 Sep 2012 12:12:59 +0000 (12:12 +0000)]
errors: indent "user token number redeclaration" context
This is the continuation of the work on the readability of errors
context.
* src/symtab.c (user_token_number_redeclaration): Use
complain_at_indent to output with increased indentation level.
* tests/input:at: Apply this change.
Signed-off-by: Akim Demaille <akim@lrde.epita.fr>
Theophile Ranquet [Wed, 26 Sep 2012 09:49:23 +0000 (11:49 +0200)]
warnings: introduce -Wdeprecated in the usage info
The deprecated warning, introduced some time ago, was not displayed in
the usage message. This patch addresses the issue.
* src/getargs.c (usage): Insert here.
Signed-off-by: Akim Demaille <akim@lrde.epita.fr>
Theophile Ranquet [Wed, 26 Sep 2012 09:49:22 +0000 (11:49 +0200)]
errors: prefix the output with "error: "
This improves readability. This is also what gcc does.
* NEWS: Document this change.
* src/complain.c (complain_at): Prefix all errors with "error: ".
(complain_at_indent, warn_at_indent): Do not prefix the context
information of errors, which are basically just indented errors.
* tests/conflicts.at, tests/glr-regression.at, tests/input.at,
tests/named-refs.at, tests/output.at, tests/push.at,
tests/regression.at, tests/skeletons.at: Apply this change.
Signed-off-by: Akim Demaille <akim@lrde.epita.fr>
Theophile Ranquet [Wed, 26 Sep 2012 09:49:21 +0000 (11:49 +0200)]
errors: indent "invalid value for %define" context
This is the continuation of the work on the readability of errors
context.
For example, what used to be:
input.y:1.9-29: invalid value for %define variable 'foo' : 'bar'
input.y:1.9-29: accepted value: 'most'
is now:
input.y:1.9-29: invalid value for %define variable 'foo' : 'bar'
input.y:1.9-29: accepted value: 'most'
* src/muscle-tab.c (muscle_percent_define_check_values): Use
complain_at_indent to output with increased indentation level.
* tests/input:at: Apply this change.
Signed-off-by: Akim Demaille <akim@lrde.epita.fr>
Theophile Ranquet [Wed, 26 Sep 2012 09:49:20 +0000 (11:49 +0200)]
errors: indent "%define var" redefinition context
This is the continuation of the work on the readability of errors
context.
For example, what used to be:
input.y:2.9-11: %define variable 'var' redefined
input.y:1.9-11: previous definition
is now:
input.y:2.9-11: %define variable 'var' redefined
input.y:1.9-11: previous definition
* src/muscle-tab.c (muscle_percent_define_insert): Use
complain_at_indent to output with increased indentation level.
* tests/input.at: Apply this change.
Signed-off-by: Akim Demaille <akim@lrde.epita.fr>
Theophile Ranquet [Wed, 26 Sep 2012 09:49:19 +0000 (11:49 +0200)]
errors: indent "symbol redeclaration" context
This is the continuation of the work on the readability of errors
context.
For example, what used to be:
input.y:5.10-24: %printer redeclaration for <field2>
input.y:3.11-25: previous declaration
is now:
input.y:5.10-24: %printer redeclaration for <field2>
input.y:3.11-25: previous declaration
* NEWS: Document this change.
* src/symtab.c (symbol_redeclaration, semantic_type_redeclaration,
user_token_number_redeclaration, default_tagged_destructor_set,
default_tagless_destructor_set, default_tagged_printer_set,
default_tagless_printer_set): Use complain_at_indent to
output with increased indentation level.
* tests/input.at: Apply this change.
Signed-off-by: Akim Demaille <akim@lrde.epita.fr>
Theophile Ranquet [Wed, 26 Sep 2012 09:49:18 +0000 (11:49 +0200)]
errors: indent "result type clash" error context
This used to be the format of the error report:
input.y:6.5-10: result type clash on merge function 'merge': [...]
input.y:2.4-9: previous declaration
In order to distinguish the actual error from the context provided, we
rather this new output:
input.y:6.5-10: result type clash on merge function 'merge': [...]
input.y:2.4-9: previous declaration
Another patch will introduce an "error: " prefix to all non-indented
lines, giving yet better readability to the reports.
* src/complain.h (SUB_INDENT): Move to here.
* src/reader.c (record_merge_function_type): Use complain_at_indent to
output with increased indentation level.
* src/scan-code.l (SUB_INDENT): Remove from here.
* tests/glr-regression.at: Apply this change.
Signed-off-by: Akim Demaille <akim@lrde.epita.fr>
Akim Demaille [Fri, 21 Sep 2012 11:35:28 +0000 (13:35 +0200)]
regen
Akim Demaille [Fri, 21 Sep 2012 09:10:37 +0000 (11:10 +0200)]
yacc: fix handling of CPP guards when no header is generated
When no header was to be generated, Bison would issue:
/* In a future release of Bison, this section will be replaced
by #include "". */
#ifndef YY_
# define YY_
It now properly generates nothing.
* data/c.m4 (b4_cpp_guard_open, b4_cpp_guard_close): Issue nothing when
the file name is empty.
* data/yacc.c: Do not generate the above comment when there is no header
to generate.
* NEWS: Update.
Akim Demaille [Thu, 20 Sep 2012 15:04:55 +0000 (17:04 +0200)]
gnulib: update
Akim Demaille [Tue, 4 Sep 2012 08:58:23 +0000 (10:58 +0200)]
maint: remove useless file
* externals/bootstrap.cfg: Remove.
This file was used by a specific build system.
It was added to the master repository by accident.
Akim Demaille [Tue, 4 Sep 2012 08:48:26 +0000 (10:48 +0200)]
update files to ignore
* doc/.gitignore: Don't ignore split info files as we don't split our
info file.
See <http://lists.gnu.org/archive/html/bug-bison/2012-08/msg00006.html>.
Akim Demaille [Tue, 4 Sep 2012 07:49:17 +0000 (09:49 +0200)]
remove useless include
* src/system.h: Don't include sys/types.h.
Reported by Eric Blake,
<http://lists.gnu.org/archive/html/bug-bison/2012-09/msg00002.html>.
(FUNCTION_PRINT): Remove, unused.
Jim Meyering [Sun, 5 Aug 2012 11:28:01 +0000 (13:28 +0200)]
use locale-indep. c_is* functions for parsing, not isspace, isprint etc
* src/parse-gram.y: Include "c-ctype.h".
(add_param): Parse with c_isspace, not isspace.
* src/parse-gram.c: Likewise.
* src/scan-gram.l: Include c-ctype.h, not ctype.h.
(SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER): Use c_isspace and c_isprint,
not ctype.h's locale-dependent functions.
Akim Demaille [Mon, 3 Sep 2012 08:56:54 +0000 (10:56 +0200)]
gnulib: update
Akim Demaille [Fri, 31 Aug 2012 16:18:13 +0000 (18:18 +0200)]
--help: include a place to report translation issues
http://lists.gnu.org/archive/html/bug-bison/2012-08/msg00007.html
shows that it is useful to help users report translation issues.
While at it, include other informative bits that the coreutils shows.
* src/getargs.c (usage): Report more URLs where the user can
refer to.
Mostly copied/pasted from coreutils' emit_ancillary_info function.
Akim Demaille [Wed, 22 Aug 2012 12:26:26 +0000 (14:26 +0200)]
tests: style changes
* tests/torture.at (AT_DATA_STACK_TORTURE): M4 style changes to
improve readability.
Fix an assertion which, because of a <= instead of ==, did not check
new_status as visibly meant.
(get_args): New.
Akim Demaille [Wed, 22 Aug 2012 11:48:55 +0000 (13:48 +0200)]
tests: fix push-pull test
* tests/torture.at: %push-pull-parser is no longer supported.
Akim Demaille [Fri, 31 Aug 2012 15:50:31 +0000 (17:50 +0200)]
yacc.c: style changes
* data/yacc.c: (yytoken): Define with initial value.
Akim Demaille [Fri, 3 Aug 2012 07:52:05 +0000 (09:52 +0200)]
maint: post-release administrivia
* NEWS: Add header line for next release.
* .prev-version: Record previous version.
* cfg.mk (old_NEWS_hash): Auto-update.
Akim Demaille [Fri, 3 Aug 2012 07:43:02 +0000 (09:43 +0200)]
version 2.6.2
* NEWS: Record release date.
Akim Demaille [Thu, 2 Aug 2012 11:51:17 +0000 (13:51 +0200)]
NEWS: update.
* NEWS: Catch up with the other changes from 2.6.1.
Akim Demaille [Thu, 2 Aug 2012 10:01:54 +0000 (12:01 +0200)]
yacc: remove trailing end of line at end of file
There are still spurious spaces at the end of some lines. But this is
addressed in the master branch, and I am reluctant to try to backport
this.
* data/yacc.c, data/glr.c, data/lalr1.cc, data/glr.cc: here.
* tests/calc.at (AT_CHECK_SPACES): New.
Use it.
Be sure not to introduce trailing empty lines in the *.y files.
* NEWS: Doc it.
* cfg.mk (syntax-check): Remove the exception.
Akim Demaille [Thu, 2 Aug 2012 06:29:18 +0000 (08:29 +0200)]
thanks: fix a contributor name
* THANKS: On his request.
Akim Demaille [Wed, 1 Aug 2012 07:59:49 +0000 (09:59 +0200)]
gnulib: update
Akim Demaille [Wed, 1 Aug 2012 06:27:28 +0000 (08:27 +0200)]
tests: synch line -> syncline, for consistency
* tests/synclines.at: Do it, as "syncline" is used consistently
everywhere else in Bison.
Akim Demaille [Wed, 1 Aug 2012 07:30:27 +0000 (09:30 +0200)]
tests: synclines: style changes
* tests/synclines.at (AT_TEST_SYNCLINE): Rename as...
(AT_TEST): this.
Use pushdef/popdef.
Formatting changes.
Use '+' instead of '*' where appropriate.
Akim Demaille [Wed, 1 Aug 2012 07:30:23 +0000 (09:30 +0200)]
tests: synclines: fix perl invocation
Reported by Summum Bonum.
* tests/synclines.at: Fix Perl invocation: its -f is not like sed's.
Akim Demaille [Tue, 31 Jul 2012 12:18:43 +0000 (14:18 +0200)]
regen
Akim Demaille [Tue, 31 Jul 2012 12:18:39 +0000 (14:18 +0200)]
c++: trailing end-of-lines in %parse-param
* src/parse-gram.y (add_param): No only skip ' ' and '\t', skip all
leading and trailing spaces.
* tests/regression.at (Lex and parse params): Check it.
* NEWS: Document it.
Akim Demaille [Tue, 31 Jul 2012 12:29:39 +0000 (14:29 +0200)]
tests: simplify
* tests/regression.at: Remove useless compilations: AT_FULL_COMPILE
includes the compilation by bison.
Akim Demaille [Mon, 30 Jul 2012 16:05:10 +0000 (18:05 +0200)]
use obstack_printf
This is not just nicer, it is also much safer, since we were
using sprintf...
* bootstrap.conf: Require it.
* src/system.h (obstack_fgrow1, obstack_fgrow2, obstack_fgrow3)
(obstack_fgrow4): Remove.
Adjust dependencies.
Akim Demaille [Mon, 30 Jul 2012 12:54:27 +0000 (14:54 +0200)]
scanner: restore a missing start condition
$ flex src/scan-skel.l
src/scan-skel.l:145: multiple <<EOF>> rules for start condition SC_AT_DIRECTIVE_ARGS
src/scan-skel.l:145: multiple <<EOF>> rules for start condition SC_AT_DIRECTIVE_SKIP_WS
This is warning, and it seems there are no means to make it an error.
* src/scan-skel.l: Restore the start-condition INITIAL for an <<EOF>>
clause.
Akim Demaille [Mon, 30 Jul 2012 12:53:58 +0000 (14:53 +0200)]
gnulib: update
Akim Demaille [Mon, 30 Jul 2012 09:47:29 +0000 (11:47 +0200)]
maint: post-release administrivia
* NEWS: Add header line for next release.
* .prev-version: Record previous version.
* cfg.mk (old_NEWS_hash): Auto-update.
Akim Demaille [Mon, 30 Jul 2012 09:18:23 +0000 (11:18 +0200)]
version 2.6.1
* NEWS: Record release date.
Akim Demaille [Mon, 30 Jul 2012 09:10:36 +0000 (11:10 +0200)]
gnulib: update
Akim Demaille [Fri, 27 Jul 2012 15:00:28 +0000 (17:00 +0200)]
maint: fix some syntax-check issues
* cfg.mk: Nuke the following warnings which are confused by our
text reports (that state that the error token is number 256).
prohibit_magic_number_exit
../../doc/bison.texi:8170:error (256)
../../tests/conflicts.at:570:error (256)
../../tests/conflicts.at:673:error (256)
../../tests/conflicts.at:811:error (256)
../../tests/conflicts.at:1154:error (256)
../../tests/regression.at:281:error (256)
../../tests/regression.at:582:error (256)
maint.mk: use EXIT_* values rather than magic number
Akim Demaille [Fri, 27 Jul 2012 14:36:07 +0000 (16:36 +0200)]
tests: do not depend on __cplusplus to decide for C++ or C output
Since we do support compiling C code with a C++ compiler.
* tests/actions.at (Qualified $$ in actions): Use AT_SKEL_CC_IF.
Akim Demaille [Fri, 27 Jul 2012 12:52:55 +0000 (14:52 +0200)]
regen
Akim Demaille [Fri, 27 Jul 2012 12:52:23 +0000 (14:52 +0200)]
synclines: remove spurious empty line
* data/bison.m4 (b4_syncline): Do not start with an empty line.
Akim Demaille [Fri, 27 Jul 2012 12:20:00 +0000 (14:20 +0200)]
also support $<foo>$ in the %initial-action
scan-code.l is already passing argument to b4_dollar_dollar for the
initial acton, but its definition (of b4_dollar_dollar) does not use
this argument.
Generalize this definition, and use it for the %initial-action too.
* data/c.m4 (b4_dollar_dollar_, b4_dollar_pushdef, b4_dollar_popdef):
Instead of expecting a pointer, require a value, and use ".".
Since they are now generic enough, move to...
* data/c-like.m4: this new file.
* data/c.m4, data/java.m4: Load it.
* data/glr.c, data/lalr1.cc, data/lalr1.java, data/yacc.c: Use
b4_dollar_pushdef for the %initial-action.
* tests/actions.at: Check that.
* data/Makefile.am: Adjust.
* NEWS, doc/bison.texi: Document.
Akim Demaille [Fri, 27 Jul 2012 11:47:24 +0000 (13:47 +0200)]
skeletons: b4_dollar_pushdef and popdef to simpify complex definitions
M4 is really making it uselessly hard to define macros that define
macros.
* data/c.m4 (b4_dollar_pushdef, b4_dollar_popdef): New.
Use it.
Akim Demaille [Fri, 27 Jul 2012 10:53:46 +0000 (12:53 +0200)]
regen
Akim Demaille [Fri, 27 Jul 2012 10:47:02 +0000 (12:47 +0200)]
printer/destructor: translate only once
Currently "%printer {...} a b c d e f" translates the {...} six times.
Not only is this bad for time and space, it also issues six times the
same warnings.
* src/symlist.h, src/symlist.c (symbol_list_destructor_set)
(symbol_list_printer_set): Take the action as code_props instead of
const char *.
* src/parse-gram.y: Translate these actions here.
* src/scan-code.h: Comment change.
* tests/input.at: Check that warnings are issued only once.
Akim Demaille [Fri, 27 Jul 2012 09:22:22 +0000 (11:22 +0200)]
factor the handling of m4 escaping
The conversion from @ to @@ and so forth is coded is too many
different places. Factor, a bit.
* src/scan-code.l: Instead of duplicating the logic of obstack_escape,
use it.
It sure is less efficient, but the cost is negligible.
This allows to factor rules that are alike.
And to factor some start-condition clauses.
* tests/input.at (Stray $ or @): New.
* NEWS: Document it.
Akim Demaille [Fri, 27 Jul 2012 10:19:01 +0000 (12:19 +0200)]
news: schedule the removal of the ";" hack
scan-code.l is significantly more complex because of this.
* NEWS: Doc it.
Akim Demaille [Fri, 27 Jul 2012 09:11:19 +0000 (11:11 +0200)]
style changes in the scanners
* src/scan-code.l, src/scan-skel.l: Use a more traditional indentation
style for start-conditions.
Prefer "continue" to a comment, for empty actions.
Strip useless {}.
Remove useless start-condition clauses.
Akim Demaille [Thu, 26 Jul 2012 15:19:04 +0000 (17:19 +0200)]
regen
Akim Demaille [Thu, 26 Jul 2012 14:52:35 +0000 (16:52 +0200)]
support $<tag>$ in printers and destructors
* src/scan-code.l (SC_SYMBOL_ACTION): Accept $<tag>$, not just $$.
* data/c.m4 (b4_dollar_dollar_): New.
(b4_symbol_actions): Let b4_dollar_dollar use b4_dollar_dollar_.
* NEWS, doc/bison.texi: Document it.
* tests/actions.at: Check this for C and C++.
Akim Demaille [Thu, 26 Jul 2012 10:06:28 +0000 (12:06 +0200)]
scan-code: factor the handling of the type in $<TYPE>$
* src/scan-code.l (fetch_type_name): New.
(handle_action_dollar): Use it.
(gt_ptr): Remove, useless.
Akim Demaille [Thu, 26 Jul 2012 12:41:55 +0000 (14:41 +0200)]
muscles: fix another occurrence of unescaped type name
* src/output.c (quoted_output): Split into...
(quoted_output, string_output): these.
Use the former when outputting a type_name.
* tests/input.at: Check this case.
* src/symtab.h: Comment changes.
Akim Demaille [Thu, 26 Jul 2012 14:44:45 +0000 (16:44 +0200)]
glr.cc: fix the handling of yydebug
* data/glr.cc (yydebug_): Remove, unused.
(set_debug_level, debug_level): Work on yydebug instead.
* doc/bison.texi, NEWS: Document this.
Akim Demaille [Thu, 26 Jul 2012 11:28:58 +0000 (13:28 +0200)]
gnulib: update
Akim Demaille [Thu, 26 Jul 2012 12:41:20 +0000 (14:41 +0200)]
formatting changes
* src/symtab.h: here.
Akim Demaille [Thu, 26 Jul 2012 14:07:26 +0000 (16:07 +0200)]
tests: fix an assertion
* tests/local.at (AT_YYLEX_DEFINE): Be sure to check the array
against its length, not its size in bytes.
Akim Demaille [Thu, 26 Jul 2012 13:20:50 +0000 (15:20 +0200)]
tests: adjust to GCC 4.8, which displays caret errors
With GCC 4.8, the tests on synclines are skipped. Transform
input.y:1:2: error: #error "1"
#error "1"
^
into
input.y:1: #error "1"
* tests/synclines.at (AT_SYNCLINES_COMPILE): Do it, using Perl instead of
sed.
Akim Demaille [Thu, 26 Jul 2012 10:14:50 +0000 (12:14 +0200)]
be sure to properly escape type names
* src/scan-code.l: Use obstack_quote when passing type_name to m4.
* tests/input.at (Code injection): New.
* NEWS: Document it.
Thanks to Paul Eggert for the wording.
Akim Demaille [Thu, 26 Jul 2012 10:12:38 +0000 (12:12 +0200)]
obstack_quote: escape and quote for M4
* src/system.h (obstack_quote): New.
* src/muscle-tab.c: Use it instead of obstack_escape where applicable.
* src/scan-code.l: Since obstack_quote supports NULL, leave type_name
as NULL instead of defaulting to "".
Akim Demaille [Thu, 26 Jul 2012 09:49:42 +0000 (11:49 +0200)]
muscles: shuffle responsabilities
* src/muscle-tab.c (muscle_boundary_grow): Be in charge of quotation,
instead of leaving this to the caller.
Akim Demaille [Thu, 26 Jul 2012 09:40:52 +0000 (11:40 +0200)]
muscles: make private functions static
* src/muscle-tab.h, src/muscle-tab.c (muscle_boundary_grow)
(muscle_location_grow): Now static.
Akim Demaille [Thu, 26 Jul 2012 09:23:55 +0000 (11:23 +0200)]
muscles: rename private functions/macros
* src/muscle-tab.c (MUSCLE_COMMON_DECODE, muscle_string_decode)
(muscle_location_decode): Not related to muscles, rename as...
(COMMON_DECODE, string_decode, location_decode): these.
Akim Demaille [Thu, 26 Jul 2012 09:19:18 +0000 (11:19 +0200)]
obstack_escape: escape M4 characters
* src/muscle-tab.h (MUSCLE_OBSTACK_SGROW): This is not related to
muscles, so move to, and rename as...
* src/system.h (obstack_escape): this.
Adjust dependencies.
Akim Demaille [Thu, 26 Jul 2012 10:16:54 +0000 (12:16 +0200)]
remove dead macro
* src/system.h (DEFAULT_TMPDIR): Remove, unused.
Akim Demaille [Wed, 25 Jul 2012 15:57:57 +0000 (17:57 +0200)]
maint: style changes
* src/scan-code.l: Remove useless braces.
Formatting changes.
Prefer NULL to 0.
* src/muscle-tab.c, src/system.h: Formatting changes.
Akim Demaille [Tue, 24 Jul 2012 11:24:20 +0000 (13:24 +0200)]
doc: avoid problems with case insensitive file systems
makeinfo --html generates index.html, and the node "Index" will result
in Index.html. On case insensitive file systems, such as on Mac OS X
by default, this results in a single, invalid, file (Texinfo 4.13).
See http://lists.gnu.org/archive/html/bug-texinfo/2012-07/msg00032.html
* doc/bison.texi (Index): Rename as...
(Index of Terms): this.
Stefano Lattarini [Sat, 21 Jul 2012 15:24:23 +0000 (17:24 +0200)]
configure: fix botched quoting
* configure.ac: In the AC_SUBST call on 'VALGRIND_PREBISON'. Without
this change, when running ./configure, I see:
...
checking for valgrind... valgrind
./configure: line 35221: -q: command not found
checking for Java compiler... gcj -C -fsource=1.3 -ftarget=1.4
...
Copyright-paperwork-exempt: yes
Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
Akim Demaille [Fri, 20 Jul 2012 08:00:51 +0000 (10:00 +0200)]
news: fix typo.
* NEWS: here.
Reported by Ben Pfaff.