The name "$end" is nice in the report, in particular it avoids that
pointed-rules (aka items) be too long. It also helps keeping them
"standard".
But it is bad in error messages, we should report "end of file" (or
maybe "end of input", this is debatable). So, unless the user already
defined the alias for the error token herself, make it "end of file".
It should even be translated if the user already translated some
tokens, so that there is now no strong reason to redefine the $end
token.
* src/output.c (prepare_symbol_names): Issue "end of file" instead of
"$end".
* data/skeletons/lalr1.java (yytnamerr_): Remove the renaming hack.
* build-aux/update-test: Accept files with names containing a "+",
such as c++.at.
* tests/actions.at, tests/c++.at, tests/conflicts.at,
* tests/glr-regression.at, tests/regression.at, tests/skeletons.at:
Adjust.
Currently EOF is handled in an adhoc way, with a #define YYEOF 0 in
the implementation file. As a result, the user has to define her own
EOF token if she wants to use it, which is a pity.
Give the $end token a visible kind name, YYEOF. Except that in C,
where enums are not scoped, we would have collisions between all the
definitions of YYEOFs in the header files, so in C, make it
<api.PREFIX>EOF.
* data/skeletons/c.m4 (YYEOF): Override its name to avoid collisions.
Unless the user already gave it a different name.
* data/skeletons/glr.c (YYEOF): Remove.
Use ]b4_symbol(0, [id])[ instead.
Add support for "pre_epilogue", for glr.cc.
* data/skeletons/glr.cc: Remove dead code (never emitted #undefs).
* data/skeletons/yacc.c
* src/parse-gram.c
* src/reader.c
* src/symtab.c
* tests/actions.at
* tests/input.at
Since Bison 2.7, output was indented four spaces for explanatory
statements. For example:
input.y:2.7-13: error: %type redeclaration for exp
input.y:1.7-11: previous declaration
Since the introduction of caret-diagnostics, it became less clear.
Remove the indentation and display submessages as in GCC:
input.y:2.7-13: error: %type redeclaration for exp
2 | %type <float> exp
| ^~~~~~~
input.y:1.7-11: note: previous declaration
1 | %type <int> exp
| ^~~~~
* src/complain.h (SUB_INDENT): Remove.
(warnings): Add "note" to the enum.
* src/complain.h, src/complain.c (complain_indent): Replace by...
(subcomplain): this.
Adjust all dependencies.
* tests/actions.at, tests/diagnostics.at, tests/glr-regression.at,
* tests/input.at, tests/named-refs.at, tests/regression.at:
Adjust expectations.
The C, C++ and D skeletons used to show the stack right after popping
the stack during the reduction. Now that the stack is printed after
reaching a new state, that has become useless:
Entering state 1
Stack now 0 1
Reducing stack by rule 5 (line 83):
$1 = token "number" (1)
-> $$ = nterm exp (1)
Stack now 0
Entering state 8
Stack now 0 8
Remove the "Stack now 0" line.
* data/skeletons/lalr1.cc, data/skeletons/lalr1.d,
* data/skeletons/lalr1.java, data/skeletons/yacc.c:
Here.
Currently, if we have long rules and series of shift, we stack states
without showing stack. Let's be more incremental, and do how the Java
skeleton does.
* data/skeletons/lalr1.cc, data/skeletons/lalr1.d,
* data/skeletons/yacc.c:
Here.
Adjust test cases.
* tests/torture.at (AT_DATA_STACK_TORTURE): Disable stack traces: this
test produces a very large stack, and showing the stack each time we
shift a token goes quadatric.
The Java skeleton displays
Reading a token:
Next token is token "number" (1)
while the other display
Reading a token: Next token is token "number" (1)
When generating logs in the scanner, the first part is separated from
the second, and the end of the scanner logs have the second part
pasted in. So let's propagate the Java way, but with the colon.
* data/skeletons/glr.c, data/skeletons/lalr1.cc, data/skeletons/lalr1.d,
* data/skeletons/lalr1.java, data/skeletons/yacc.c: Do it.
Adjust test cases and doc.
The current code for yysyntax_error for %define parse.error verbose is
fishy (given that YYEMPTY is -2, invalid argument for yytname[]):
static int
yysyntax_error ([...])
{
YYPTRDIFF_T yysize0 = yytnamerr (YY_NULLPTR, yytname[yytoken]);
[...]
if (yytoken != YYEMPTY)
A nearby comment reports
The only way there can be no lookahead present (in yychar) is if
this state is a consistent state with a default action. Thus,
detecting the absence of a lookahead is sufficient to determine
that there is no unexpected or expected token to report. In that
case, just report a simple "syntax error".
So it _is_ possible to call yysyntax_error with yytoken == YYEMPTY,
albeit quite difficult when meaning to, so virtually impossible by
accident (after all, there was never a bug report about this).
I failed to produce a test case, but Joel E. Denny provided me with
one (added to the test suite below). The yacc.c skeleton fails on
this, and once fixed dies on a second problem. The glr.c skeleton was
also dying, but immediately of this second problem.
Indeed we were not allocating space for the error message's final \0.
This was hidden by the fact that we only had error messages with at
least an unexpected token displayed, so with at least one "%s" in the
format string, whose size (2) was included (incorrectly) in the final
size of the message (where the %s have been replaced by the actual
content).
* data/skeletons/glr.c, data/skeletons/yacc.c (yysyntax_error):
Do not invoke yytnamerr on YYEMPTY.
Clarify the computation of the length of the _final_ error message,
with the NUL terminator but without the '%s's.
* tests/conflicts.at (Syntax error in consistent error state):
New, contributed by Joel E. Denny.
We still have a few old C casts in lalr1.cc, let's get rid of them.
Reported by Frank Heckenbach.
Actually, let's monitor all our casts using easy to grep macros.
Let's use these macros to use the C++ standard casts when we are in
C++.
* data/skeletons/c.m4 (b4_cast_define): New.
* data/skeletons/glr.c, data/skeletons/glr.cc,
* data/skeletons/lalr1.cc, data/skeletons/stack.hh,
* data/skeletons/yacc.c:
Use it and/or its casts.
* tests/actions.at, tests/cxx-type.at,
* tests/glr-regression.at, tests/headers.at, tests/torture.at,
* tests/types.at:
Use YY_CAST instead of C casts.
* configure.ac (warn_cxx): Add -Wold-style-cast.
* doc/bison.texi: Disable it.
This patch contains more fixes to prefer signed to unsigned
integer types, as modern tools like 'gcc -fsanitize=undefined'
can check for signed integer overflow but not unsigned overflow.
* NEWS: Document the API change.
* boostrap.conf (gnulib_modules): Add intprops.
* data/skeletons/glr.c: Include stddef.h and stdint.h,
since this skeleton can assume C99 or later.
(YYSIZEMAX): Now signed, and the minimum of SIZE_MAX and PTRDIFF_MAX.
(yybool) [!__cplusplus]: Now signed (which is how bool behaves).
(YYTRANSLATE): Avoid use of unsigned, and make the macro
safe even for values greater than UINT_MAX.
(yytnamerr, struct yyGLRState, struct yyGLRStateSet, struct yyGLRStack)
(yyaddDeferredAction, yyinitStateSet, yyinitGLRStack)
(yyexpandGLRStack, yymarkStackDeleted, yyremoveDeletes)
(yyglrShift, yyglrShiftDefer, yy_reduce_print, yydoAction)
(yyglrReduce, yysplitStack, yyreportTree, yycompressStack)
(yyprocessOneStack, yyreportSyntaxError, yyrecoverSyntaxError)
(yyparse, yy_yypstack, yypstack, yypdumpstack):
* tests/input.at (Torturing the Scanner):
Prefer ptrdiff_t to size_t.
* data/skeletons/c++.m4 (b4_yytranslate_define):
* src/AnnotationList.c (AnnotationList__computePredecessorAnnotations):
* src/AnnotationList.h (AnnotationIndex):
* src/InadequacyList.h (InadequacyListNodeCount):
* src/closure.c (closure_new):
* src/complain.c (error_message, complains, complain_indent)
(complain_args, duplicate_directive, duplicate_rule_directive):
* src/gram.c (nritems, ritem_print, grammar_dump):
* src/ielr.c (ielr_compute_ritem_sees_lookahead_set)
(ielr_item_has_lookahead, ielr_compute_annotation_lists)
(ielr_compute_lookaheads):
* src/location.c (columns, boundary_print, location_print):
* src/muscle-tab.c (muscle_percent_define_insert)
(muscle_percent_define_check_values):
* src/output.c (prepare_rules, prepare_actions):
* src/parse-gram.y (id, handle_require):
* src/reader.c (record_merge_function_type, packgram):
* src/reduce.c (nuseless_productions, nuseless_nonterminals)
(inaccessable_symbols):
* src/relation.c (relation_print):
* src/scan-code.l (variant, variant_table_size, variant_count)
(variant_add, get_at_spec, show_sub_message, show_sub_messages)
(parse_ref):
* src/scan-gram.l (<SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>)
(scan_integer, convert_ucn_to_byte, handle_syncline):
* src/scan-skel.l (at_complain):
* src/symtab.c (complain_symbol_redeclared)
(complain_semantic_type_redeclared, complain_class_redeclared)
(symbol_class_set, complain_user_token_number_redeclared):
* src/tables.c (conflict_tos, conflrow, conflict_table)
(conflict_list, save_row, pack_vector):
* tests/local.at (AT_YYLEX_DEFINE(c)):
Prefer signed to unsigned integer.
* data/skeletons/lalr1.cc (yy_lac_check_):
* tests/actions.at (_AT_CHECK_PRINTER_AND_DESTRUCTOR):
* tests/local.at (AT_YYLEX_DEFINE(c)):
Omit now-unnecessary casts.
* data/skeletons/location.cc (b4_location_define):
* doc/bison.texi (Mfcalc Lexer, C++ position, C++ location):
Prefer int to unsigned for line and column numbers.
Change example to abort explicitly on memory exhaustion,
and fix an off-by-one bug that led to undefined behavior.
* data/skeletons/stack.hh (stack::operator[]):
Also allow ptrdiff_t indexes.
(stack::pop, slice::slice, slice::operator[]):
Index arg is now ptrdiff_t, not int.
(stack::ssize): New method.
(slice::range_): Now ptrdiff_t, not int.
* data/skeletons/yacc.c (b4_state_num_type): Remove.
All uses replaced by b4_int_type.
(YY_CONVERT_INT_BEGIN, YY_CONVERT_INT_END): New macros.
(yylac, yyparse): Use them around conversions that -Wconversion
would give false alarms about. Omit unnecessary casts.
(yy_stack_print): Use int rather than unsigned, and omit
a cast that doesn’t seem to be needed here any more.
* examples/c++/variant.yy (yylex):
* examples/c++/variant-11.yy (yylex):
Omit no-longer-needed conversions to unsigned.
* src/InadequacyList.c (InadequacyList__new_conflict):
Don’t assume *node_count is unsigned.
* src/output.c (muscle_insert_unsigned_table):
Remove; no longer used.
Currently we remove the rhs to install %empty instead.
* src/reader.c (grammar_rule_check_and_complete): Insert the missing
%empty in front of the rhs, not in replacement thereof.
* tests/actions.at (Add missing %empty): Check that.
The "identifier and colon" of a rule is implemented as a single token,
but whose location is only that of the identifier (so that messages
about the lhs of a rule are accurate). When reducing empty rules, the
default location is the single point location on the end of the
previous symbol. As a consequence, when Bison parses a grammar, the
location of the right-hand side of an empty rule is based on the
lhs, *independently of the position of the colon*. And the colon can
be way farther, separated by comments, white spaces, including empty
lines.
As a result, some messages look really bad. For instance:
$ cat foo.y
%%
foo : /* empty */
bar
: /* empty */
gives
$ bison -Wall foo.y
foo.y:2.4: warning: empty rule without %empty [-Wempty-rule]
2 | foo : /* empty */
| ^
foo.y:3.4: warning: empty rule without %empty [-Wempty-rule]
3 | bar
| ^
The carets are not at the right column, not even the right line.
This commit passes the colon "again" after the "id colon" token, which
gives more accurate locations for these messages:
$ bison -Wall foo.y
foo.y:2.10: warning: empty rule without %empty [-Wempty-rule]
2 | foo : /* empty */
| ^
foo.y:4.2: warning: empty rule without %empty [-Wempty-rule]
4 | : /* empty */
| ^
* src/scan-gram.l (SC_AFTER_IDENTIFIER): Rollback the colon, so that
we scan it again afterwards.
(INITIAL): Scan colons.
* src/parse-gram.y (COLON): New.
(rules): Parse the colon after the rule's id_colon (and possible
named reference).
* tests/actions.at, tests/conflicts.at, tests/diagnostics.at,
* tests/existing.at: Adjust.
Currently, when we quote the source file, we indent it with one space,
and preserve tabulations, so there is a discrepancy and the visual
rendering is bad. One way out is to indent with a tab instead of a
space, but then this space can be used for more information. This is
what GCC9 does. Let's play copy cats.
See
https://lists.gnu.org/archive/html/bison-patches/2019-04/msg00025.htmlhttps://developers.redhat.com/blog/2019/03/08/usability-improvements-in-gcc-9/https://gcc.gnu.org/onlinedocs/gccint/Guidelines-for-Diagnostics.html#Guidelines-for-Diagnostics
* src/location.c (location_caret): Prefix quoted lines with the line
number and a pipe, fitting 8 columns.
* tests/actions.at, tests/c++.at, tests/conflicts.at,
* tests/diagnostics.at, tests/input.at, tests/java.at,
* tests/named-refs.at, tests/reduce.at, tests/regression.at,
* tests/sets.at: Adjust expectations.
Partly by "./build-aux/update-test tests/testsuite.dir/*/testsuite.log"
repeatedly, and partly by hand.
Currently the caller must specify the ./ prefix to its command. Let's
avoid that: it will be nicer to read, make it easier to have a version
that works for Java and C/C++.
* tests/local.at (AT_PARSER_CHECK): Prefix the command with ./.
Adjust callers.
Reported by Uxio Prego.
https://lists.gnu.org/archive/html/help-bison/2018-11/msg00031.html
We also need to move the unreachable 'goto' to a reachable place,
otherwise clang complains about the code being unreachable anyway.
See also https://bugs.llvm.org/show_bug.cgi?id=39736.
Interestingly, we don't have to apply that trick to
`#define YYCDEBUG if (false) std::cerr`, clang does not warn when the
code comes from macro expansion.
* configure.ac: Use -Wunreachable-code when supported.
* data/lalr1.cc, data/yacc.c: Pacify clang's warning about `if (0)`
by using a macro.
Another possibility was to move this statement to a reachable place.
* tests/actions.at, tests/c++.at: Avoid generating unreachable code.
The previous name is too obscure, and the other macros for C++ use
CXX, not CC.
* tests/local.at (AT_SKEL_CC_IF, AT_SKEL_JAVA_IF): Rename as...
(AT_CXX_IF, AT_JAVA_IF): these.
Adjust callers.
The code was already using midrule only, never mid_rule. This is
simpler to remember, and matches a similar change we made from
look-ahead to lookahead.
* NEWS, doc/bison.texi, src/reader.c, src/scan-code.h, src/scan-code.l
* tests/actions.at, tests/c++.at, tests/existing.at: here.
* tests/actions.at: Fix some log messages.
Prefer #error to fprintf: it fixes the invalid use of yyoutput in
%destructor, and it is an even stronger check: that the code is not
even emitted. The portability of #error is not really a problem here,
since the point is anyway to have the compilation fail.
* tests/input.at (_AT_UNUSED_VALUES_DECLARATIONS): Check
typed mid-rule actions.
* tests/report.at (Reports): Check that types of typed mid-rule
actions are reported.
* tests/actions.at (Typed mid-rule actions): Check that
the values of typed mid-rule actions are correct.
* data/lalr1.cc, doc/bison.texi, etc/bench.pl.in, examples/variant.yy,
* tests/actions.at, tests/atlocal.in, tests/c++.at, tests/headers.at,
* tests/local.at, tests/types.at:
Don't use std::endl, it flushes uselessly, and is considered bad
style.
Currently lalr1.cc makes an out-of-bound access when trying to read @1
in rules with an empty rhs (i.e., when there is no @1) that raises an
error (YYERROR).
glr.c already gracefully handles this by using @$ as initial location
for the errors. Let's do that in yacc.c and lalr1.cc.
* data/lalr1.cc, data/yacc.c: Use @$ to initialize the error location.
* tests/actions.at: Check that case.
There are warnings (-Wextra) in generated C++ code:
ltlparse.cc: In member function 'ltlyy::parser::symbol_number_type
ltlyy::parser::by_state::type_get() const':
ltlparse.cc:452:33: warning: enumeral and non-enumeral type in
conditional expression
return state == empty_state ? empty_symbol : yystos_[state];
Reported by Alexandre Duret-Lutz.
It turns out that -Wall and -Wextra were disabled because of a stupid
typo.
* configure.ac: Fix the stupid typo.
* data/lalr1.cc, src/AnnotationList.c, src/InadequacyList.c,
* src/ielr.c, src/print.c, src/scan-code.l, src/symlist.c,
* src/symlist.h, src/symtab.c, src/tables.c, tests/actions.at,
* tests/calc.at, tests/cxx-type.at, tests/glr-regression.at,
* tests/named-refs.at, tests/torture.at:
Fix warnings, mostly issues about variables used only with assertions,
which are disabled with -DNDEBUG.
This fixes test 241 on xLC:
"input.y", line 42.11: 1540-0274 (S) The name lookup for "report" did not find a declaration.
"input.y", line 42.11: 1540-1292 (I) Static declarations are not considered for a function call if the function is not qualified.
where report is:
static void
report (std::ostream& yyo, int ival, float fval)
{
yyo << "ival: " << ival << ", fval: " << fval;
}
and line 42 is:
%printer { report (yyo, $$, $<fval>$); } <ival>;
It turns out that indeed this function must not be declared static,
<http://stackoverflow.com/a/17662745/1353549>. Let's put it into an
anonymous namespace.
Reported by Thomas Jahns.
http://lists.gnu.org/archive/html/bug-bison/2015-01/msg00059.html
* tests/actions.at (Qualified $$ in actions): Don't use "static",
prefer anonymous namespace.
-Wsign-compare was disabled for bison's own code, following gnulib's
approach. However, the generated parsers should not trigger such
warnings.
Reported by Efi Fogel.
http://lists.gnu.org/archive/html/help-bison/2013-04/msg00018.html
See also http://stackoverflow.com/questions/16101062 for the weird
"-(unsigned)i" piece of code.
* configure.ac (warn_tests): Enable -Wsign-compare.
* data/location.cc (position::add_): New.
(position::lines, position::columns): Use it.
* tests/actions.at (AT_CHECK_PRINTER_AND_DESTRUCTOR): Fix signedness issues.
