Currently the core of the initial state is limited to the single rule
on $accept.
* src/lr0.c (generate_states): There may now be several rules on
$accept.
* src/graphviz.c (conclude_red): Recognize "final" transitions by the
fact that we reduce to "$accept".
* src/print.c (print_reduction): Likewise.
* src/print-xml.c (print_reduction): Likewise.
Now that the parser can read several start symbols, let's process
them, and create the corresponding rules.
* src/parse-gram.y (grammar_declaration): Accept a list of start symbols.
* src/reader.h, src/reader.c (grammar_start_symbol_set): Rename as...
(grammar_start_symbols_set): this.
* src/reader.h, src/reader.c (start_flag): Replace with...
(start_symbols): this.
* src/reader.c (grammar_start_symbols_set): Build a list of start
symbols.
(switching_token, create_start_rules): New.
(check_and_convert_grammar): Use them to turn the list of start
symbols into a set of rules.
* src/reduce.c (nonterminals_reduce): Don't complain about $accept,
it's an internal detail.
(reduce_grammar): Complain about all the start symbols that don't
derive sentences.
* src/symtab.c (startsymbol, startsymbol_loc): Remove, replaced by
start_symbols.
symbols_pack): Move the check about the start symbols
to...
* src/symlist.c (check_start_symbols): here.
Adjust to multiple start symbols.
* tests/reduce.at (Empty Language): Generalize into...
(Bad start symbols): this.
Currently this example crashes on input such as "T (x) + y;".
The same example with glr.c works properly.
* examples/c++/glr/Makefile, examples/c++/glr/README.md,
* examples/c++/glr/c++-types.test, examples/c++/glr/c++-types.yy,
* examples/c++/glr/local.mk, examples/c++/local.mk: New.
Based on examples/c/glr/c++-types.y.
The D skeleton was not properly supporting @1 etc.
Reported by Adela Vais.
https://lists.gnu.org/r/bison-patches/2020-09/msg00049.html
* data/skeletons/d.m4 (b4_rhs_location): Fix it.
* tests/calc.at: Check the support of @n for all the skeletons.
Based on the test case 668 (cxx-type.at:437) "GLR: Merge conflicting
parses, pure, locations".
* examples/c/glr/Makefile, examples/c/glr/README.md,
* examples/c/glr/c++-types.test, examples/c/glr/c++-types.y,
* examples/c/glr/local.mk: New.
This is consistent with --defines being deprecated in favor of
--header. The directive %defines is also too similar to %define.
And %header matches nicely with api.header.name.
* src/scan-gram.l (%defines): Deprecate to %header.
(%header): Scan it.
* src/parse-gram.y (PERCENT_DEFINES): Replace with...
(PERCENT_HEADER): this.
* data/skeletons/lalr1.java
* doc/bison.texi
* tests/actions.at, tests/c++.at, tests/calc.at, tests/conflicts.at,
* tests/input.at, tests/java.at, tests/local.at, tests/output.at,
* tests/synclines.at, tests/types.at:
Convert most tests to check %header instead of %defines.
The name "defines" is incorrect, the generated file contains far more
than just #defines.
* src/getargs.h, src/getargs.c (-H, --header): New option.
With optional argument, just like --defines, --xml, etc.
(defines_flag): Rename as...
(header_flag): this.
Adjust dependencies.
* data/skeletons/bison.m4, data/skeletons/c.m4, data/skeletons/glr.c,
* data/skeletons/glr.cc, data/skeletons/glr2.cc, data/skeletons/lalr1.cc,
* data/skeletons/yacc.c:
Adjust.
* examples, doc/bison.texi: Adjust.
* tests/headers.at, tests/local.at, tests/output.at: Convert most
tests from using --defines to using --header.
231. conflicts.at:1096: testing Syntax error in consistent error state: glr2.cc ...
tests/conflicts.at:1096: $CXX $CXXFLAGS $CPPFLAGS $LDFLAGS -o input input.cc $LIBS
input.cc: In member function 'YYRESULTTAG glr_stack::yyresolveValue(glr_state*)':
input.cc:2674:36: error: 'yysval' may be used uninitialized in this function [-Werror=uninitialized]
Do not initialize the variable: this way ASAN can really make sure we
do set it to a proper value.
If we initialize it, ASAN would report nothing.
* data/skeletons/c.m4 (YY_IGNORE_MAYBE_UNINITIALIZED_BEGIN): Disable
GCC 4.6's -Wuninitialized.
* data/skeletons/glr2.cc: Disable the warning locally.
231. conflicts.at:1096: testing Syntax error in consistent error state: glr2.cc ...
tests/conflicts.at:1096: $CXX $CXXFLAGS $CPPFLAGS $LDFLAGS -o input input.cc $LIBS
input.cc: In function 'int yyparse(yy::parser&)':
input.cc:3147:41: error: 'yyarg' may be used uninitialized in this function [-Werror=maybe-uninitialized]
return yytnamerr_ (yytname_[yysymbol]);
^
input.cc:2058:34: note: 'yyarg' was declared here
yy::parser::symbol_kind_type yyarg[YYERROR_VERBOSE_ARGS_MAXIMUM];
^
* data/skeletons/glr2.cc (yyreportSyntaxError): Initialize yyarg.
Fix a warning triggered in GCC (at least from 4.6 to 4.9):
input.cc: In constructor 'glr_stack_item::glr_stack_item(bool)':
input.cc:1371:5: error: declaration of 'is_state' shadows a member of 'this' [-Werror=shadow]
: is_state_(is_state)
^
* data/skeletons/glr2.cc (glr_stack_item): Alpha-convert.
On the CI, tests fail with GCC 4.6 to GCC 6 as follows:
tests/synclines.at:440: COLUMNS=1000; export COLUMNS; NO_TERM_HYPERLINKS=1; export NO_TERM_HYPERLINKS; bison --color=no -fno-caret -o \"\\\"\".cc \"\\\"\".y
tests/synclines.at:440: $CXX $CXXFLAGS $CPPFLAGS $LDFLAGS -o \"\\\"\" \"\\\"\".cc $LIBS
stderr:
"\"".cc: In member function 'glr_state& glr_stack_item::getState()':
"\"".cc:1404:47: error: dereferencing type-punned pointer will break strict-aliasing rules [-Werror=strict-aliasing]
return *reinterpret_cast<glr_state*>(&raw_);
^
"\"".cc: In member function 'const glr_state& glr_stack_item::getState() const':
"\"".cc:1408:53: error: dereferencing type-punned pointer will break strict-aliasing rules [-Werror=strict-aliasing]
return *reinterpret_cast<const glr_state*>(&raw_);
^
"\"".cc: In member function 'semantic_option& glr_stack_item::getOption()':
"\"".cc:1413:53: error: dereferencing type-punned pointer will break strict-aliasing rules [-Werror=strict-aliasing]
return *reinterpret_cast<semantic_option*>(&raw_);
^
"\"".cc: In member function 'const semantic_option& glr_stack_item::getOption() const':
"\"".cc:1417:59: error: dereferencing type-punned pointer will break strict-aliasing rules [-Werror=strict-aliasing]
return *reinterpret_cast<const semantic_option*>(&raw_);
^
See also be6fa942ac.
* data/skeletons/glr2.cc (glr_stack_item): Use a temporary void*
variable to avoid type-punning issues with reinterpret_cast.
Currently the compiler attributes are defined in
b4_shared_declarations (that can in the header if it exists, otherwise
in the implementation file). This is not needed, only the
implementation file needs them.
Besides, glr2.cc was also defining these macros in the implementation
file, so we had two definitions.
* data/skeletons/glr.cc, data/skeletons/glr2.cc: Define the compiler
attribute macros only in the implementation files.
* tests/regression.at (Lex and parse params): Generate a header, to
make it easy to check that the header is self-sufficient.
input.cc: In constructor 'glr_stack_item::glr_stack_item(bool)':
input.cc:1423:5: error: declaration of 'isState' shadows a member of 'this' [-Werror=shadow]
: isState_(isState) {
^
test.cc:1165:45: error: declaration of 'begin' shadows a member of 'this' [-Werror=shadow]
test.cc:1167:45: error: declaration of 'end' shadows a member of 'this' [-Werror=shadow]
* data/skeletons/glr2.cc (isState): Rename as...
(is_state): this.
Formatting changes.
(reduceToOneStack): Rename variables to avoid name clashes.
Commit af000bab11 ("doc: work around
Texinfo 6.7 bug"), published in 3.4.91, added a dependency on the
"all" target.
This is a super bad idea, since "make all" will run this
target *before* "all", which builds bison. It turns out that this new
dependency actually needed bison to be built. So all the regular
process (i) build $(BUILT_SOURCES) and then (ii) build bison, was
wrecked since some of the $(BUILT_SOURCES) depended on bison...
It was "easy" to see in the logs of "make V=1" because we were
building bison files (such as src/files.o) *before* displaying the
banner for "all-recursive". With this fix, we finally get again the
proper sequence:
rm -f examples/c/reccalc/scan.stamp examples/c/reccalc/scan.stamp.tmp
/opt/local/libexec/gnubin/mkdir -p examples/c/reccalc
touch examples/c/reccalc/scan.stamp.tmp
flex -oexamples/c/reccalc/scan.c --header=examples/c/reccalc/scan.h ./examples/c/reccalc/scan.l
mv examples/c/reccalc/scan.stamp.tmp examples/c/reccalc/scan.stamp
rm -f lib/fcntl.h-t lib/fcntl.h && \
{ echo '/* DO NOT EDIT! GENERATED AUTOMATICALLY! */'; \
...
} > lib/fcntl.h-t && \
mv lib/fcntl.h-t lib/fcntl.h
...
mv -f lib/alloca.h-t lib/alloca.h
make all-recursive
Reported by Mingli Yu <[email protected]>.
https://github.com/akimd/bison/issues/31https://lists.gnu.org/r/bison-patches/2020-05/msg00055.html
Reported by Claudio Calvelli <[email protected]>.
https://lists.gnu.org/r/bug-bison/2020-09/msg00001.htmlhttps://bugs.gentoo.org/716516
* doc/local.mk (all): Rename as...
(all-local): this.
So that we don't compete with BUILT_SOURCES.
The yyerror stand-alone function was used to bounce from glr.c's call
to yyerror to glr.cc's parser.error. Now that glr.c is out of the
way, just directly use parser.error.
* data/skeletons/glr2.cc (yyerror): Remove.
Adjust callers.
(b4_yyerror_args, b4_lyyerror_args, b4_pure_formals): Remove.
Now unused.
On a case such as
%%
exp
: empty "a"
| "a" empty
empty
: %empty
we used to display
warning: shift/reduce conflict on token "a" [-Wcounterexamples]
Example: • "a"
Shift derivation
exp
↳ 2: • "a" empty
↳ 2: ε
Example: • "a"
Reduce derivation
exp
↳ 1: empty "a"
↳ 3: •
where the shift derivation shows an item "2: empty → ε", with an
explicit "ε", but the reduce derivation shows "3: empty → •", without
"ε".
For consistency, let's always show ε/%empty in rules with an empty
rhs:
Reduce derivation
exp
↳ 1: empty "a"
↳ 3: ε •
* src/derivation.c (derivation_width, derivation_print_tree_impl):
Always show ε/%empty in counterexamples.
* tests/diagnostics.at: Check that case.
* tests/conflicts.at, tests/counterexample.at: Adjust.
* data/skeletons/glr2.cc: Fix some documentation.
Be consistent between class/struct.
(yydoAction, yyresolveAction): Avoid passing yyparser where useless.
Reported by Dennis Clarke.
https://lists.gnu.org/r/bug-bison/2020-08/msg00013.html
* examples/d/local.mk, examples/java/calc/local.mk,
* examples/java/simple/local.mk: Pass bison's options before its
argument, in case we're in POSIX mode.
An assertion failed when the last character is a '\' and we're in a
character or a string.
Reported by Agency for Defense Development.
https://lists.gnu.org/r/bug-bison/2020-08/msg00009.html
* src/scan-gram.l: Catch unterminated escapes.
* tests/input.at (Unexpected end of file): New.
Reported by Agency for Defense Development.
https://lists.gnu.org/r/bug-bison/2020-08/msg00008.html
On an empty such as
%token FOO
BAR
FOO 0
%%
input: %empty
we crash because when we find FOO 0, we decrement ntokens (since FOO
was discovered to be EOF, which is already known to be a token, so we
increment ntokens for it, and need to cancel this). This "works well"
when EOF is properly defined in one go, but here it is first defined
and later only assign token code 0. In the meanwhile BAR was given
the token number that we just decremented.
To fix this, assign symbol numbers after parsing, not during parsing,
so that we also saw all the explicit token codes. To maintain the
current numbers (I'd like to keep no difference in the output, not
just equivalence), we need to make sure the symbols are numbered in
the same order: that of appearance in the source file. So we need the
locations to be correct, which was almost the case, except for nterms
that appeared several times as LHS (i.e., several times as "foo:
..."). Fixing the use of location_of_lhs sufficed (it appears it was
intended for this use, but its implementation was unfinished: it was
always set to "false" only).
* src/symtab.c (symbol_location_as_lhs_set): Update location_of_lhs.
(symbol_code_set): Remove broken hack that decremented ntokens.
(symbol_class_set, dummy_symbol_get): Don't set number, ntokens and
nnterms.
(symbol_check_defined): Do it.
(symbols): Don't count nsyms here.
Actually, don't count nsyms at all: let it be done in...
* src/reader.c (check_and_convert_grammar): here. Define nsyms from
ntokens and nnterms after parsing.
* tests/input.at (EOF redeclared): New.
* examples/c/bistromathic/bistromathic.test: Adjust the traces: in
"%nterm <double> exp %% input: ...", exp used to be numbered before
input.
* cfg.mk (_space_before_paren_exempt): Be less laxist.
* src/output.c, src/reader.c: Fix space before paren issues.
Pacify the warnings where applicable.
Older versions of GCC (4.1.2 here) don't like repeated typedefs.
CC src/bison-parse-simulation.o
src/parse-simulation.c:61: error: redefinition of typedef 'parse_state'
src/parse-simulation.h:74: error: previous declaration of 'parse_state' was here
make: *** [Makefile:7876: src/bison-parse-simulation.o] Error 1
Reported by Nelson H. F. Beebe.
* src/parse-simulation.c (parse_state): Don't typedef,
parse-simulation.h did it already.
This reverts commit d0bec3175f (which
should have read "We have a clash...", not "With have a clash...").
Now that `max()` was renamed `max_int()`, we can use `max` again, as
elsewhere in the code.
* src/counterexample.c (visited_hasher): Alpha reconversion.
Reported by Maarten De Braekeleer.
https://lists.gnu.org/r/bison-patches/2020-07/msg00080.html
We don't want to use gnulib's min and max macros, since we use
function calls in min/max arguments.
* src/location.c (max_int, min_int): Move to...
* src/system.h: here.
* src/counterexample.c, src/derivation.c: Use max_int instead of max.
From
foo.y:1.7-11: error: %type redeclaration for bar
1 | %type <foo> bar bar
| ^~~~~
foo.y:1.7-11: note: previous declaration
1 | %type <foo> bar bar
| ^~~~~
to
foo.y:1.17-19: error: %type redeclaration for bar
1 | %type <foo> bar bar
| ^~~
foo.y:1.13-15: note: previous declaration
1 | %type <foo> bar bar
| ^~~
* src/symlist.h, src/symlist.c (symbol_list_type_set): There's no need
for the tag's location, use that of the symbol.
* src/parse-gram.y: Adjust.
* tests/input.at: Adjust.
We crash if the input contains a string containing a NUL byte.
Reported by Suhwan Song.
https://lists.gnu.org/r/bug-bison/2020-07/msg00051.html
* src/flex-scanner.h (STRING_FREE): Avoid accidental use of
last_string.
* src/scan-gram.l: Don't call STRING_FREE without calling
STRING_FINISH first.
* tests/input.at (Invalid inputs): Check that case.
The original Texinfo macros introducing colors were made for
diagnostics, which are printed in bold. So by copy-paste accident the
styles we introduced for counterexamples were also in bold. They
should not.
* doc/bison.texi: Separate the styling of diagnostics from the styling
for counterexamples.
Don't use bold in the latter case.
With have a clash with the "max" function.
src/counterexample.c: In function 'visited_hasher':
src/counterexample.c:720:48: error: declaration of 'max' shadows a global declaration [-Werror=shadow]
src/counterexample.c:116:12: error: shadowed declaration is here [-Werror=shadow]
* src/counterexample.c (visited_hasher): Alpha conversion.
Fix 6b78e50cef, "cex: make "rerun with
'-Wcex'" a note instead of a warning"
* tests/conflicts.at (-W versus %expect and %expect-rr): Fix
expectations.
Unfortunately I found no way to use the ↳ glyph in Texinfo, so I used
@arrow{} instead, which has a different width, so we have to have all
the examples doubled, once for TeX, another for the rest of the world.
* doc/bison.texi: Use the current display in the examples.
* doc/calc.y, doc/ids.y, doc/if-then-else.y, doc/sequence.y: New.
Currently the suggestion to rerun is a -Wother warning:
warning: 2 shift/reduce conflicts [-Wconflicts-sr]
warning: rerun with option '-Wcounterexamples' to generate conflict counterexamples [-Wother]
Instead, let's attach it as a subnote of the diagnostic (in the
current case, -Wconflicts-sr):
warning: 2 shift/reduce conflicts [-Wconflicts-sr]
note: rerun with option '-Wcounterexamples' to generate conflict counterexamples
* src/conflicts.c (conflicts_print): Do that.
Adjust the test suite.
From
input.y: warning: reduce/reduce conflict on token $end [-Wcounterexamples]
Example: A b .
First derivation
a
`-> A b .
Second derivation
a
`-> A b
`-> b .
to
input.y: warning: reduce/reduce conflict on token $end [-Wcounterexamples]
Example: A b .
First reduce derivation
a
`-> A b .
Second reduce derivation
a
`-> A b
`-> b .
* src/counterexample.c (print_counterexample): here.
Compute the width of the labels to properly align the values.
* tests/conflicts.at, tests/counterexample.at, tests/diagnostics.at,
* tests/report.at: Adjust.
Now that the derivation is no longer printed on one line, aligning the
example and the derivation is no longer useful. It can actually be
harmful, as it makes the overall structure less clear.
* src/derivation.h, src/derivation.c (derivation_print_leaves): Remove
the `prefix` argument.
* src/counterexample.c (print_counterexample): Put the example next to
its label.
* tests/conflicts.at, tests/counterexample.at, tests/diagnostics.at,
* tests/report.at: Adjust.
Now that we use complain, the "sections" are clearer.
* src/counterexample.c (print_counterexample): Use the empty line only
in reports.
* tests/counterexample.at, tests/diagnostics.at, tests/report.at: Adjust.
Akim Demaille