The "shadows a global declaration" warning in GCC 4.0 was a bit
annoying. It does not like that a type name be used in a prototype of
a function (not the implementation, just the declaration):
In file included from src/LR0.c:38:
src/reader.h:56: warning: declaration of 'named_ref' shadows a
global declaration
src/named-ref.h:35: warning: shadowed declaration is here
It does not like either when a global variable name is used in a
prototype. Flex 2.5.37 generates this prototype:
void gram_set_debug (int debug_flag );
* src/getargs.h, src/getargs.c (debug_flag): Rename as...
(debug): this.
Adjust dependencies.
* src/reader.h: Don't use "named_ref" as a formal argument name.
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.
* 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.
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.
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.
* 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.
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.
$ 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.
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.
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.
* 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.
* 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++.
* 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.
* 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.
* 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 "".
* 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.
* 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.
Using http://github.com/lyda/misspell-check, massage its
output into sed commands to perform the suggested changes.
Initially, I filtered out the THRU->Through changes, because
that failed to retain capitalization in the grammar token.
Instead, do this manually, beforehand:
sed -i s/THRU/THROUGH/ tests/existing.at
git ls-files|misspellings -f -|perl -nl \
-e '/^(.*?)\[(\d+)\]: (\w+) -> "(.*?)"$/ or next;' \
-e '($file,$n,$l,$r)=($1,$2,$3,$4); $q="'\''"; $r=~s/$q/$q\\$q$q/g;'\
-e 'print "sed -i $q${n}s!$l!$r!$q $file"'|bash
Some versions of Flex, possibly modified by the distribution package
maintainers, have incompatible signatures. Since newer versions of
Flex prototype their functions, avoid the conflicts in that case.
Reported by Stefano Lattarini.
<http://lists.gnu.org/archive/html/bug-bison/2012-05/msg00012.html>.
* src/flex-scanner.h (FLEX_VERSION_GT): New.
Use it to issue prototypes for flex-generated functions only for
versions up to 2.5.31, in accordance with the comment.
See commit dc9701e848.
Use it to define yylex_destroy when needed.
* src/parse-gram.y (add_param): Use xmemdup0 in place of
xmalloc+memcpy, and strspn in place of an open-coded loop.
Co-authored-by: Akim Demaille <akim@lrde.epita.fr>
* src/output.c (output_skeleton): Use memcpy, not strncpy,
since the source is known to fit in the destination buffer.
* src/parse-gram.y (%skeleton): Likewise.
