* data/skeletons/yacc.c: here.
This is more logical for the time stamps, but it's also required by
following patches: the shared declarations are also in charge of
handling api.value.type=union. So far, they are run in the
implementation file in both cases (with or without header). But if we
run them only in the header, then the implementation file is emited
with incorrect support for api.value.type=union.
Arguably we should not have such dependencies. This is because we
have side-effects in our backend (redefining the symbols' type and
type_tag). In the future we should find a better solution for this,
without sacrificing the independence of the backend from bison
itself (i.e., I don't think we should handle api.value.type=union in
bison, leave it to m4).
Currently we generate things like:
#line 683 "src/parse-gram.y" /* yacc.c:316 */
The first part is of course very important: compilers point the users
to their grammar file rather than into the generated parser. The
second part points to the place in the skeletons that generated this
piece of code.
This dependency on the Bison skeletons generates lots of useless 'git
diff'. This location is useless for the regular user (who does not
care about the skeletons) and is actually not useful for Bison
developpers too (I never used this to locate the code in skeletons
that generated output). So disable it completely. If someone thinks
this was actually useful, a %define variable should be provided to
control the level of verbosity of '#line', in replacement of
--no-lines.
So now, generate:
#line 683 "src/parse-gram.y"
* data/skeletons/bison.m4 (b4_sync_end): Emit nothing.
* src/lalr.c (follows_print): Just print the symbol tag.
Take and print a title.
Indent the output.
Use it to print the various steps of the computation.
(lookahead_tokens_print): Fix a lie: the number displayed is not the
number of tokens.
Don't display states that don't even have reductions.
* src/relation.h, src/relation.c (relation_print): Accept and use a
title.
Don't print empty rows.
Indent the output.
Adjust dependencies.
* src/lalr.c (initialize_goto_follows): Print 'reads' in traces.
* configure.ac (DCFLAGS): Define.
* tests/atlocal.in: Receive it.
* data/skeletons/d.m4 (api.parser.class): Remove spurious YY.
* data/skeletons/lalr1.d (yylex): Return an int instead of a
YYTokenType, so that we can use characters as tokens.
* examples/d/calc.y: Adjust.
* tests/local.at: Initial support for D.
(AT_D_IF, AT_DATA_GRAMMAR(D), AT_YYERROR_DECLARE(d))
(AT_YYERROR_DECLARE_EXTERN(d), AT_YYERROR_DEFINE(d))
(AT_MAIN_DEFINE(d), AT_COMPILE_D, AT_LANG_COMPILE(d), AT_LANG_EXT(d)):
New.
* tests/calc.at: Initial support for D.
* tests/headers.at
* examples/d/calc.y: Exit with failure on errors.
Remove useless operators (=, !) meant for the test suite.
Add unary + for symmetry.
* examples/d/calc.test: Adjust expectations.
Commit db34f79889 renames the variable F
as goto_follows, but forgot to rename this function.
* src/lalr.c (initialize_F): Rename as...
(initialize_goto_follows): this.
I need to be able to read includes and goto_follows.
* src/relation.h, src/relation.c (relation_print): Provide a means to
pretty-print the nodes of the relation.
* src/lalr.c (goto_print, follows_print): New.
(set_goto_map): Use goto_print.
(build_relations): Show INCLUDES.
(compute_FOLLOWS): Rename as...
(compute_follows): this.
Show FOLLOWS.
* src/lalr.c (build_relations): Reduce the scopes.
Instead of keeping rp alive in two different loops, clarify the second
one by having an index on the path we traverse (i.e., use that index
to compute the source state _and_ the symbol that labels the
transition).
This allows to turn an obscure 'while'-loop in a clearer (IMHO)
'for'-loop. We also consume more variables (by introducing p instead
of making more side effects on length), but we're in 2019, I don't
think this matters. What does matter is that (IMHO again), this is
now clearer.
Also, use clearer names.
* tests/local.at (AT_MAIN_DEFINE(java)): Exit failure on failure.
(AT_PARSER_CHECK): If in Java, run AT_JAVA_PARSER_CHECK.
* tests/conflicts.at (AT_CONSISTENT_ERRORS_CHECK): Simplify.
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.
* tests/conflicts.at (AT_YYLEX_PROTOTYPE): Don't define it, leave that
task to AT_DATA_GRAMMAR.
But be honest: tell AT_BISON_OPTION_PUSHDEFS all the options we use.
Reported by Hans Åberg.
https://lists.gnu.org/archive/html/help-bison/2019-02/msg00064.html
* src/files.c (spec_graph_file): Use `*.gv` when 3.4 or better,
otherwise `*.dot`.
* src/parse-gram.y (handle_require): Pretend we are already 3.4.
* doc/bison.texi: Adjust.
* tests/local.at, tests/output.at: Exercise this.
Currently, we have errors:
./doc/bison.texi:13005: node `History' lacks menu item for `Yacc' despite being its Up target
* doc/bison.texi (History): Add the local menu.
While at it, add a few index entries.
* examples/c/lexcalc/parse.y: Formatting/comment changes.
(line): Don't return a value.
Print the result here, which avoids printing a value for lines with an
error.
(yyerror): Be sure to increment the pointed, not the pointer...
* examples/c/lexcalc/lexcalc.test: Check errors.
* examples/c/lexcalc/local.mk: Fix a dependency.
While hacking on the computation of the automaton, I had yystate being
equal to -1, and the parser loops. Let's catch this when
parser.assert is enabled.
* data/skeletons/yacc.c (YY_ASSERT): New.
Use it.
Not using the name YYASSERT, to make it clear that this is private.
glr.c should probably move to YY_ASSERT too.
Also, while at it, report 'Entering state...' even before growing the
stacks.
* examples/c/calc/local.mk, examples/c/lexcalc/local.mk,
* examples/c/mfcalc/local.mk, examples/c/rpcalc/local.mk:
Regenerate the files if dependencies have changed.
I don't think this style:
/* If buffer is full, make it bigger. */
if (i == length)
{
length *= 2;
symbuf = (char *) realloc (symbuf, length + 1);
}
/* Add this character to the buffer. */
symbuf[i++] = c;
/* Get another character. */
c = getchar ();
or more readable than
/* If buffer is full, make it bigger. */
if (i == length)
{
length *= 2;
symbuf = (char *) realloc (symbuf, length + 1);
}
/* Add this character to the buffer. */
symbuf[i++] = c;
/* Get another character. */
c = getchar ();
Actually, I think the latter is more readable, and helps with width
issues in the PDF. As a matter of fact, I would happily move to //
only for single line comments.
* doc/bison.texi: A single space before closing comments.
* doc/bison.texi: Prefer 'fun' to 'fnct'.
Reduce local variable scopes.
Prefer strdup to malloc + strcpy.
Avoid gratuitous casts.
Use simpler names (e.g., 'name' instead of 'fname').
Avoid uses of 0 for NULL.
Avoid using NULL when possible (e.g., 'p' instead of 'p != NULL').
Prefer union names to casts (e.g. 'yylval.VAR = s' instead of
'*((symrec**) &yylval) = s').
Give arguments a name in fun declarations.
Use our typedefs instead of duplicating them (func_t).
Stop promoting an explicit $$ = $1;, it should be implicit (Bison
might be able to eliminate useless chain rules).
Help a bit Texinfo by making smaller groups.
Rely on the C compiler to call function pointers (prefer
'$1->value.fun ($3)' to (*($1->value.fnctptr))($3)').
Currently we have no simple example: rpcalc in reverse Polish, mfcalc
has functions, and lexcalc is using lex.
* examples/c/calc/Makefile, examples/c/calc/calc.y,
* examples/c/calc/calc.test, examples/c/calc/local.mk: New.
