I never understood why we book ngotos+1 slots for relations between
gotos: there are at most ngotos images, not ngotos+1 (and "includes"
does have cases where a goto is in relation with itself, so it's not
ngotos-1).
Maybe bbf37f2534 explains the +1: a bug
left us register a goto several times on occasion, and the +1 might
have been a means to avoid this problem in most cases. Now that this
bug is addressed, we should no longer overbook memory, if only for the
clarity of the code ("why ngotos+1 instead of ngotos?").
* src/lalr.c: A goto has at most ngotos images, not ngotos+1.
While at it, avoid useless repeated call to map_goto introduced in
bbf37f2534.
Reported by wcventure.
http://lists.gnu.org/archive/html/bug-bison/2019-03/msg00008.html
* src/symtab.c (complain_class_redeclared): Don't print empty
locations.
There can only be empty locations for predefined symbols. And the
only symbol that is lexically available is the error token. So this
appears to be the only possible way to have an error involving an
empty location.
* tests/input.at (Symbol class redefinition): Check it.
The "includes" relation [DeRemer 1982] is between gotos, so of course,
for a given goto, there cannot be more that ngotos (number of gotos)
images. But we manipulate the set of images of a goto as a list,
without checking that an image was not already introduced. So we can
"register" way more images than ngotos, leading to a crash (heap
buffer overflow).
Reported by wcventure.
http://lists.gnu.org/archive/html/bug-bison/2019-03/msg00007.html
For the records, this bug is present in the first committed version of
Bison.
* src/lalr.c (build_relations): Don't insert the same goto several
times.
* tests/sets.at (Build Relations): New.
It seems that not many people read these logs: the error was
introduced in 2001 (3067fbef53),
* src/gram.c (grammar_dump): Fix the headers of the table: remove
duplicate display of "Ritem Range".
While at it, remove duplicate display of the rule number (and remove
an incorrect comment about it: these numbers _are_ equal).
* tests/sets.at (Reduced Grammar): Use useless rule, nterm and token
in the example.
Suggested by Yuri at
<http://lists.gnu.org/archive/html/bison-patches/2012-01/msg00000.html>.
The improvement is marginal for most grammars, but notable for large
grammars (e.g., PosgreSQL's postgre.y), and very large for the
sample.y grammar submitted by Yuri in
http://lists.gnu.org/archive/html/bison-patches/2012-01/msg00012.html.
Measured with --trace=time -fsyntax-only.
parser action tables postgre.y sample.y
Before 0,129 (44%) 37,095 (99%)
After 0,117 (42%) 5,046 (93%)
* src/tables.c (pos): Replace this set of integer coded as an unsorted
array or integers with...
(pos_set): this bitset.
* src/reader.c (grammar_rule_check_and_complete): When 'p' and 'lhs'
are aliases, prefer the latter, for clarity and consistency.
(grammar_current_rule_begin): Avoid 'p', current_rule suffices.
* src/gram.h, src/gram.c: Comment changes.
ptdr# calc.tab.c
Bruno Haible just added a default implementation of libtextstyle's
interface when the library is not available.
https://lists.gnu.org/archive/html/bison-patches/2019-03/msg00025.html
* gnulib: Update.
* bootstrap.conf: Replace libtextstyle with libtextstyle-optional.
* src/complain.c, src/getargs.c: Remove now useless cpp guards.
The variable spec_defines_file denotes the name of the generated
header. Its name is derived from --defines/%defines, whose name in
turn is derived from the fact that the header, in Yacc, contained the
Not only does the header now contain a lot more than just the token
definitions, but we no longer even generate macros, but an enum...
Let's modernize our vocabulary.
* src/files.h, src/files.c (spec_defines_file): Rename as...
(spec_header_file): this.
Currently when --defines is used, we generate a header, and paste an
exact copy of it into the generated parser implementation file. Let's
provide a means to #include it instead.
We don't do it by default because of the Autotools' ylwrap. This
program wraps invocations of yacc (that uses a fixed output name:
y.tab.c, y.tab.h, y.output) to support a more modern naming
scheme (dir/foo.y -> dir/foo.tab.c, dir/foo.tab.h, etc.). It does
that by renaming the generated files, and then by running sed to
propagate these renamings inside the files themselves.
Unfortunately Automake's Makefiles uses Bison as if it were Yacc (with
--yacc or with -o y.tab.c) and invoke bison via ylwrap. As a
consequence, as far as Bison is concerned, the output files are
y.tab.c and y.tab.h, so it emits '#include "y.tab.h"'. So far, so
good. But now ylwrap processes this '#include "y.tab.h"' into
'#include "dir/foo.tab.h"', which is not guaranteed to always work.
So, let's do the Right Thing when the output file is not y.tab.c, in
which case the user should %define api.header.include. Binding this
behavior to --yacc is tempting, but we recently told people to stop
using --yacc (as it also enables the Yacc warnings), but rather to use
-o y.tab.c.
Yacc.c is the only skeleton concerned: all the others do include their
header.
* data/skeletons/yacc.c (b4_header_include_if): New.
(api.header.include): Provide a default value when the output is not
y.tab.c.
* src/parse-gram.y (api.header.include): Define.
Running with CC='gcc-mp-8 -fsanitize=undefined' revealed Undefined
Behaviors.
https://lists.gnu.org/archive/html/bison-patches/2019-03/msg00008.html
* src/state.c (errs_new): Don't call memcpy with NULL as source.
* src/location.c (add_column_width): Don't assume that the column
argument is nonnegative: the scanner sometimes "backtracks" (e.g., see
ROLLBACK_CURRENT_TOKEN and DEPRECATED) in which case we can have
negative column numbers (temporarily).
Found in test 3 (Invalid inputs).
* 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.
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.
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.
The format is inconsistent. For instance most sections are
indented (including "Terminals unused in grammar" for instance), but
the sections "Terminals, with rules where they appear" and
"Nonterminals, with rules where they appear" are not. Let's indent
them. Also, these two sections try to wrap the output to avoid lines
too long. Yet we don't do that in the rest of the file, for instance
when listing the lookaheads of an item.
For instance in the case of Bison's parse-gram.output we go from:
Terminals, with rules where they appear
"end of file" (0) 0
error (256) 28 88
"string" <char*> (258) 9 13 16 17 20 23 24 109 116
[...]
Nonterminals, with rules where they appear
$accept (58)
on left: 0
input (59)
on left: 1, on right: 0
prologue_declarations (60)
on left: 2 3, on right: 1 3
prologue_declaration (61)
on left: 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 22 23 24
25 26 27 28 29, on right: 3
[...]
to
Terminals, with rules where they appear
"end of file" (0) 0
error (256) 28 88
"string" <char*> (258) 9 13 16 17 20 23 24 109 116
[...]
Nonterminals, with rules where they appear
$accept (58)
on left: 0
input (59)
on left: 1
on right: 0
prologue_declarations (60)
on left: 2 3
on right: 1 3
prologue_declaration (61)
on left: 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 22 23 24 25 26 27 28 29
on right: 3
[...]
* src/print.c (END_TEST): Remove.
(print_terminal_symbols): Don't try to wrap the output.
(print_nonterminal_symbols): Likewise.
Make two different lines for occurrences on the left, and occurrence
on the rhs of the rules.
Indent by 4 and 8, not 3.
* src/reduce.c (reduce_output): Indent by 4, not 3.
* tests/conflicts.at, tests/existing.at, tests/reduce.at,
* tests/regression.at, tests/report.at:
Adjust.
This should not be used to generate parsers. My point is actually to
facilitate debugging (when tweaking the generation of the LR(0)
automaton for instance, not carying -yet- about lookaheads).
* src/reader.c (prepare_percent_define_front_end_variables): Add lr(0).
* src/conflicts.c (set_conflicts): Be robust to reds not having
lookaheads at all.
* src/ielr.c (LrType, lr_type_get): Adjust.
(ielr): Implement support for LR(0).
* src/lalr.c (lalr_free): Don't free LA when it's not computed.
* maint:
maint: post-release administrivia
version 3.3.2
style: minor fixes
NEWS: named constructors are preferable to symbol_type ctors
gram: fix handling of nterms in actions when some are unused
style: rename local variable
CI: update the ICC serial number for travis-ci.org
Since Bison 3.3, semantic values in rule actions (i.e., '$...') are
passed to the m4 backend as the symbol number. Unfortunately, when
there are unused symbols, the symbols are renumbered _after_ the
numbers were used in the rule actions. As a result, the evaluation of
the skeleton failed because it used non existing symbol numbers.
Which is the happy scenario: we could use numbers of other existing
symbols...
Reported by Balázs Scheidler.
http://lists.gnu.org/archive/html/bug-bison/2019-01/msg00044.html
Translating the rule actions after the symbol renumbering moves too
many parts in bison. Relying on the symbol identifiers is more
troublesome than it might first seem: some don't have an
identifier (tokens with only a literal string), some might have a
complex one (tokens with a literal string with characters special for
M4). Well, these are tokens, but nterms also have issues: "dummy"
nterms (for midrule actions) are named $@32 etc. which is risky for
M4.
Instead, let's simply give M4 the mapping between the old numbers and
the new ones. To avoid confusion between old and new numbers, always
emit pre-renumbering numbers as "orig NUM".
* data/README: Give details about "orig NUM".
* data/skeletons/bison.m4 (__b4_symbol, _b4_symbol): Resolve the
"orig NUM".
* src/output.c (prepare_symbol_definitions): Pass nterm_map to m4.
* src/reduce.h, src/reduce.c (nterm_map): Extract it from
nonterminals_reduce, to make it public.
(reduce_free): Free it.
* src/scan-code.l (handle_action_dollar): When referring to a nterm,
use "orig NUM".
* tests/reduce.at (Useless Parts): New, based Balázs Scheidler's
report.
