diff options
author | Akim Demaille <akim@lrde.epita.fr> | 2012-07-27 16:22:45 +0200 |
---|---|---|
committer | Akim Demaille <akim@lrde.epita.fr> | 2012-07-27 16:35:49 +0200 |
commit | 4323e0dac386d777d070c68564f1c0041b06935d (patch) | |
tree | 187677008a28b67fc0e0f031f365f83b57f54ab6 | |
parent | 9e41ba026ab2e6c7e3a56c4a092a48059be05bc4 (diff) | |
parent | 7b18c112316339685df9ba15f44741c608a29145 (diff) | |
download | bison-4323e0dac386d777d070c68564f1c0041b06935d.tar.gz |
Merge remote-tracking branch 'origin/maint'
* origin/maint: (29 commits)
regen
synclines: remove spurious empty line
also support $<foo>$ in the %initial-action
skeletons: b4_dollar_pushdef and popdef to simpify complex definitions
regen
printer/destructor: translate only once
factor the handling of m4 escaping
news: schedule the removal of the ";" hack
style changes in the scanners
regen
support $<tag>$ in printers and destructors
scan-code: factor the handling of the type in $<TYPE>$
muscles: fix another occurrence of unescaped type name
glr.cc: fix the handling of yydebug
gnulib: update
formatting changes
tests: fix an assertion
tests: adjust to GCC 4.8, which displays caret errors
be sure to properly escape type names
obstack_quote: escape and quote for M4
muscles: shuffle responsabilities
muscles: make private functions static
muscles: rename private functions/macros
obstack_escape: escape M4 characters
remove dead macro
maint: style changes
doc: avoid problems with case insensitive file systems
configure: fix botched quoting
news: fix typo.
Conflicts:
NEWS
data/c.m4
data/glr.cc
data/lalr1.cc
examples/rpcalc/local.mk
src/muscle-tab.h
src/output.c
src/parse-gram.c
src/parse-gram.h
src/parse-gram.y
src/scan-code.l
src/symlist.c
src/symlist.h
src/symtab.h
tests/calc.at
-rw-r--r-- | .gitignore | 2 | ||||
-rw-r--r-- | NEWS | 37 | ||||
-rw-r--r-- | THANKS | 1 | ||||
-rw-r--r-- | TODO | 4 | ||||
-rw-r--r-- | configure.ac | 2 | ||||
-rw-r--r-- | data/bison.m4 | 12 | ||||
-rw-r--r-- | data/c-like.m4 | 43 | ||||
-rw-r--r-- | data/c.m4 | 1 | ||||
-rw-r--r-- | data/glr.c | 6 | ||||
-rw-r--r-- | data/glr.cc | 1 | ||||
-rw-r--r-- | data/java.m4 | 1 | ||||
-rw-r--r-- | data/lalr1.cc | 20 | ||||
-rw-r--r-- | data/lalr1.java | 6 | ||||
-rw-r--r-- | data/local.mk | 1 | ||||
-rw-r--r-- | data/yacc.c | 11 | ||||
-rw-r--r-- | doc/bison.texi | 31 | ||||
m--------- | gnulib | 0 | ||||
-rw-r--r-- | src/muscle-tab.c | 88 | ||||
-rw-r--r-- | src/muscle-tab.h | 83 | ||||
-rw-r--r-- | src/output.c | 30 | ||||
-rw-r--r-- | src/parse-gram.y | 13 | ||||
-rw-r--r-- | src/scan-code.h | 5 | ||||
-rw-r--r-- | src/scan-code.l | 155 | ||||
-rw-r--r-- | src/scan-skel.l | 32 | ||||
-rw-r--r-- | src/symlist.c | 9 | ||||
-rw-r--r-- | src/symlist.h | 5 | ||||
-rw-r--r-- | src/symtab.h | 12 | ||||
-rw-r--r-- | src/system.h | 87 | ||||
-rw-r--r-- | tests/actions.at | 137 | ||||
-rw-r--r-- | tests/input.at | 99 | ||||
-rw-r--r-- | tests/local.at | 2 | ||||
-rw-r--r-- | tests/synclines.at | 25 |
32 files changed, 668 insertions, 293 deletions
@@ -11,6 +11,7 @@ /INSTALL /Makefile /Makefile.in +/README-release /_* /a.exe /a.out @@ -31,4 +32,3 @@ /patches /releases /stamp-h* -/README-release @@ -129,6 +129,41 @@ GNU Bison NEWS * Noteworthy changes in release ?.? (????-??-??) [?] + Bison no longer executes user-specified M4 code when processing a grammar. + +** Future Changes + + In addition to the removal of the features announced in Bison 2.6, the + next major release will remove the "Temporary hack for adding a semicolon + to the user action", as announced in the release 2.5. Instead of: + + exp: exp "+" exp { $$ = $1 + $3 }; + + write: + + exp: exp "+" exp { $$ = $1 + $3; }; + +** Bug fixes + +*** Type names are now properly escaped. + +*** glr.cc: set_debug_level and debug_level work as expected. + +*** Stray @ or $ in actions + + While Bison used to warn about stray $ or @ in action rules, it did not + for other actions such as printers, destructors, or initial actions. It + now does. + +** Type names in actions + + For consistency with rule actions, it is now possible to qualify $$ by a + type-name in destructors, printers, and initial actions. For instance: + + %printer { fprintf (yyo, "(%d, %f)", $<ival>$, $<fval>$); } <*> <>; + + will display two values for each typed and untyped symbol (provided + that YYSTYPE has both "ival" and "fval" fields). * Noteworthy changes in release 2.6 (2012-07-19) [stable] @@ -137,7 +172,7 @@ GNU Bison NEWS The next major release of Bison will drop support for the following deprecated features. Please report disagreements to bug-bison@gnu.org. -*** K&C parsers +*** K&R C parsers Support for generating parsers in K&R C will be removed. Parsers generated for C support ISO C90, and are tested with ISO C99 and ISO C11 @@ -13,6 +13,7 @@ Anthony Heading ajrh@ajrh.net Arnold Robbins arnold@skeeve.com Art Haas ahaas@neosoft.com Baron Schwartz baron@sequent.org +Ben Pfaff blp@cs.stanford.edu Benoit Perrot benoit.perrot@epita.fr Bernd Kiefer kiefer@dfki.de Bert Deknuydt Bert.Deknuydt@esat.kuleuven.ac.be @@ -1,4 +1,8 @@ * Short term +** scan-code.l +Avoid variables for format strings, as then GCC cannot check them. +show_sub_messages should call show_sub_message. + ** Variable names. What should we name `variant' and `lex_symbol'? diff --git a/configure.ac b/configure.ac index cf42d2c6..328d9294 100644 --- a/configure.ac +++ b/configure.ac @@ -177,7 +177,7 @@ case $VALGRIND:$host_os in # VALGRIND+=' --suppressions=$(abs_top_srcdir)/build-aux/darwin11.4.0.valgrind' VALGRIND=;; *:*) - AC_SUBST([VALGRIND_PREBISON], [$VALGRIND -q]);; + AC_SUBST([VALGRIND_PREBISON], ["$VALGRIND -q"]);; esac AM_MISSING_PROG([AUTOM4TE], [autom4te]) diff --git a/data/bison.m4 b/data/bison.m4 index 1db75a8e..f2126ef4 100644 --- a/data/bison.m4 +++ b/data/bison.m4 @@ -376,19 +376,17 @@ m4_define([b4_symbol_action_location], # Same as in C, but using references instead of pointers. m4_define([b4_symbol_action], [b4_symbol_if([$1], [has_$2], -[m4_pushdef([b4_dollar_dollar], - [b4_symbol_value([(*yyvaluep)], - b4_symbol_if([$1], [has_type], - [b4_symbol([$1], [type])]))])dnl -m4_pushdef([b4_at_dollar], [(*yylocationp)])dnl +[b4_dollar_pushdef([(*yyvaluep)], + b4_symbol_if([$1], [has_type], + [m4_dquote(b4_symbol([$1], [type]))]), + [(*yylocationp)])dnl b4_symbol_case_([$1])[]dnl b4_syncline([b4_symbol([$1], [$2_line])], ["b4_symbol([$1], [$2_file])"]) b4_symbol([$1], [$2]) b4_syncline([@oline@], [@ofile@]) break; -m4_popdef([b4_at_dollar])dnl -m4_popdef([b4_dollar_dollar])dnl +b4_dollar_popdef[]dnl ])]) diff --git a/data/c-like.m4 b/data/c-like.m4 new file mode 100644 index 00000000..732d18d4 --- /dev/null +++ b/data/c-like.m4 @@ -0,0 +1,43 @@ + -*- Autoconf -*- + +# Common code for C-like languages (C, C++, Java, etc.) + +# Copyright (C) 2012 Free Software Foundation, Inc. + +# This program is free software: you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation, either version 3 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program. If not, see <http://www.gnu.org/licenses/>. + +# b4_dollar_dollar_(VALUE, FIELD, DEFAULT-FIELD) +# ---------------------------------------------- +# If FIELD (or DEFAULT-FIELD) is non-null, return "VALUE.FIELD", +# otherwise just VALUE. Be sure to pass "(VALUE)" is VALUE is a +# pointer. +m4_define([b4_dollar_dollar_], +[b4_symbol_value([$1], + m4_if([$2], [[]], + [[$3]], [[$2]]))]) + +# b4_dollar_pushdef(VALUE-POINTER, DEFAULT-FIELD, LOCATION) +# b4_dollar_popdef +# --------------------------------------------------------- +# Define b4_dollar_dollar for VALUE and DEFAULT-FIELD, +# and b4_at_dollar for LOCATION. +m4_define([b4_dollar_pushdef], +[m4_pushdef([b4_dollar_dollar], + [b4_dollar_dollar_([$1], m4_dquote($][1), [$2])])dnl +m4_pushdef([b4_at_dollar], [$3])dnl +]) +m4_define([b4_dollar_popdef], +[m4_popdef([b4_at_dollar])dnl +m4_popdef([b4_dollar_dollar])dnl +]) @@ -17,6 +17,7 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. +m4_include(b4_pkgdatadir/[c-like.m4]) # b4_tocpp(STRING) # ---------------- @@ -2294,12 +2294,10 @@ yyrecoverSyntaxError (yyGLRStack* yystackp]b4_user_formals[) #endif ]) m4_ifdef([b4_initial_action], [ -m4_pushdef([b4_at_dollar], [yylloc])dnl -m4_pushdef([b4_dollar_dollar], [yylval])dnl +b4_dollar_pushdef([yylval], [], [yylloc])dnl /* User initialization code. */ b4_user_initial_action -m4_popdef([b4_dollar_dollar])dnl -m4_popdef([b4_at_dollar])])dnl +b4_dollar_popdef])[]dnl [ if (! yyinitGLRStack (yystackp, YYINITDEPTH)) goto yyexhaustedlab; diff --git a/data/glr.cc b/data/glr.cc index 81e80278..7b3ea3f4 100644 --- a/data/glr.cc +++ b/data/glr.cc @@ -202,6 +202,7 @@ m4_pushdef([b4_parse_param], m4_defn([b4_parse_param_orig]))dnl void ]b4_parser_class_name[::set_debug_level (debug_level_type l) { + // Actually, it is yydebug which is really used. yydebug = l; } diff --git a/data/java.m4 b/data/java.m4 index cd70852c..988ac39a 100644 --- a/data/java.m4 +++ b/data/java.m4 @@ -17,6 +17,7 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. +m4_include(b4_pkgdatadir/[c-like.m4]) # b4_comment(TEXT) # ---------------- diff --git a/data/lalr1.cc b/data/lalr1.cc index 96936bbe..313dba27 100644 --- a/data/lalr1.cc +++ b/data/lalr1.cc @@ -98,19 +98,19 @@ m4_define([b4_rhs_location], # Same as in C, but using references instead of pointers. m4_define([b4_symbol_action], [b4_symbol_if([$1], [has_$2], -[m4_pushdef([b4_dollar_dollar], - [b4_symbol_value_template([yysym.value], - b4_symbol_if([$1], [has_type], - [b4_symbol([$1], [type])]))])dnl -m4_pushdef([b4_at_dollar], [yysym.location])dnl +[m4_pushdef([b4_symbol_value], m4_defn([b4_symbol_value_template]))[]dnl +b4_dollar_pushdef([yysym.value], + b4_symbol_if([$1], [has_type], + [m4_dquote(b4_symbol([$1], [type]))]), + [yysym.location])dnl b4_symbol_case_([$1]) b4_syncline([b4_symbol([$1], [$2_line])], ["b4_symbol([$1], [$2_file])"]) b4_symbol([$1], [$2]) b4_syncline([@oline@], [@ofile@]) break; -m4_popdef([b4_at_dollar])dnl -m4_popdef([b4_dollar_dollar])dnl +m4_popdef([b4_symbol_value])[]dnl +b4_dollar_popdef[]dnl ])]) @@ -670,12 +670,10 @@ m4_if(b4_prefix, [yy], [], YYCDEBUG << "Starting parse" << std::endl; ]m4_ifdef([b4_initial_action], [ -m4_pushdef([b4_at_dollar], [yyla.location])dnl -m4_pushdef([b4_dollar_dollar], [yyla.value])dnl +b4_dollar_pushdef([yyla.value], [], [yyla.location])dnl /* User initialization code. */ b4_user_initial_action -m4_popdef([b4_dollar_dollar])dnl -m4_popdef([b4_at_dollar])])dnl +b4_dollar_popdef])[]dnl [ /* Initialize the stack. The initial state will be set in yynewstate, since the latter expects the semantical and the diff --git a/data/lalr1.java b/data/lalr1.java index d3729c6e..1cfaef38 100644 --- a/data/lalr1.java +++ b/data/lalr1.java @@ -502,12 +502,10 @@ b4_lexer_if([[ yyerrstatus_ = 0; ]m4_ifdef([b4_initial_action], [ -m4_pushdef([b4_at_dollar], [yylloc])dnl -m4_pushdef([b4_dollar_dollar], [yylval])dnl +b4_dollar_pushdef([yylval], [], [yylloc])dnl /* User initialization code. */ b4_user_initial_action -m4_popdef([b4_dollar_dollar])dnl -m4_popdef([b4_at_dollar])])dnl +b4_dollar_popdef])[]dnl [ /* Initialize the stack. */ yystack.push (yystate, yylval]b4_locations_if([, yylloc])[); diff --git a/data/local.mk b/data/local.mk index 6337f769..b8db2d64 100644 --- a/data/local.mk +++ b/data/local.mk @@ -18,6 +18,7 @@ dist_pkgdata_DATA = \ data/bison.m4 \ data/c++-skel.m4 \ data/c++.m4 \ + data/c-like.m4 \ data/c-skel.m4 \ data/c.m4 \ data/glr.c \ diff --git a/data/yacc.c b/data/yacc.c index b3864e90..7376f900 100644 --- a/data/yacc.c +++ b/data/yacc.c @@ -1484,17 +1484,16 @@ b4_c_function_def([[yyparse]], [[int]], b4_parse_param)[ yylloc.first_column = yylloc.last_column = ]b4_location_initial_column[; #endif]]) m4_ifdef([b4_initial_action],[ -m4_pushdef([b4_at_dollar], [m4_define([b4_at_dollar_used])yylloc])dnl -m4_pushdef([b4_dollar_dollar], [m4_define([b4_dollar_dollar_used])yylval])dnl +b4_dollar_pushdef([m4_define([b4_dollar_dollar_used])yylval], [], + [m4_define([b4_at_dollar_used])yylloc])dnl /* User initialization code. */ b4_user_initial_action -m4_popdef([b4_dollar_dollar])dnl -m4_popdef([b4_at_dollar])])dnl +b4_dollar_popdef[]dnl m4_ifdef([b4_dollar_dollar_used],[[ yyvsp[0] = yylval; ]])dnl m4_ifdef([b4_at_dollar_used], [[ yylsp[0] = yylloc; -]])[ - goto yysetstate; +]])])dnl +[ goto yysetstate; /*------------------------------------------------------------. | yynewstate -- Push a new state, which is found in yystate. | diff --git a/doc/bison.texi b/doc/bison.texi index 2d06fec9..62ff0459 100644 --- a/doc/bison.texi +++ b/doc/bison.texi @@ -110,7 +110,7 @@ Reference sections: * Glossary:: Basic concepts are explained. * Copying This Manual:: License for copying this manual. * Bibliography:: Publications cited in this manual. -* Index:: Cross-references to the text. +* Index of Terms:: Cross-references to the text. @detailmenu --- The Detailed Node Listing --- @@ -4634,9 +4634,9 @@ code. @deffn {Directive} %initial-action @{ @var{code} @} @findex %initial-action Declare that the braced @var{code} must be invoked before parsing each time -@code{yyparse} is called. The @var{code} may use @code{$$} and -@code{@@$} --- initial value and location of the lookahead --- and the -@code{%parse-param}. +@code{yyparse} is called. The @var{code} may use @code{$$} (or +@code{$<@var{tag}>$}) and @code{@@$} --- initial value and location of the +lookahead --- and the @code{%parse-param}. @end deffn For instance, if your locations use a file name, you may use @@ -4674,11 +4674,11 @@ symbol is automatically discarded. @deffn {Directive} %destructor @{ @var{code} @} @var{symbols} @findex %destructor Invoke the braced @var{code} whenever the parser discards one of the -@var{symbols}. -Within @var{code}, @code{$$} designates the semantic value associated -with the discarded symbol, and @code{@@$} designates its location. -The additional parser parameters are also available (@pxref{Parser Function, , -The Parser Function @code{yyparse}}). +@var{symbols}. Within @var{code}, @code{$$} (or @code{$<@var{tag}>$}) +designates the semantic value associated with the discarded symbol, and +@code{@@$} designates its location. The additional parser parameters are +also available (@pxref{Parser Function, , The Parser Function +@code{yyparse}}). When a symbol is listed among @var{symbols}, its @code{%destructor} is called a per-symbol @code{%destructor}. @@ -4816,10 +4816,11 @@ Decl, , Freeing Discarded Symbols}). @c This is the same text as for %destructor. Invoke the braced @var{code} whenever the parser displays one of the @var{symbols}. Within @var{code}, @code{yyoutput} denotes the output stream -(a @code{FILE*} in C, and an @code{std::ostream&} in C++), -@code{$$} designates the semantic value associated with the symbol, and -@code{@@$} its location. The additional parser parameters are also -available (@pxref{Parser Function, , The Parser Function @code{yyparse}}). +(a @code{FILE*} in C, and an @code{std::ostream&} in C++), @code{$$} (or +@code{$<@var{tag}>$}) designates the semantic value associated with the +symbol, and @code{@@$} its location. The additional parser parameters are +also available (@pxref{Parser Function, , The Parser Function +@code{yyparse}}). The @var{symbols} are defined as for @code{%destructor} (@pxref{Destructor Decl, , Freeing Discarded Symbols}.): they can be per-type (e.g., @@ -12375,8 +12376,8 @@ London, Department of Computer Science, TR-00-12 (December 2000). @uref{http://www.cs.rhul.ac.uk/research/languages/publications/tomita_style_1.ps} @end table -@node Index -@unnumbered Index +@node Index of Terms +@unnumbered Index of Terms @printindex cp diff --git a/gnulib b/gnulib -Subproject 2f67aa79c6f13961eda685144130692dc9f7b95 +Subproject dbd914496c99c52220e5f5ba4121d6cb55fb3be diff --git a/src/muscle-tab.c b/src/muscle-tab.c index 47d9fb65..1168a9c3 100644 --- a/src/muscle-tab.c +++ b/src/muscle-tab.c @@ -177,10 +177,10 @@ static void muscle_syncline_grow (char const *key, location loc) { char *extension = NULL; - obstack_fgrow1 (&muscle_obstack, "]b4_syncline(%d, [[", loc.start.line); - MUSCLE_OBSTACK_SGROW (&muscle_obstack, - quotearg_style (c_quoting_style, loc.start.file)); - obstack_sgrow (&muscle_obstack, "]])["); + obstack_fgrow1 (&muscle_obstack, "]b4_syncline(%d, ", loc.start.line); + obstack_quote (&muscle_obstack, + quotearg_style (c_quoting_style, loc.start.file)); + obstack_sgrow (&muscle_obstack, ")["); obstack_1grow (&muscle_obstack, 0); extension = obstack_finish (&muscle_obstack); muscle_grow (key, extension, ""); @@ -205,11 +205,11 @@ void muscle_pair_list_grow (const char *muscle, const char *a1, const char *a2) { char *pair; - obstack_sgrow (&muscle_obstack, "[[["); - MUSCLE_OBSTACK_SGROW (&muscle_obstack, a1); - obstack_sgrow (&muscle_obstack, "]], [["); - MUSCLE_OBSTACK_SGROW (&muscle_obstack, a2); - obstack_sgrow (&muscle_obstack, "]]]"); + obstack_sgrow (&muscle_obstack, "["); + obstack_quote (&muscle_obstack, a1); + obstack_sgrow (&muscle_obstack, ", "); + obstack_quote (&muscle_obstack, a2); + obstack_sgrow (&muscle_obstack, "]"); obstack_1grow (&muscle_obstack, 0); pair = obstack_finish (&muscle_obstack); muscle_grow (muscle, pair, ",\n"); @@ -259,53 +259,61 @@ muscle_find (char const *key) } -void +/* In the format `file_name:line.column', append BOUND to MUSCLE. Use + digraphs for special characters in the file name. */ + +static void muscle_boundary_grow (char const *key, boundary bound) { char *extension; - MUSCLE_OBSTACK_SGROW (&muscle_obstack, bound.file); + obstack_sgrow (&muscle_obstack, "[["); + obstack_escape (&muscle_obstack, bound.file); obstack_1grow (&muscle_obstack, ':'); obstack_fgrow1 (&muscle_obstack, "%d", bound.line); obstack_1grow (&muscle_obstack, '.'); obstack_fgrow1 (&muscle_obstack, "%d", bound.column); + obstack_sgrow (&muscle_obstack, "]]"); obstack_1grow (&muscle_obstack, '\0'); extension = obstack_finish (&muscle_obstack); muscle_grow (key, extension, ""); obstack_free (&muscle_obstack, extension); } -void + +/* In the format `[[file_name:line.column]], [[file_name:line.column]]', + append LOC to MUSCLE. Use digraphs for special characters in each + file name. */ + +static void muscle_location_grow (char const *key, location loc) { - muscle_grow (key, "[[", ""); muscle_boundary_grow (key, loc.start); - muscle_grow (key, "]], [[", ""); + muscle_grow (key, "", ", "); muscle_boundary_grow (key, loc.end); - muscle_grow (key, "]]", ""); } -#define MUSCLE_COMMON_DECODE(Value) \ - case '$': \ - aver (*++(Value) == ']'); \ - aver (*++(Value) == '['); \ - obstack_sgrow (&muscle_obstack, "$"); \ - break; \ - case '@': \ - switch (*++(Value)) \ - { \ - case '@': obstack_sgrow (&muscle_obstack, "@" ); break; \ - case '{': obstack_sgrow (&muscle_obstack, "[" ); break; \ - case '}': obstack_sgrow (&muscle_obstack, "]" ); break; \ - default: aver (false); break; \ - } \ - break; \ - default: \ - obstack_1grow (&muscle_obstack, *(Value)); \ +#define COMMON_DECODE(Value) \ + case '$': \ + aver (*++(Value) == ']'); \ + aver (*++(Value) == '['); \ + obstack_sgrow (&muscle_obstack, "$"); \ + break; \ + case '@': \ + switch (*++(Value)) \ + { \ + case '@': obstack_sgrow (&muscle_obstack, "@" ); break; \ + case '{': obstack_sgrow (&muscle_obstack, "[" ); break; \ + case '}': obstack_sgrow (&muscle_obstack, "]" ); break; \ + default: aver (false); break; \ + } \ + break; \ + default: \ + obstack_1grow (&muscle_obstack, *(Value)); \ break; -/* Reverse of MUSCLE_OBSTACK_SGROW. */ +/* Reverse of obstack_escape. */ static char * -muscle_string_decode (char const *key) +string_decode (char const *key) { char const *value; char *value_decoded; @@ -317,7 +325,7 @@ muscle_string_decode (char const *key) do { switch (*value) { - MUSCLE_COMMON_DECODE (value) + COMMON_DECODE (value) case '[': case ']': aver (false); @@ -332,7 +340,7 @@ muscle_string_decode (char const *key) /* Reverse of muscle_location_grow. */ static location -muscle_location_decode (char const *key) +location_decode (char const *key) { location loc; char const *value = muscle_find_const (key); @@ -342,7 +350,7 @@ muscle_location_decode (char const *key) while (*++value) switch (*value) { - MUSCLE_COMMON_DECODE (value) + COMMON_DECODE (value) case '[': aver (false); break; @@ -491,7 +499,7 @@ muscle_percent_define_get (char const *variable) variable, ")"); muscle_insert (usage_name, ""); - value = muscle_string_decode (name); + value = string_decode (name); if (!value) value = xstrdup (""); return value; @@ -505,7 +513,7 @@ muscle_percent_define_get_loc (char const *variable) if (!muscle_find_const (loc_name)) complain (fatal, _("%s: undefined %%define variable %s"), "muscle_percent_define_get_loc", quote (variable)); - return muscle_location_decode (loc_name); + return location_decode (loc_name); } char const * @@ -610,7 +618,7 @@ muscle_percent_define_check_values (char const * const *values) name = UNIQSTR_CONCAT ("percent_define(", *variablep, ")"); - value = muscle_string_decode (name); + value = string_decode (name); if (value) { for (++values; *values; ++values) diff --git a/src/muscle-tab.h b/src/muscle-tab.h index 5f701ae4..1f33f8c1 100644 --- a/src/muscle-tab.h +++ b/src/muscle-tab.h @@ -35,63 +35,48 @@ void muscle_free (void); extern struct obstack muscle_obstack; #define MUSCLE_INSERT_BOOL(Key, Value) \ -do { \ - int v__ = Value; \ - MUSCLE_INSERT_INT (Key, v__); \ -} while(0) + do { \ + int v__ = Value; \ + MUSCLE_INSERT_INT (Key, v__); \ + } while (0) #define MUSCLE_INSERT_INT(Key, Value) \ -do { \ - obstack_fgrow1 (&muscle_obstack, "%d", Value); \ - obstack_1grow (&muscle_obstack, 0); \ - muscle_insert (Key, obstack_finish (&muscle_obstack)); \ -} while(0) + do { \ + obstack_fgrow1 (&muscle_obstack, "%d", Value); \ + obstack_1grow (&muscle_obstack, 0); \ + muscle_insert (Key, obstack_finish (&muscle_obstack)); \ + } while (0) #define MUSCLE_INSERT_LONG_INT(Key, Value) \ -do { \ - obstack_fgrow1 (&muscle_obstack, "%ld", Value); \ - obstack_1grow (&muscle_obstack, 0); \ - muscle_insert (Key, obstack_finish (&muscle_obstack)); \ -} while(0) + do { \ + obstack_fgrow1 (&muscle_obstack, "%ld", Value); \ + obstack_1grow (&muscle_obstack, 0); \ + muscle_insert (Key, obstack_finish (&muscle_obstack)); \ + } while (0) /* Key -> Value, but don't apply escaping to Value. */ #define MUSCLE_INSERT_STRING_RAW(Key, Value) \ -do { \ - obstack_sgrow (&muscle_obstack, Value); \ - obstack_1grow (&muscle_obstack, 0); \ - muscle_insert (Key, obstack_finish (&muscle_obstack)); \ -} while(0) + do { \ + obstack_sgrow (&muscle_obstack, Value); \ + obstack_1grow (&muscle_obstack, 0); \ + muscle_insert (Key, obstack_finish (&muscle_obstack)); \ + } while (0) /* Key -> Value, applying M4 escaping to Value. */ #define MUSCLE_INSERT_STRING(Key, Value) \ -do { \ - MUSCLE_OBSTACK_SGROW (&muscle_obstack, Value); \ - obstack_1grow (&muscle_obstack, 0); \ - muscle_insert (Key, obstack_finish (&muscle_obstack)); \ -} while(0) - -#define MUSCLE_OBSTACK_SGROW(Obstack, Value) \ -do { \ - char const *p__; \ - for (p__ = Value; *p__; p__++) \ - switch (*p__) \ - { \ - case '$': obstack_sgrow (Obstack, "$]["); break; \ - case '@': obstack_sgrow (Obstack, "@@" ); break; \ - case '[': obstack_sgrow (Obstack, "@{" ); break; \ - case ']': obstack_sgrow (Obstack, "@}" ); break; \ - default: obstack_1grow (Obstack, *p__); break; \ - } \ -} while(0) + do { \ + obstack_escape (&muscle_obstack, Value); \ + obstack_1grow (&muscle_obstack, 0); \ + muscle_insert (Key, obstack_finish (&muscle_obstack)); \ + } while (0) #define MUSCLE_INSERT_C_STRING(Key, Value) \ -do { \ - MUSCLE_OBSTACK_SGROW (&muscle_obstack, \ - quotearg_style (c_quoting_style, \ - Value)); \ - obstack_1grow (&muscle_obstack, 0); \ - muscle_insert (Key, obstack_finish (&muscle_obstack)); \ -} while(0) + do { \ + obstack_escape (&muscle_obstack, \ + quotearg_style (c_quoting_style, Value)); \ + obstack_1grow (&muscle_obstack, 0); \ + muscle_insert (Key, obstack_finish (&muscle_obstack)); \ + } while (0) /* Append VALUE to the current value of KEY. If KEY did not already exist, create it. Use MUSCLE_OBSTACK. De-allocate the previously @@ -113,14 +98,6 @@ void muscle_code_grow (const char *key, const char *value, location loc); void muscle_pair_list_grow (const char *muscle, const char *a1, const char *a2); -/* In the format `[[file_name:line.column]], [[file_name:line.column]]', append - LOC to MUSCLE. Use digraphs for special characters in each file name. */ -void muscle_location_grow (char const *key, location loc); - -/* In the format `file_name:line.column', append BOUND to MUSCLE. Use digraphs - for special characters in the file name. */ -void muscle_boundary_grow (char const *key, boundary bound); - /* Grow KEY for the occurrence of the name USER_NAME at LOC appropriately for use with b4_check_user_names in ../data/bison.m4. USER_NAME is not escaped with digraphs, so it must not contain `[' or `]'. */ diff --git a/src/output.c b/src/output.c index 26a6e7fe..c227be22 100644 --- a/src/output.c +++ b/src/output.c @@ -108,29 +108,39 @@ GENERATE_MUSCLE_INSERT_TABLE(muscle_insert_symbol_number_table, symbol_number) GENERATE_MUSCLE_INSERT_TABLE(muscle_insert_state_number_table, state_number) -/*--------------------------------------------------------------------. -| Print to OUT a representation of STRING escaped both for C and M4. | -`--------------------------------------------------------------------*/ +/*----------------------------------------------------------------. +| Print to OUT a representation of CP quoted and escaped for M4. | +`----------------------------------------------------------------*/ static void -escaped_output (FILE *out, char const *string) +quoted_output (FILE *out, char const *cp) { - char const *p; fprintf (out, "[["); - for (p = quotearg_style (c_quoting_style, string); *p; p++) - switch (*p) + for (; *cp; cp++) + switch (*cp) { case '$': fputs ("$][", out); break; case '@': fputs ("@@", out); break; case '[': fputs ("@{", out); break; case ']': fputs ("@}", out); break; - default: fputc (*p, out); break; + default: fputc (*cp, out); break; } fprintf (out, "]]"); } +/*----------------------------------------------------------------. +| Print to OUT a representation of STRING quoted and escaped both | +| for C and M4. | +`----------------------------------------------------------------*/ + +static void +string_output (FILE *out, char const *string) +{ + quoted_output (out, quotearg_style (c_quoting_style, string)); +} + /*------------------------------------------------------------------. | Prepare the muscles related to the symbols: translate, tname, and | @@ -174,7 +184,7 @@ prepare_symbols (void) if (i) obstack_1grow (&format_obstack, ' '); - MUSCLE_OBSTACK_SGROW (&format_obstack, cp); + obstack_escape (&format_obstack, cp); free (cp); obstack_1grow (&format_obstack, ','); j += width; @@ -355,7 +365,7 @@ user_actions_output (FILE *out) fprintf (out, "b4_%scase(%d, [b4_syncline(%d, ", rules[r].is_predicate ? "predicate_" : "", r + 1, rules[r].action_location.start.line); - escaped_output (out, rules[r].action_location.start.file); + string_output (out, rules[r].action_location.start.file); fprintf (out, ")\n[ %s]])\n\n", rules[r].action); } fputs ("])\n\n", out); diff --git a/src/parse-gram.y b/src/parse-gram.y index 9d9d1a42..51896994 100644 --- a/src/parse-gram.y +++ b/src/parse-gram.y @@ -389,10 +389,15 @@ grammar_declaration: } | code_props_type "{...}" generic_symlist { - symbol_list *list; - for (list = $3; list; list = list->next) - symbol_list_code_props_set (list, $1, @2, $2); - symbol_list_free ($3); + code_props code; + code_props_symbol_action_init (&code, $2, @2); + code_props_translate_code (&code); + { + symbol_list *list; + for (list = $3; list; list = list->next) + symbol_list_code_props_set (list, $1, &code); + symbol_list_free ($3); + } } | "%default-prec" { diff --git a/src/scan-code.h b/src/scan-code.h index 2dfb1704..348d9e3f 100644 --- a/src/scan-code.h +++ b/src/scan-code.h @@ -45,7 +45,10 @@ typedef struct code_props { CODE_PROPS_SYMBOL_ACTION, CODE_PROPS_RULE_ACTION } kind; - /** \c NULL iff \c code_props::kind is \c CODE_PROPS_NONE. */ + /** + * \c NULL iff \c code_props::kind is \c CODE_PROPS_NONE. + * Memory is allocated in an obstack freed elsewhere. + */ char const *code; /** Undefined iff \c code_props::code is \c NULL. */ location location; diff --git a/src/scan-code.l b/src/scan-code.l index f6efeffd..2fe880ed 100644 --- a/src/scan-code.l +++ b/src/scan-code.l @@ -47,6 +47,9 @@ YY_DECL; #define YY_USER_ACTION location_compute (loc, &loc->end, yytext, yyleng); +static char *fetch_type_name (char *cp, char const **type_name, + location dollar_loc); + static void handle_action_dollar (symbol_list *rule, char *cp, location dollar_loc); static void handle_action_at (symbol_list *rule, char *cp, location at_loc); @@ -96,11 +99,15 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" int braces_level = 0; /* Whether a semicolon is probably needed. - The heuristic is that a semicolon is not needed after '{', '}', ';', - or a C preprocessor directive, and that whitespaces and comments - do not affect this flag. - Note that '{' does not need a semicolon because of '{}'. - A semicolon may be needed before a cpp direcive, but don't bother. */ + + The heuristic is that a semicolon is not needed after '{', '}', + ';', or a C preprocessor directive, and that whitespaces and + comments do not affect this flag. Note that '{' does not need a + semicolon because of '{}'. A semicolon may be needed before a + cpp directive, but don't bother. + + While it is maintained in several start-conditions (factoring + opportunities), it is meaningful only for SC_RULE_ACTION. */ bool need_semicolon = false; /* Whether in a C preprocessor directive. Don't use a start condition @@ -158,7 +165,8 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" } -<SC_RULE_ACTION,SC_SYMBOL_ACTION>{ +<SC_RULE_ACTION,SC_SYMBOL_ACTION> +{ "'" { STRING_GROW; BEGIN SC_CHARACTER; @@ -177,42 +185,31 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" STRING_GROW; BEGIN SC_LINE_COMMENT; } + [$@] { + complain_at (*loc, Wother, _("stray '%s'"), yytext); + obstack_escape (&obstack_for_string, yytext); + need_semicolon = true; + } + [\[\]] { + obstack_escape (&obstack_for_string, yytext); + need_semicolon = true; + } } <SC_RULE_ACTION> { "$"("<"{tag}">")?{ref} { - ref_tail_fields = 0; + ref_tail_fields = NULL; handle_action_dollar (self->rule, yytext, *loc); - if (ref_tail_fields) { + if (ref_tail_fields) obstack_sgrow (&obstack_for_string, ref_tail_fields); - } need_semicolon = true; } "@"{ref} { - ref_tail_fields = 0; + ref_tail_fields = NULL; handle_action_at (self->rule, yytext, *loc); - if (ref_tail_fields) { + if (ref_tail_fields) obstack_sgrow (&obstack_for_string, ref_tail_fields); - } - need_semicolon = true; - } - "$" { - complain_at (*loc, Wother, _("stray '$'")); - obstack_sgrow (&obstack_for_string, "$]["); - need_semicolon = true; - } - "@" { - complain_at (*loc, Wother, _("stray '@'")); - obstack_sgrow (&obstack_for_string, "@@"); - need_semicolon = true; - } - "[" { - obstack_sgrow (&obstack_for_string, "@{"); - need_semicolon = true; - } - "]" { - obstack_sgrow (&obstack_for_string, "@}"); need_semicolon = true; } @@ -253,8 +250,12 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" <SC_SYMBOL_ACTION> { - "$$" { - obstack_sgrow (&obstack_for_string, "]b4_dollar_dollar["); + "$"("<"{tag}">")?"$" { + const char *type_name = NULL; + fetch_type_name (yytext + 1, &type_name, *loc)[-1] = 0; + obstack_sgrow (&obstack_for_string, "]b4_dollar_dollar("); + obstack_quote (&obstack_for_string, type_name); + obstack_sgrow (&obstack_for_string, ")["); self->is_value_used = true; } "@$" { @@ -264,29 +265,17 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" } - /*-----------------------------------------. - | Escape M4 quoting characters in C code. | - `-----------------------------------------*/ - <*> { - \$ obstack_sgrow (&obstack_for_string, "$]["); - \@ obstack_sgrow (&obstack_for_string, "@@"); - \[ obstack_sgrow (&obstack_for_string, "@{"); - \] obstack_sgrow (&obstack_for_string, "@}"); -} - - /*-----------------------------------------------------. - | By default, grow the string obstack with the input. | - `-----------------------------------------------------*/ + /* Escape M4 quoting characters in C code. */ + [$@\[\]] obstack_escape (&obstack_for_string, yytext); -<*>.|\n STRING_GROW; + /* By default, grow the string obstack with the input. */ + .|\n STRING_GROW; /* End of processing. */ -<*><<EOF>> { - STRING_FINISH; - return last_string; - } + <<EOF>> STRING_FINISH; return last_string; +} %% @@ -335,7 +324,7 @@ typedef struct not visible from current midrule. */ #define VARIANT_NOT_VISIBLE_FROM_MIDRULE (1 << 2) -static variant *variant_table = 0; +static variant *variant_table = NULL; static unsigned variant_table_size = 0; static unsigned variant_count = 0; @@ -357,7 +346,7 @@ static void variant_table_free (void) { free (variant_table); - variant_table = 0; + variant_table = NULL; variant_table_size = variant_count = 0; } @@ -681,6 +670,32 @@ parse_ref (char *cp, symbol_list *rule, int rule_length, int max_left_semantic_context = 0; +/* If CP points to a typename (i.e., <.*?>), set TYPE_NAME to its + beginning (i.e., after the opening "<", and return the pointer + immediately after it. */ + +static +char * +fetch_type_name (char *cp, char const **type_name, + location dollar_loc) +{ + if (*cp == '<') + { + *type_name = ++cp; + while (*cp != '>') + ++cp; + + /* The '>' symbol will be later replaced by '\0'. Original + 'text' is needed for error messages. */ + ++cp; + if (untyped_var_seen) + complain_at (dollar_loc, complaint, + _("explicit type given in untyped grammar")); + tag_seen = true; + } + return cp; +} + /*------------------------------------------------------------------. | TEXT is pointing to a wannabee semantic value (i.e., a '$'). | | | @@ -694,7 +709,6 @@ handle_action_dollar (symbol_list *rule, char *text, location dollar_loc) { char const *type_name = NULL; char *cp = text + 1; - char *gt_ptr = 0; symbol_list *effective_rule; int effective_rule_length; int n; @@ -711,27 +725,14 @@ handle_action_dollar (symbol_list *rule, char *text, location dollar_loc) } /* Get the type name if explicit. */ - if (*cp == '<') - { - type_name = ++cp; - while (*cp != '>') - ++cp; - - /* The '>' symbol will be later replaced by '\0'. Original - 'text' is needed for error messages. */ - gt_ptr = cp; - ++cp; - if (untyped_var_seen) - complain_at (dollar_loc, complaint, - _("explicit type given in untyped grammar")); - tag_seen = true; - } + cp = fetch_type_name (cp, &type_name, dollar_loc); n = parse_ref (cp, effective_rule, effective_rule_length, rule->midrule_parent_rhs_index, text, dollar_loc, '$'); - if (gt_ptr) - *gt_ptr = '\0'; + /* End type_name. */ + if (type_name) + cp[-1] = '\0'; switch (n) { @@ -759,11 +760,11 @@ handle_action_dollar (symbol_list *rule, char *text, location dollar_loc) } else untyped_var_seen = true; - type_name = ""; } - obstack_fgrow1 (&obstack_for_string, - "]b4_lhs_value([%s])[", type_name); + obstack_sgrow (&obstack_for_string, "]b4_lhs_value("); + obstack_quote (&obstack_for_string, type_name); + obstack_sgrow (&obstack_for_string, ")["); rule->action_props.is_value_used = true; break; @@ -781,12 +782,12 @@ handle_action_dollar (symbol_list *rule, char *text, location dollar_loc) quote (effective_rule->content.sym->tag)); else untyped_var_seen = true; - type_name = ""; } - obstack_fgrow3 (&obstack_for_string, - "]b4_rhs_value(%d, %d, [%s])[", - effective_rule_length, n, type_name); + obstack_fgrow2 (&obstack_for_string, + "]b4_rhs_value(%d, %d, ", effective_rule_length, n); + obstack_quote (&obstack_for_string, type_name); + obstack_sgrow (&obstack_for_string, ")["); if (n > 0) symbol_list_n_get (effective_rule, n)->action_props.is_value_used = true; diff --git a/src/scan-skel.l b/src/scan-skel.l index 17382a73..b79e5832 100644 --- a/src/scan-skel.l +++ b/src/scan-skel.l @@ -72,8 +72,8 @@ static void fail_for_invalid_at (char const *at); "@@" fputc ('@', yyout); "@{" fputc ('[', yyout); "@}" fputc (']', yyout); -"@`" /* Empty. Used by b4_cat in ../data/bison.m4. */ -@\n /* Likewise. */ +"@`" continue; /* Used by b4_cat in ../data/bison.m4. */ +@\n continue; "@oline@" fprintf (yyout, "%d", out_lineno + 1); "@ofile@" QPUTS (outname); @@ -91,7 +91,7 @@ static void fail_for_invalid_at (char const *at); \n out_lineno++; ECHO; [^@\n]+ ECHO; -<INITIAL><<EOF>> { +<<EOF>> { if (outname) { free (outname); @@ -100,15 +100,15 @@ static void fail_for_invalid_at (char const *at); return EOF; } -<SC_AT_DIRECTIVE_ARGS>{ - [^@]+ { STRING_GROW; } +<SC_AT_DIRECTIVE_ARGS> +{ + [^@]+ STRING_GROW; - "@@" { obstack_1grow (&obstack_for_string, '@'); } - "@{" { obstack_1grow (&obstack_for_string, '['); } - "@}" { obstack_1grow (&obstack_for_string, ']'); } - "@`" /* Empty. Useful for starting an argument - that begins with whitespace. */ - @\n /* Empty. */ + "@@" obstack_1grow (&obstack_for_string, '@'); + "@{" obstack_1grow (&obstack_for_string, '['); + "@}" obstack_1grow (&obstack_for_string, ']'); + "@`" continue; /* For starting an argument that begins with whitespace. */ + @\n continue; @[,)] { if (at_directive_argc >= AT_DIRECTIVE_ARGC_MAX) @@ -131,15 +131,17 @@ static void fail_for_invalid_at (char const *at); } } - @.? { fail_for_invalid_at (yytext); } + @.? fail_for_invalid_at (yytext); } -<SC_AT_DIRECTIVE_SKIP_WS>{ - [ \t\r\n] +<SC_AT_DIRECTIVE_SKIP_WS> +{ + [ \t\r\n] continue; . { yyless (0); BEGIN SC_AT_DIRECTIVE_ARGS; } } -<SC_AT_DIRECTIVE_ARGS,SC_AT_DIRECTIVE_SKIP_WS>{ +<SC_AT_DIRECTIVE_ARGS,SC_AT_DIRECTIVE_SKIP_WS> +{ <<EOF>> { complain (fatal, _("unclosed %s directive in skeleton"), at_directive_argv[0]); diff --git a/src/symlist.c b/src/symlist.c index bd9f3c23..2bbedb58 100644 --- a/src/symlist.c +++ b/src/symlist.c @@ -191,15 +191,12 @@ symbol_list_null (symbol_list *node) void symbol_list_code_props_set (symbol_list *node, code_props_type kind, - location loc, char const *code) + code_props const *cprops) { - code_props cprops; - code_props_symbol_action_init (&cprops, code, loc); - code_props_translate_code (&cprops); switch (node->content_type) { case SYMLIST_SYMBOL: - symbol_code_props_set (node->content.sym, kind, &cprops); + symbol_code_props_set (node->content.sym, kind, cprops); if (node->content.sym->status == undeclared) node->content.sym->status = used; break; @@ -207,7 +204,7 @@ symbol_list_code_props_set (symbol_list *node, code_props_type kind, semantic_type_code_props_set (semantic_type_get (node->content.sem_type->tag, &node->content.sem_type->location), - kind, &cprops); + kind, cprops); if (node->content.sem_type->status == undeclared) node->content.sem_type->status = used; break; diff --git a/src/symlist.h b/src/symlist.h index dd634083..03ffebef 100644 --- a/src/symlist.h +++ b/src/symlist.h @@ -113,9 +113,8 @@ uniqstr symbol_list_n_type_name_get (symbol_list *l, location loc, int n); /* Check whether the node is a border element of a rule. */ bool symbol_list_null (symbol_list *node); -/** Set the \c \%destructor or \c \%printer for \c node as \c code at - \c loc. */ +/** Set the \c \%destructor or \c \%printer for \c node as \c cprops. */ void symbol_list_code_props_set (symbol_list *node, code_props_type kind, - location loc, char const *code); + code_props const *cprops); #endif /* !SYMLIST_H_ */ diff --git a/src/symtab.h b/src/symtab.h index 93778008..49eebd4c 100644 --- a/src/symtab.h +++ b/src/symtab.h @@ -92,8 +92,14 @@ struct symbol /** The location of its first occurrence. */ location location; - /** Its \c \%type. */ + /** Its \c \%type. + + Beware that this is the type_name as was entered by the user, + including silly things such as "]" if she entered "%token <]> t". + Therefore, when outputting type_name to M4, be sure to escape it + into "@}". See quoted_output for instance. */ uniqstr type_name; + /** Its \c \%type's location. */ location type_location; @@ -101,9 +107,9 @@ struct symbol symbol. Access this field only through <tt>symbol</tt>'s interface functions. For - Example, if <tt>symbol::destructor = NULL</tt> (resp. <tt>symbol::printer + example, if <tt>symbol::destructor = NULL</tt> (resp. <tt>symbol::printer = NULL</tt>), a default \c \%destructor (resp. \%printer) or a per-type - \c symbol_destructor_printer_get will compute the corect one. */ + \c symbol_destructor_printer_get will compute the correct one. */ code_props props[CODE_PROPS_SIZE]; symbol_number number; diff --git a/src/system.h b/src/system.h index 5613a92f..1d0af8fa 100644 --- a/src/system.h +++ b/src/system.h @@ -177,32 +177,75 @@ typedef size_t uintptr_t; obstack_grow (Obs, Str, strlen (Str)) #define obstack_fgrow1(Obs, Format, Arg1) \ -do { \ - char buf[4096]; \ - sprintf (buf, Format, Arg1); \ - obstack_grow (Obs, buf, strlen (buf)); \ -} while (0) + do { \ + char buf[4096]; \ + sprintf (buf, Format, Arg1); \ + obstack_grow (Obs, buf, strlen (buf)); \ + } while (0) #define obstack_fgrow2(Obs, Format, Arg1, Arg2) \ -do { \ - char buf[4096]; \ - sprintf (buf, Format, Arg1, Arg2); \ - obstack_grow (Obs, buf, strlen (buf)); \ -} while (0) + do { \ + char buf[4096]; \ + sprintf (buf, Format, Arg1, Arg2); \ + obstack_grow (Obs, buf, strlen (buf)); \ + } while (0) #define obstack_fgrow3(Obs, Format, Arg1, Arg2, Arg3) \ -do { \ - char buf[4096]; \ - sprintf (buf, Format, Arg1, Arg2, Arg3); \ - obstack_grow (Obs, buf, strlen (buf)); \ -} while (0) + do { \ + char buf[4096]; \ + sprintf (buf, Format, Arg1, Arg2, Arg3); \ + obstack_grow (Obs, buf, strlen (buf)); \ + } while (0) #define obstack_fgrow4(Obs, Format, Arg1, Arg2, Arg3, Arg4) \ -do { \ - char buf[4096]; \ - sprintf (buf, Format, Arg1, Arg2, Arg3, Arg4); \ - obstack_grow (Obs, buf, strlen (buf)); \ -} while (0) + do { \ + char buf[4096]; \ + sprintf (buf, Format, Arg1, Arg2, Arg3, Arg4); \ + obstack_grow (Obs, buf, strlen (buf)); \ + } while (0) + + +/* Output Str escaped for our postprocessing (i.e., escape M4 special + characters). + + For instance "[foo]" -> "@{foo@}", "$$" -> "$][$][". */ + +# define obstack_escape(Obs, Str) \ + do { \ + char const *p__; \ + for (p__ = Str; *p__; p__++) \ + switch (*p__) \ + { \ + case '$': obstack_sgrow (Obs, "$]["); break; \ + case '@': obstack_sgrow (Obs, "@@" ); break; \ + case '[': obstack_sgrow (Obs, "@{" ); break; \ + case ']': obstack_sgrow (Obs, "@}" ); break; \ + default: obstack_1grow (Obs, *p__ ); break; \ + } \ + } while (0) + + +/* Output Str both quoted for M4 (i.e., embed in [[...]]), and escaped + for our postprocessing (i.e., escape M4 special characters). If + Str is empty (or NULL), output "[]" instead of "[[]]" as it make M4 + programming easier (m4_ifval can be used). + + For instance "[foo]" -> "[[@{foo@}]]", "$$" -> "[[$][$][]]". */ + +# define obstack_quote(Obs, Str) \ + do { \ + char const* obstack_quote_p = Str; \ + if (obstack_quote_p && obstack_quote_p[0]) \ + { \ + obstack_sgrow (Obs, "[["); \ + obstack_escape (Obs, obstack_quote_p); \ + obstack_sgrow (Obs, "]]"); \ + } \ + else \ + obstack_sgrow (Obs, "[]"); \ + } while (0) + + @@ -218,10 +261,6 @@ do { \ # define TAB_EXT ".tab" #endif -#ifndef DEFAULT_TMPDIR -# define DEFAULT_TMPDIR "/tmp" -#endif - /*---------------------. diff --git a/tests/actions.at b/tests/actions.at index 120c4012..bde0961b 100644 --- a/tests/actions.at +++ b/tests/actions.at @@ -1255,6 +1255,143 @@ AT_CHECK_ACTION_LOCATIONS([[%destructor]]) AT_CHECK_ACTION_LOCATIONS([[%printer]]) +## ------------------------- ## +## Qualified $$ in actions. ## +## ------------------------- ## + +# Check that we can used qualified $$ (v.g., $<type>$) not only in +# rule actions, but also where $$ is valid: %printer and %destructor. +# +# FIXME: Not actually checking %desctructor, but it's the same code as +# %printer... +# +# To do that, use a semantic value that has two fields (sem_type), +# declare symbols to have only one of these types (INT, float), and +# use $<type>$ to get the other one. Including for symbols that are +# not typed (UNTYPED). + +m4_pushdef([AT_TEST], +[AT_SETUP([[Qualified $$ in actions: $1]]) + +AT_BISON_OPTION_PUSHDEFS([%locations %skeleton "$1"]) + +AT_DATA_GRAMMAR([[input.y]], +[[%skeleton "$1" +%defines // FIXME: Mandated by lalr1.cc in Bison 2.6. +%locations // FIXME: Mandated by lalr1.cc in Bison 2.6. +%debug +%code requires +{ +# include <stdio.h> + + typedef struct sem_type + { + int ival; + float fval; + } sem_type; + +# define YYSTYPE sem_type + +#ifdef __cplusplus +# include <iostream> + static void + report (std::ostream& yyo, int ival, float fval) + { + yyo << "ival: " << ival << ", fval: " << fval; + } +#else + static void + report (FILE* yyo, int ival, float fval) + { + fprintf (yyo, "ival: %d, fval: %1.1f", ival, fval); + } +#endif +} + +%code +{ + ]AT_YYERROR_DECLARE[ + ]AT_YYLEX_DECLARE[ +} + +%token UNTYPED +%token <ival> INT +%type <fval> float +%printer { report (yyo, $$, $<fval>$); } <ival>; +%printer { report (yyo, $<ival>$, $$ ); } <fval>; +%printer { report (yyo, $<ival>$, $<fval>$); } <>; + +]AT_SKEL_CC_IF([[ +/* The lalr1.cc skeleton, for backward compatibility, defines + a constructor for position that initializes the filename. The + glr.cc skeleton does not (and in fact cannot: location/position + are stored in a union, from which objects with constructors are + excluded in C++). */ +%initial-action { + @$.initialize (); +} +]])[ + +%initial-action +{ + $<ival>$ = 42; + $<fval>$ = 4.2; +} + +%% +float: UNTYPED INT +{ + $$ = $<fval>1 + $<fval>2; + $<ival>$ = $<ival>1 + $][2; +}; +%% +]AT_YYERROR_DEFINE[ +]AT_YYLEX_DEFINE(AT_SKEL_CC_IF([[{yy::parser::token::UNTYPED, + yy::parser::token::INT, + EOF}]], + [[{UNTYPED, INT, EOF}]]), + [AT_VAL.ival = toknum * 10; AT_VAL.fval = toknum / 10.0;])[ +int +main (void) +{]AT_SKEL_CC_IF([[ + yy::parser p; + p.set_debug_level(1); + return p.parse ();]], [[ + yydebug = 1; + return yyparse ();]])[ +} +]]) + +AT_FULL_COMPILE([[input]]) +AT_PARSER_CHECK([./input], 0, [], [stderr]) +# Don't be too picky on the traces, GLR is not exactly the same. Keep +# only the lines from the printer. +# +# Don't care about locations. FIXME: remove their removal when Bison +# supports C++ without locations. +AT_CHECK([[sed -ne 's/([-0-9.]*: /(/;/ival:/p' stderr]], 0, +[[Reading a token: Next token is token UNTYPED (ival: 10, fval: 0.1) +Shifting token UNTYPED (ival: 10, fval: 0.1) +Reading a token: Next token is token INT (ival: 20, fval: 0.2) +Shifting token INT (ival: 20, fval: 0.2) + $][1 = token UNTYPED (ival: 10, fval: 0.1) + $][2 = token INT (ival: 20, fval: 0.2) +-> $$ = nterm float (ival: 30, fval: 0.3) +Cleanup: popping nterm float (ival: 30, fval: 0.3) +]]) + +AT_BISON_OPTION_POPDEFS + +AT_CLEANUP +]) + +AT_TEST([yacc.c]) +AT_TEST([glr.c]) +AT_TEST([lalr1.cc]) +AT_TEST([glr.cc]) + +m4_popdef([AT_TEST]) + ## ----------------------------------------------- ## ## Fix user actions without a trailing semicolon. ## ## ----------------------------------------------- ## diff --git a/tests/input.at b/tests/input.at index 71505cf8..d6f2084a 100644 --- a/tests/input.at +++ b/tests/input.at @@ -1491,3 +1491,102 @@ AT_TEST([%define api.prefix foo], [-p bar], [input.y:1.9-18]) m4_popdef([AT_TEST]) AT_CLEANUP + + +## -------------- ## +## Stray $ or @. ## +## -------------- ## + +AT_SETUP([[Stray $ or @]]) + +# Give %printer and %destructor "<*> exp TOK" instead of "<*>" to +# check that the warnings are reported once, not three times. + +AT_DATA_GRAMMAR([[input.y]], +[[%type <TYPE> exp +%token <TYPE> TOK TOK2 +%destructor { $%; @%; } <*> exp TOK; +%initial-action { $%; @%; }; +%printer { $%; @%; } <*> exp TOK; +%% +exp: TOK { $%; @%; $$ = $1; }; +]]) + +AT_BISON_CHECK([[input.y]], 0, [], +[[input.y:11.19: warning: stray '$' [-Wother] +input.y:11.23: warning: stray '@' [-Wother] +input.y:12.19: warning: stray '$' [-Wother] +input.y:12.23: warning: stray '@' [-Wother] +input.y:13.19: warning: stray '$' [-Wother] +input.y:13.23: warning: stray '@' [-Wother] +input.y:15.19: warning: stray '$' [-Wother] +input.y:15.23: warning: stray '@' [-Wother] +]]) + +AT_CLEANUP + + + +## ---------------- ## +## Code injection. ## +## ---------------- ## + + +AT_SETUP([[Code injection]]) + +m4_pattern_allow([^m4_errprintn$]) + +# AT_TEST([MACRO]) +# ---------------- +# Try to have MACRO be run by bison. +m4_pushdef([AT_TEST], +[AT_DATA([[input.y]], +[[%type <$1(DEAD %type)> exp +%token <$1(DEAD %token)> a +%token b +%initial-action +{ + $$; + $<$1(DEAD %initial-action)>$ +}; +%printer +{ + $$ + $<$1(DEAD %printer)>$ +} <> <*>; +%lex-param +{ + $1(DEAD %lex-param) +}; +%parse-param +{ + $1(DEAD %parse-param) +}; +%% +exp: + a a[name] b + { + $$; + $][1; + $<$1(DEAD action 1)>$ + $<$1(DEAD action 2)>1 + $<$1(DEAD action 3)>name + $<$1(DEAD action 4)>0 + ; + }; +]]) + +# FIXME: Provide a means to iterate over all the skeletons. +AT_BISON_CHECK([[-d input.y]]) +AT_BISON_CHECK([[-d -S glr.c input.y]]) +AT_BISON_CHECK([[-d -S lalr1.cc input.y]]) +AT_BISON_CHECK([[-d -S glr.cc input.y]]) +AT_BISON_CHECK([[ -S lalr1.java input.y]]) +]) + +AT_TEST([m4_errprintn]) +AT_TEST([@:>@m4_errprintn]) + +m4_popdef([AT_TEST]) + +AT_CLEANUP diff --git a/tests/local.at b/tests/local.at index c79fc08e..4499e13f 100644 --- a/tests/local.at +++ b/tests/local.at @@ -338,7 +338,7 @@ static static size_t toknum = 0; int res; ]AT_USE_LEX_ARGS[; - assert (toknum < sizeof input); + assert (toknum < sizeof input / sizeof input[0]); res = input[toknum++]; ]$2[;]AT_LOCATION_IF([[ ]AT_LOC_FIRST_LINE[ = ]AT_LOC_LAST_LINE[ = 1; diff --git a/tests/synclines.at b/tests/synclines.at index 8d742d53..7e7122f9 100644 --- a/tests/synclines.at +++ b/tests/synclines.at @@ -27,6 +27,11 @@ AT_BANNER([[User Actions.]]) # errors. m4_define([AT_SYNCLINES_COMPILE], [AT_CHECK([$CC $CFLAGS $CPPFLAGS -c $1], [ignore], [], [stderr]) + +# Transform stderr into something like this: +# +# input.y:4: #error "4" +# # In case GCC displays column information, strip it down. # # input.y:4:2: #error "4" or @@ -42,6 +47,11 @@ m4_define([AT_SYNCLINES_COMPILE], # => # input.y:4: #error "8" # +# The message may include a caret-error: +# +# input.y:1:2: error: #error "1" +# #error "1" +# ^ # # And possibly distcc adds its bits. # @@ -54,12 +64,15 @@ m4_define([AT_SYNCLINES_COMPILE], # distcc[35882] (dcc_connect_by_name) ERROR: failed to look up host "chrisimac": Unknown host # distcc[35882] Warning: failed to distribute input.c to chrisimac/4, running locally instead -AT_CHECK([[sed -e '/^distcc\[[0-9]*\] /d' \ - -e 's/^\([^:]*:[^:.]*\)[.:][^:]*:\(.*\)$/\1:\2/' \ - -e 's/^\([^:]*:[^:]*:\)[^@%:@]*\( @%:@error\)/\1\2/' \ - -e "/^[^:]*: In function '[^\']*':$/d" \ - stderr]], - 0, [stdout]) +AT_CHECK([[perl -p -0777 -f - stderr <<\EOF + s/^distcc\[\d+\] .*\n//gm; + s/^([^:]*:\d+)[.:][^:]*:(.*)$/$][1:$][2/gm; + s/^([^:]*:\d+:)[^#]*( #error)/$][1$][2/gm; + s/^[^:]*: In function '[^']*':\n//gm; + s/^\ +#error.*\n\ *\^\n//gm; +EOF +]], + 0, [stdout]) ]) # AT_TEST_SYNCLINE(TITLE, INPUT, ERROR-MSG) |