Remove a dead item.

[bison.git] / TODO

diff --git a/TODO b/TODO
index a46c66f13bca15a412d24f4b205b8d6a86f7a7e7..434053356856d2c91172cf5695b273c7f16c0e89 100644 (file)
--- a/TODO
+++ b/TODO
@@ -1,15 +1,256 @@
 -*- outline -*-
 
 -*- outline -*-
 

-* src/reader.c
-Check and cleanup for CPP-out code.    [ok]

 
 

-* src/output.c
-Check and cleanup for CPP-out code.    [ok]
-** %%input-line is not correct at this time.
-Rely with bison.simple.
+* URGENT: Prologue
+The %union is declared after the user C declarations. It can be
+a problem if YYSTYPE is declared after the user part.

 
 

-* Broken options ?.
-** %no-lines           []
+Actually, the real problem seems that the %union ought to be output
+where it was defined.  For instance, in gettext/intl/plural.y, we
+have:
+
+       %{
+       ...
+       #include "gettextP.h"
+       ...
+       %}
+
+       %union {
+         unsigned long int num;
+         enum operator op;
+         struct expression *exp;
+       }
+
+       %{
+       ...
+       static int yylex PARAMS ((YYSTYPE *lval, const char **pexp));
+       ...
+       %}
+
+Where the first part defines struct expression, the second uses it to
+define YYSTYPE, and the last uses YYSTYPE.  Only this order is valid.
+
+Note that we have the same problem with GCC.
+
+I suggest splitting the prologue into pre-prologue and post-prologue.
+The reason is that:
+
+1. we keep language independance as it is the skeleton that joins the
+two prologues (there is no need for the engine to encode union yystype
+and to output it inside the prologue, which breaks the language
+independance of the generator)
+
+2. that makes it possible to have several %union in input.  I think
+this is a pleasant (but useless currently) feature, but in the future,
+I want a means to %include other bits of grammars, and _then_ it will
+be important for the various bits to define their needs in %union.
+
+When implementing multiple-%union support, bare the following in mind:
+
+- when --yacc, this must be flagged as an error.  Don't make it fatal
+  though.
+
+- The #line must now appear *inside* the definition of yystype.
+  Something like
+
+       {
+       #line 12 "foo.y"
+         int ival;
+       #line 23 "foo.y"
+         char *sval;
+       }
+
+* Language independent actions
+
+Currently bison, the generator, transforms $1, $$ and so forth into
+direct C code, manipulating the stacks.  This is problematic, because
+(i) it means that if we want more languages, we need to update the
+generator, and (ii), it forces names everywhere (e.g., the C++
+skeleton would be happy to use other naming schemes, and actually,
+even other accessing schemes).
+
+Therefore we want
+
+1. the generator to replace $1, etc. by M4 macro invocations
+   (b4_dollar(1), b4_at(3), b4_dollar_dollar) etc.
+
+2. the skeletons to define these macros.
+
+But currently the actions are double-quoted, to protect them from M4
+evaluation.  So we need to:
+
+3. stop quoting them
+
+4. change the [ and ] in the actions into @<:@ and @:>@
+
+5. extend the postprocessor to maps these back onto [ and ].
+
+* Coding system independence
+Paul notes:
+
+       Currently Bison assumes 8-bit bytes (i.e. that UCHAR_MAX is
+       255).  It also assumes that the 8-bit character encoding is
+       the same for the invocation of 'bison' as it is for the
+       invocation of 'cc', but this is not necessarily true when
+       people run bison on an ASCII host and then use cc on an EBCDIC
+       host.  I don't think these topics are worth our time
+       addressing (unless we find a gung-ho volunteer for EBCDIC or
+       PDP-10 ports :-) but they should probably be documented
+       somewhere.
+
+* Using enums instead of int for tokens.
+Paul suggests:
+
+   #ifndef YYTOKENTYPE
+   # if defined (__STDC__) || defined (__cplusplus)
+      /* Put the tokens into the symbol table, so that GDB and other debuggers
+         know about them.  */
+      enum yytokentype {
+        FOO = 256,
+        BAR,
+        ...
+      };
+      /* POSIX requires `int' for tokens in interfaces.  */
+   #  define YYTOKENTYPE int
+   # endif
+   #endif
+   #define FOO 256
+   #define BAR 257
+   ...
+
+* Output directory
+Akim:
+
+| I consider this to be a bug in bison:
+|
+| /tmp % mkdir src
+| /tmp % cp ~/src/bison/tests/calc.y src
+| /tmp % mkdir build && cd build
+| /tmp/build % bison ../src/calc.y
+| /tmp/build % cd ..
+| /tmp % ls -l build src
+| build:
+| total 0
+|
+| src:
+| total 32
+| -rw-r--r--    1 akim     lrde        27553 oct  2 16:31 calc.tab.c
+| -rw-r--r--    1 akim     lrde         3335 oct  2 16:31 calc.y
+|
+|
+| Would it be safe to change this behavior to something more reasonable?
+| Do you think some people depend upon this?
+
+Jim:
+
+Is it that behavior documented?
+If so, then it's probably not reasonable to change it.
+I've Cc'd the automake list, because some of automake's
+rules use bison through $(YACC) -- though I'll bet they
+all use it in yacc-compatible mode.
+
+Pavel:
+
+Hello, Jim and others!
+
+> Is it that behavior documented?
+> If so, then it's probably not reasonable to change it.
+> I've Cc'd the automake list, because some of automake's
+> rules use bison through $(YACC) -- though I'll bet they
+> all use it in yacc-compatible mode.
+
+Yes, Automake currently used bison in Automake-compatible mode, but it
+would be fair for Automake to switch to the native mode as long as the
+processed files are distributed and "missing" emulates bison.
+
+In any case, the makefiles should specify the output file explicitly
+instead of relying on weird defaults.
+
+> | src:
+> | total 32
+> | -rw-r--r--    1 akim     lrde        27553 oct  2 16:31 calc.tab.c
+> | -rw-r--r--    1 akim     lrde         3335 oct  2 16:31 calc.y
+
+This is not _that_ ugly as it seems - with Automake you want to put
+sources where they belong - to the source directory.
+
+> | This is not _that_ ugly as it seems - with Automake you want to put
+> | sources where they belong - to the source directory.
+>
+> The difference source/build you are referring to is based on Automake
+> concepts.  They have no sense at all for tools such as bison or gcc
+> etc.  They have input and output.  I do not want them to try to grasp
+> source/build.  I want them to behave uniformly: output *here*.
+
+I realize that.
+
+It's unfortunate that the native mode of Bison behaves in a less uniform
+way than the yacc mode. I agree with your point. Bison maintainters may
+want to fix it along with the documentation.
+
+
+* Unit rules
+Maybe we could expand unit rules, i.e., transform
+
+       exp: arith | bool;
+       arith: exp '+' exp;
+       bool: exp '&' exp;
+
+into
+
+       exp: exp '+' exp | exp '&' exp;
+
+when there are no actions.  This can significantly speed up some
+grammars.
+
+* Stupid error messages
+An example shows it easily:
+
+src/bison/tests % ./testsuite -k calc,location,error-verbose -l
+GNU Bison 1.49a test suite test groups:
+
+ NUM: FILENAME:LINE      TEST-GROUP-NAME
+      KEYWORDS
+
+  51: calc.at:440        Calculator --locations --yyerror-verbose
+  52: calc.at:442        Calculator --defines --locations --name-prefix=calc --verbose --yacc --yyerror-verbose
+  54: calc.at:445        Calculator --debug --defines --locations --name-prefix=calc --verbose --yacc --yyerror-verbose
+src/bison/tests % ./testsuite 51 -d
+## --------------------------- ##
+## GNU Bison 1.49a test suite. ##
+## --------------------------- ##
+ 51: calc.at:440       ok
+## ---------------------------- ##
+## All 1 tests were successful. ##
+## ---------------------------- ##
+src/bison/tests % cd ./testsuite.dir/51
+tests/testsuite.dir/51 % echo "()" | ./calc
+1.2-1.3: parse error, unexpected ')', expecting error or "number" or '-' or '('
+
+* yyerror, yyprint interface
+It should be improved, in particular when using Bison features such as
+locations, and YYPARSE_PARAMS.  For the time being, it is recommended
+to #define yyerror and yyprint to steal internal variables...
+
+* read_pipe.c
+This is not portable to DOS for instance.  Implement a more portable
+scheme.  Sources of inspiration include GNU diff, and Free Recode.
+
+* Memory leaks in the generator
+A round of memory leak clean ups would be most welcome.  Dmalloc,
+Checker GCC, Electric Fence, or Valgrind: you chose your tool.
+
+* Memory leaks in the parser
+The same applies to the generated parsers.  In particular, this is
+critical for user data: when aborting a parsing, when handling the
+error token etc., we often throw away yylval without giving a chance
+of cleaning it up to the user.
+
+* --graph
+Show reductions.       []
+
+* Broken options ?
+** %no-lines           [ok]

 ** %no-parser          []
 ** %pure-parser                []
 ** %semantic-parser    []
 ** %no-parser          []
 ** %pure-parser                []
 ** %semantic-parser    []


@@ -22,19 +263,214 @@ Maybe transfered in lex.c.
 *** %name-prefix       []
 
 ** Skeleton strategy.  []
 *** %name-prefix       []
 
 ** Skeleton strategy.  []

-Useful for %no_parser. []
+Must we keep %no-parser?
+            %token-table?

 *** New skeletons.     []
 
 *** New skeletons.     []
 

-* src/macrotab.[ch]
-Removing warnings when compiling. (gcc-warnings).      [ok]
-

 * src/print_graph.c
 Find the best graph parameters.        []
 
 * doc/bison.texinfo
 * src/print_graph.c
 Find the best graph parameters.        []
 
 * doc/bison.texinfo

-Update informations about ERROR_VERBOSE.       []
-Add explainations about skeleton muscles.      []
-Add explainations about %skeleton.     []
+** Update
+informations about ERROR_VERBOSE.      []
+** Add explainations about
+skeleton muscles.      []
+%skeleton.             []
+
+* testsuite
+** tests/pure-parser.at        []
+New tests.
+
+* Debugging parsers
+
+From Greg McGary:
+
+akim demaille <akim.demaille@epita.fr> writes:
+
+> With great pleasure!  Nonetheless, things which are debatable
+> (or not, but just `big') should be discuss in `public': something
+> like help- or bug-bison@gnu.org is just fine.  Jesse and I are there,
+> but there is also Jim and some other people.
+
+I have no idea whether it qualifies as big or controversial, so I'll
+just summarize for you.  I proposed this change years ago and was
+surprised that it was met with utter indifference!
+
+This debug feature is for the programs/grammars one develops with
+bison, not for debugging bison itself.  I find that the YYDEBUG
+output comes in a very inconvenient format for my purposes.
+When debugging gcc, for instance, what I want is to see a trace of
+the sequence of reductions and the line#s for the semantic actions
+so I can follow what's happening.  Single-step in gdb doesn't cut it
+because to move from one semantic action to the next takes you through
+lots of internal machinery of the parser, which is uninteresting.
+
+The change I made was to the format of the debug output, so that it
+comes out in the format of C error messages, digestible by emacs
+compile mode, like so:
+
+grammar.y:1234: foo: bar(0x123456) baz(0x345678)
+
+where "foo: bar baz" is the reduction rule, whose semantic action
+appears on line 1234 of the bison grammar file grammar.y.  The hex
+numbers on the rhs tokens are the parse-stack values associated with
+those tokens.  Of course, yytype might be something totally
+incompatible with that representation, but for the most part, yytype
+values are single words (scalars or pointers).  In the case of gcc,
+they're most often pointers to tree nodes.  Come to think of it, the
+right thing to do is to make the printing of stack values be
+user-definable.  It would also be useful to include the filename &
+line# of the file being parsed, but the main filename & line# should
+continue to be that of grammar.y
+
+Anyway, this feature has saved my life on numerous occasions.  The way
+I customarily use it is to first run bison with the traces on, isolate
+the sequence of reductions that interests me, put those traces in a
+buffer and force it into compile-mode, then visit each of those lines
+in the grammar and set breakpoints with C-x SPACE.  Then, I can run
+again under the control of gdb and stop at each semantic action.
+With the hex addresses of tree nodes, I can inspect the values
+associated with any rhs token.
+
+You like?
+
+* input synclines
+Some users create their foo.y files, and equip them with #line.  Bison
+should recognize these, and preserve them.
+
+* BTYacc
+See if we can integrate backtracking in Bison.  Contact the BTYacc
+maintainers.
+
+* Automaton report
+Display more clearly the lookaheads for each item.
+
+* RR conflicts
+See if we can use precedence between rules to solve RR conflicts.  See
+what POSIX says.
+
+* Precedence
+It is unfortunate that there is a total order for precedence.  It
+makes it impossible to have modular precedence information.  We should
+move to partial orders.
+
+* Parsing grammars
+Rewrite the reader in Bison.
+
+* Problems with aliases
+From: "Baum, Nathan I" <s0009525@chelt.ac.uk>
+Subject: Token Alias Bug
+To: "'bug-bison@gnu.org'" <bug-bison@gnu.org>
+
+I've noticed a bug in bison. Sadly, our eternally wise sysadmins won't let
+us use CVS, so I can't find out if it's been fixed already...
+
+Basically, I made a program (in flex) that went through a .y file looking
+for "..."-tokens, and then outputed a %token
+line for it. For single-character ""-tokens, I reasoned, I could just use
+[%token 'A' "A"]. However, this causes Bison to output a [#define 'A' 65],
+which cppp chokes on, not unreasonably. (And even if cppp didn't choke, I
+obviously wouldn't want (char)'A' to be replaced with (int)65 throughout my
+code.
+
+Bison normally forgoes outputing a #define for a character token. However,
+it always outputs an aliased token -- even if the token is an alias for a
+character token. We don't want that. The problem is in /output.c/, as I
+recall. When it outputs the token definitions, it checks for a character
+token, and then checks for an alias token. If the character token check is
+placed after the alias check, then it works correctly.
+
+Alias tokens seem to be something of a kludge. What about an [%alias "..."]
+command...
+
+       %alias T_IF "IF"
+
+Hmm. I can't help thinking... What about a --generate-lex option that
+creates an .l file for the alias tokens used... (Or an option to make a
+gperf file, etc...)
+
+* Presentation of the report file
+From: "Baum, Nathan I" <s0009525@chelt.ac.uk>
+Subject: Token Alias Bug
+To: "'bug-bison@gnu.org'" <bug-bison@gnu.org>
+
+I've also noticed something, that whilst not *wrong*, is inconvienient: I
+use the verbose mode to help find the causes of unresolved shift/reduce
+conflicts. However, this mode insists on starting the .output file with a
+list of *resolved* conflicts, something I find quite useless. Might it be
+possible to define a -v mode, and a -vv mode -- Where the -vv mode shows
+everything, but the -v mode only tells you what you need for examining
+conflicts? (Or, perhaps, a "*** This state has N conflicts ***" marker above
+each state with conflicts.)
+
+
+* $undefined
+From Hans:
+- If the Bison generated parser experiences an undefined number in the
+character range, that character is written out in diagnostic messages, an
+addition to the $undefined value.
+
+Suggest: Change the name $undefined to undefined; looks better in outputs.
+
+* Default Action
+From Hans:
+- For use with my C++ parser, I transported the "switch (yyn)" statement
+that Bison writes to the bison.simple skeleton file. This way, I can remove
+the current default rule $$ = $1 implementation, which causes a double
+assignment to $$ which may not be OK under C++, replacing it with a
+"default:" part within the switch statement.
+
+Note that the default rule $$ = $1, when typed, is perfectly OK under C,
+but in the C++ implementation I made, this rule is different from
+$<type_name>$ = $<type_name>1. I therefore think that one should implement
+a Bison option where every typed default rule is explicitly written out
+(same typed ruled can of course be grouped together).
+
+* Pre and post actions.
+From: Florian Krohm <florian@edamail.fishkill.ibm.com>
+Subject: YYACT_EPILOGUE
+To: bug-bison@gnu.org
+X-Sent: 1 week, 4 days, 14 hours, 38 minutes, 11 seconds ago
+
+The other day I had the need for explicitly building the parse tree. I
+used %locations for that and defined YYLLOC_DEFAULT to call a function
+that returns the tree node for the production. Easy. But I also needed
+to assign the S-attribute to the tree node. That cannot be done in
+YYLLOC_DEFAULT, because it is invoked before the action is executed.
+The way I solved this was to define a macro YYACT_EPILOGUE that would
+be invoked after the action. For reasons of symmetry I also added
+YYACT_PROLOGUE. Although I had no use for that I can envision how it
+might come in handy for debugging purposes.
+All is needed is to add
+
+#if YYLSP_NEEDED
+    YYACT_EPILOGUE (yyval, (yyvsp - yylen), yylen, yyloc, (yylsp - yylen));
+#else
+    YYACT_EPILOGUE (yyval, (yyvsp - yylen), yylen);
+#endif
+
+at the proper place to bison.simple. Ditto for YYACT_PROLOGUE.
+
+I was wondering what you think about adding YYACT_PROLOGUE/EPILOGUE
+to bison. If you're interested, I'll work on a patch.
+
+-----
+
+Copyright (C) 2001, 2002 Free Software Foundation, Inc.
+
+This file is part of GNU Autoconf.
+
+GNU Autoconf 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 2, or (at your option)
+any later version.
+
+GNU Autoconf 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.

 
 

-* testsuite.
-** tests/reduce.at     [ok]
+You should have received a copy of the GNU General Public License
+along with autoconf; see the file COPYING.  If not, write to
+the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+Boston, MA 02111-1307, USA.