X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/b4cbf822eba2133e123d165603bc5542c5662b33..52f32190aa3e60eb079f8e374cf683094c5529ad:/TODO diff --git a/TODO b/TODO index 27d2c084..168b4e03 100644 --- a/TODO +++ b/TODO @@ -1,49 +1,97 @@ -*- outline -*- +* Header guards -* URGENT: Prologue -The %union is declared after the user C declarations. It can be -a problem if YYSTYPE is declared after the user part. +From Franc,ois: should we keep the directory part in the CPP guard? -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" - ... - %} +* Yacc.c: CPP Macros - %union { - unsigned long int num; - enum operator op; - struct expression *exp; - } +Do some people use YYPURE, YYLSP_NEEDED like we do in the test suite? +They should not: it is not documented. But if they need to, let's +find something clean (not like YYLSP_NEEDED...). + + +* URGENT: Documenting C++ output +Write a first documentation for C++ output. + + +* Documentation +Before releasing, make sure the documentation ("Understanding your +parser") refers to the current `output' format. + + +* GLR & C++ +Currently, the GLR parser cannot compile with a C++ compiler. + + +* Report + +** GLR +How would Paul like to display the conflicted actions? In particular, +what when two reductions are possible on a given lookahead, but one is +part of $default. Should we make the two reductions explicit, or just +keep $default? See the following point. + +** Disabled Reductions +See `tests/conflicts.at (Defaulted Conflicted Reduction)', and decide +what we want to do. - %{ - ... - static int yylex PARAMS ((YYSTYPE *lval, const char **pexp)); - ... - %} +** Documentation +Extend with error productions. The hard part will probably be finding +the right rule so that a single state does not exhibit too many yet +undocumented ``features''. Maybe an empty action ought to be +presented too. Shall we try to make a single grammar with all these +features, or should we have several very small grammars? -Where the first part defines struct expression, the second uses it to -define YYSTYPE, and the last uses YYSTYPE. Only this order is valid. +** --report=conflict-path +Provide better assistance for understanding the conflicts by providing +a sample text exhibiting the (LALR) ambiguity. See the paper from +DeRemer and Penello: they already provide the algorithm. -Note that we have the same problem with GCC. -I suggest splitting the prologue into pre-prologue and post-prologue. -The reason is that: +* Extensions -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) +** %destructor +I think we should document it as experimental, and allow its use in +the next releases. But we also need to port it to GLR. What about +lalr1.cc? Well, read what Hans reported, maybe we don't want +%detructor. On the other hand, there is no reason not to provide it: +users can avoid its use. -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. +** $foo +Have a look at the Lemon parser generator: instead of $1, $2 etc. they +can name the values. This is much more pleasant. For instance: + + exp (res): exp (a) '+' exp (b) { $res = $a + $b; }; + +I love this. I have been bitten too often by the removal of the +symbol, and forgetting to shift all the $n to $n-1. If you are +unlucky, it compiles... + +** $-1 +We should find a means to provide an access to values deep in the +stack. For instance, instead of + + baz: qux { $$ = $-1 + $0 + $1; } + +we should be able to have: + + foo($foo) bar($bar) baz($bar): qux($qux) { $baz = $foo + $bar + $qux; } + +Or something like this. + + +** yysymprint interface +It should be improved, in particular when using Bison features such as +locations, and YYPARSE_PARAMS. For the time being, it is almost +recommended to yyprint to steal internal variables... + +** Several %unions +I think this is a pleasant (but useless currently) feature, but in the +future, I want a means to %include other bits of grammars, and _then_ +it will be important for the various bits to define their needs in +%union. When implementing multiple-%union support, bare the following in mind: @@ -60,30 +108,63 @@ When implementing multiple-%union support, bare the following in mind: char *sval; } -* Language independent actions +** %if and the like +It should be possible to have %if/%else/%endif. The implementation is +not clear: should it be lexical or syntactic. Vadim Maslow thinks it +must be in the scanner: we must not parse what is in a switched off +part of %if. Akim Demaille thinks it should be in the parser, so as +to avoid falling into another CPP mistake. + +** -D, --define-muscle NAME=VALUE +To define muscles via cli. Or maybe support directly NAME=VALUE? + + +* Unit rules +Maybe we could expand unit rules, i.e., transform + + exp: arith | bool; + arith: exp '+' exp; + bool: exp '&' exp; + +into + + exp: exp '+' exp | exp '&' exp; + +when there are no actions. This can significantly speed up some +grammars. I can't find the papers. In particular the book `LR +parsing: Theory and Practice' is impossible to find, but according to +`Parsing Techniques: a Practical Guide', it includes information about +this issue. Does anybody have it? + + + +* Documentation + +** History/Bibliography +Some history of Bison and some bibliography would be most welcome. +Are there any Texinfo standards for bibliography? -Currently bison, the generator, transforms $1, $$ 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. +* Java, Fortran, 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: +** Java -3. stop quoting them +There are a couple of proposed outputs: -4. change the [ and ] in the actions into @<:@ and @:>@ +- BYACC/J + which is based on Byacc. + + +- Bison Java + which is based on Bison. + + +Sebastien Serrurier (serrur_s@epita.fr) is working on this: he is +expected to contact the authors, design the output, and implement it +into Bison. -5. extend the postprocessor to maps these back onto [ and ]. * Coding system independence Paul notes: @@ -98,184 +179,17 @@ Paul notes: 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 - ... - -> I'm in favor of -> -> %token FOO 256 -> %token BAR 257 -> -> and Bison moves error into 258. - -Yes, I think that's a valid extension too, if the user doesn't define -the token number for error. - -* 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 [] ** %token-table [] -** Options which could use parse_dquoted_param (). -Maybe transfered in lex.c. -*** %skeleton [ok] -*** %output [] -*** %file-prefix [] -*** %name-prefix [] - ** Skeleton strategy. [] Must we keep %no-parser? %token-table? -*** New skeletons. [] * src/print_graph.c Find the best graph parameters. [] @@ -283,7 +197,7 @@ Find the best graph parameters. [] * doc/bison.texinfo ** Update informations about ERROR_VERBOSE. [] -** Add explainations about +** Add explanations about skeleton muscles. [] %skeleton. [] @@ -291,127 +205,52 @@ skeleton muscles. [] ** tests/pure-parser.at [] New tests. -* Debugging parsers - -From Greg McGary: - -akim demaille 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. +See if we can integrate backtracking in Bison. Charles-Henri de +Boysson is working on this, and already has some +results. Vadim Maslow, the maintainer of BTYacc was contacted, and we +stay in touch with him. Adjusting the Bison grammar parser will be +needed to support some extra BTYacc features. This is less urgent. -* Automaton report -Display more clearly the lookaheads for each item. +** Keeping the conflicted actions +First, analyze the differences between byacc and btyacc (I'm referring +to the executables). Find where the conflicts are preserved. + +** Compare with the GLR tables +See how isomorphic the way BTYacc and the way the GLR adjustments in +Bison are compatible. *As much as possible* one should try to use the +same implementation in the Bison executables. I insist: it should be +very feasible to use the very same conflict tables. + +** Adjust the skeletons +Import the skeletons for C and C++. + +** Improve the skeletons +Have them support yysymprint, yydestruct and so forth. -* RR conflicts -See if we can use precedence between rules to solve RR conflicts. See -what POSIX says. * Precedence + +** Partial order 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" -Subject: Token Alias Bug -To: "'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" -Subject: Token Alias Bug -To: "'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.) +move to partial orders (sounds like series/parallel orders to me). + +This will be possible with a Bison parser for the grammar, as it will +make it much easier to extend the grammar. + +** Correlation b/w precedence and associativity +Also, I fail to understand why we have to assign the same +associativity to operators with the same precedence. For instance, +why can't I decide that the precedence of * and / is the same, but the +latter is nonassoc? + +If there is really no profound motivation, we should find a new syntax +to allow specifying this. + +** RR conflicts +See if we can use precedence between rules to solve RR conflicts. See +what POSIX says. * $undefined @@ -422,6 +261,7 @@ 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 @@ -436,6 +276,20 @@ $$ = $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). +Note: Robert Anisko handles this. He knows how to do it. + + +* Warnings +It would be nice to have warning support. See how Autoconf handles +them, it is fairly well described there. It would be very nice to +implement this in such a way that other programs could use +lib/warnings.[ch]. + +Don't work on this without first announcing you do, as I already have +thought about it, and know many of the components that can be used to +implement it. + + * Pre and post actions. From: Florian Krohm Subject: YYACT_EPILOGUE @@ -464,23 +318,27 @@ 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. +* Move to Graphviz +Well, VCG seems really dead. Move to Graphviz instead. Also, equip +the parser with a means to create the (visual) parse tree. + ----- -Copyright (C) 2001, 2002 Free Software Foundation, Inc. +Copyright (C) 2001, 2002, 2003 Free Software Foundation, Inc. -This file is part of GNU Autoconf. +This file is part of GNU Bison. -GNU Autoconf is free software; you can redistribute it and/or modify +GNU Bison 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, +GNU Bison 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 autoconf; see the file COPYING. If not, write to +along with Bison; see the file COPYING. If not, write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.