]> git.saurik.com Git - bison.git/blame_incremental - TODO
* src/output.c (action_row): Let default_rule be always a rule
[bison.git] / TODO
... / ...
CommitLineData
1-*- outline -*-
2
3* URGENT: Documenting C++ output
4Write a first documentation for C++ output.
5
6* value_components_used
7Was defined but not used: where was it coming from? It can't be to
8check if %union is used, since the user is free to $<foo>n on her
9union, doesn't she?
10
11* yyerror, yyprint interface
12It should be improved, in particular when using Bison features such as
13locations, and YYPARSE_PARAMS. For the time being, it is recommended
14to #define yyerror and yyprint to steal internal variables...
15
16* documentation
17Explain $axiom (and maybe change its name: BTYacc names it `goal',
18byacc `$accept' probably based on AT&T Yacc, Meta `Start'...).
19Complete the glossary (item, axiom, ?).
20
21* Error messages
22Some are really funky. For instance
23
24 type clash (`%s' `%s') on default action
25
26is really weird. Revisit them all.
27
28* Report documentation
29Extend with error. The hard part will probably be finding the right
30rule so that a single state does not exhibit to many yet undocumented
31``features''. Maybe an empty action ought to be presented too. Shall
32we try to make a single grammar with all these features, or should we
33have several very small grammars?
34
35* Documentation
36Some history of Bison and some bibliography would be most welcome.
37Are there any Texinfo standards for bibliography?
38
39* Several %unions
40I think this is a pleasant (but useless currently) feature, but in the
41future, I want a means to %include other bits of grammars, and _then_
42it will be important for the various bits to define their needs in
43%union.
44
45When implementing multiple-%union support, bare the following in mind:
46
47- when --yacc, this must be flagged as an error. Don't make it fatal
48 though.
49
50- The #line must now appear *inside* the definition of yystype.
51 Something like
52
53 {
54 #line 12 "foo.y"
55 int ival;
56 #line 23 "foo.y"
57 char *sval;
58 }
59
60* --report=conflict-path
61Provide better assistance for understanding the conflicts by providing
62a sample text exhibiting the (LALR) ambiguity. See the paper from
63DeRemer and Penello: they already provide the algorithm.
64
65* Coding system independence
66Paul notes:
67
68 Currently Bison assumes 8-bit bytes (i.e. that UCHAR_MAX is
69 255). It also assumes that the 8-bit character encoding is
70 the same for the invocation of 'bison' as it is for the
71 invocation of 'cc', but this is not necessarily true when
72 people run bison on an ASCII host and then use cc on an EBCDIC
73 host. I don't think these topics are worth our time
74 addressing (unless we find a gung-ho volunteer for EBCDIC or
75 PDP-10 ports :-) but they should probably be documented
76 somewhere.
77
78* Unit rules
79Maybe we could expand unit rules, i.e., transform
80
81 exp: arith | bool;
82 arith: exp '+' exp;
83 bool: exp '&' exp;
84
85into
86
87 exp: exp '+' exp | exp '&' exp;
88
89when there are no actions. This can significantly speed up some
90grammars. I can't find the papers. In particular the book `LR
91parsing: Theory and Practice' is impossible to find, but according to
92`Parsing Techniques: a Practical Guide', it includes information about
93this issue. Does anybody have it?
94
95* Stupid error messages
96An example shows it easily:
97
98src/bison/tests % ./testsuite -k calc,location,error-verbose -l
99GNU Bison 1.49a test suite test groups:
100
101 NUM: FILENAME:LINE TEST-GROUP-NAME
102 KEYWORDS
103
104 51: calc.at:440 Calculator --locations --yyerror-verbose
105 52: calc.at:442 Calculator --defines --locations --name-prefix=calc --verbose --yacc --yyerror-verbose
106 54: calc.at:445 Calculator --debug --defines --locations --name-prefix=calc --verbose --yacc --yyerror-verbose
107src/bison/tests % ./testsuite 51 -d
108## --------------------------- ##
109## GNU Bison 1.49a test suite. ##
110## --------------------------- ##
111 51: calc.at:440 ok
112## ---------------------------- ##
113## All 1 tests were successful. ##
114## ---------------------------- ##
115src/bison/tests % cd ./testsuite.dir/51
116tests/testsuite.dir/51 % echo "()" | ./calc
1171.2-1.3: parse error, unexpected ')', expecting error or "number" or '-' or '('
118
119* read_pipe.c
120This is not portable to DOS for instance. Implement a more portable
121scheme. Sources of inspiration include GNU diff, and Free Recode.
122
123* Memory leaks in the generator
124A round of memory leak clean ups would be most welcome. Dmalloc,
125Checker GCC, Electric Fence, or Valgrind: you chose your tool.
126
127* --graph
128Show reductions. []
129
130* Broken options ?
131** %no-lines [ok]
132** %no-parser []
133** %pure-parser []
134** %token-table []
135** Options which could use parse_dquoted_param ().
136Maybe transfered in lex.c.
137*** %skeleton [ok]
138*** %output []
139*** %file-prefix []
140*** %name-prefix []
141
142** Skeleton strategy. []
143Must we keep %no-parser?
144 %token-table?
145*** New skeletons. []
146
147* src/print_graph.c
148Find the best graph parameters. []
149
150* doc/bison.texinfo
151** Update
152informations about ERROR_VERBOSE. []
153** Add explainations about
154skeleton muscles. []
155%skeleton. []
156
157* testsuite
158** tests/pure-parser.at []
159New tests.
160
161* input synclines
162Some users create their foo.y files, and equip them with #line. Bison
163should recognize these, and preserve them.
164
165* BTYacc
166See if we can integrate backtracking in Bison. Contact the BTYacc
167maintainers.
168
169* RR conflicts
170See if we can use precedence between rules to solve RR conflicts. See
171what POSIX says.
172
173* Precedence
174It is unfortunate that there is a total order for precedence. It
175makes it impossible to have modular precedence information. We should
176move to partial orders.
177
178This will be possible with a Bison parser for the grammar, as it will
179make it much easier to extend the grammar.
180
181* $undefined
182From Hans:
183- If the Bison generated parser experiences an undefined number in the
184character range, that character is written out in diagnostic messages, an
185addition to the $undefined value.
186
187Suggest: Change the name $undefined to undefined; looks better in outputs.
188
189* Default Action
190From Hans:
191- For use with my C++ parser, I transported the "switch (yyn)" statement
192that Bison writes to the bison.simple skeleton file. This way, I can remove
193the current default rule $$ = $1 implementation, which causes a double
194assignment to $$ which may not be OK under C++, replacing it with a
195"default:" part within the switch statement.
196
197Note that the default rule $$ = $1, when typed, is perfectly OK under C,
198but in the C++ implementation I made, this rule is different from
199$<type_name>$ = $<type_name>1. I therefore think that one should implement
200a Bison option where every typed default rule is explicitly written out
201(same typed ruled can of course be grouped together).
202
203Note: Robert Anisko handles this. He knows how to do it.
204
205* Warnings
206It would be nice to have warning support. See how Autoconf handles
207them, it is fairly well described there. It would be very nice to
208implement this in such a way that other programs could use
209lib/warnings.[ch].
210
211Don't work on this without first announcing you do, as I already have
212thought about it, and know many of the components that can be used to
213implement it.
214
215* Pre and post actions.
216From: Florian Krohm <florian@edamail.fishkill.ibm.com>
217Subject: YYACT_EPILOGUE
218To: bug-bison@gnu.org
219X-Sent: 1 week, 4 days, 14 hours, 38 minutes, 11 seconds ago
220
221The other day I had the need for explicitly building the parse tree. I
222used %locations for that and defined YYLLOC_DEFAULT to call a function
223that returns the tree node for the production. Easy. But I also needed
224to assign the S-attribute to the tree node. That cannot be done in
225YYLLOC_DEFAULT, because it is invoked before the action is executed.
226The way I solved this was to define a macro YYACT_EPILOGUE that would
227be invoked after the action. For reasons of symmetry I also added
228YYACT_PROLOGUE. Although I had no use for that I can envision how it
229might come in handy for debugging purposes.
230All is needed is to add
231
232#if YYLSP_NEEDED
233 YYACT_EPILOGUE (yyval, (yyvsp - yylen), yylen, yyloc, (yylsp - yylen));
234#else
235 YYACT_EPILOGUE (yyval, (yyvsp - yylen), yylen);
236#endif
237
238at the proper place to bison.simple. Ditto for YYACT_PROLOGUE.
239
240I was wondering what you think about adding YYACT_PROLOGUE/EPILOGUE
241to bison. If you're interested, I'll work on a patch.
242
243* Move to Graphviz
244Well, VCG seems really dead. Move to Graphviz instead. Also, equip
245the parser with a means to create the (visual) parse tree.
246
247-----
248
249Copyright (C) 2001, 2002 Free Software Foundation, Inc.
250
251This file is part of GNU Bison.
252
253GNU Bison is free software; you can redistribute it and/or modify
254it under the terms of the GNU General Public License as published by
255the Free Software Foundation; either version 2, or (at your option)
256any later version.
257
258GNU Bison is distributed in the hope that it will be useful,
259but WITHOUT ANY WARRANTY; without even the implied warranty of
260MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
261GNU General Public License for more details.
262
263You should have received a copy of the GNU General Public License
264along with Bison; see the file COPYING. If not, write to
265the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
266Boston, MA 02111-1307, USA.