]> git.saurik.com Git - bison.git/blame - TODO
Update.
[bison.git] / TODO
CommitLineData
416bd7a9
MA
1-*- outline -*-
2
52d1aeee
MA
3* NEWS
4Sort from 1.31 NEWS.
5
bcb05e75
MA
6* Prologue
7The %union is declared after the user C declarations. It can be
704a47c4 8a problem if YYSTYPE is declared after the user part. []
bcb05e75 9
704a47c4
AD
10Actually, the real problem seems that the %union ought to be output
11where it was defined. For instance, in gettext/intl/plural.y, we
12have:
13
14 %{
15 ...
16 #include "gettextP.h"
17 ...
18 %}
19
20 %union {
21 unsigned long int num;
22 enum operator op;
23 struct expression *exp;
24 }
25
26 %{
27 ...
28 static int yylex PARAMS ((YYSTYPE *lval, const char **pexp));
29 ...
30 %}
31
32Where the first part defines struct expression, the second uses it to
33define YYSTYPE, and the last uses YYSTYPE. Only this order is valid.
bcb05e75
MA
34
35* --graph
36Show reductions. []
37
704a47c4 38* Broken options ?
c3995d99 39** %no-lines [ok]
04a76783 40** %no-parser []
fbbf9b3b 41** %pure-parser []
04a76783
MA
42** %semantic-parser []
43** %token-table []
44** Options which could use parse_dquoted_param ().
45Maybe transfered in lex.c.
46*** %skeleton [ok]
47*** %output []
48*** %file-prefix []
49*** %name-prefix []
ec93a213 50
fbbf9b3b 51** Skeleton strategy. []
c3a8cbaa
MA
52Must we keep %no-parser?
53 %token-table?
fbbf9b3b 54*** New skeletons. []
416bd7a9 55
c111e171 56* src/print_graph.c
31b53af2 57Find the best graph parameters. []
63c2d5de
MA
58
59* doc/bison.texinfo
1a4648ff 60** Update
c3a8cbaa 61informations about ERROR_VERBOSE. []
1a4648ff 62** Add explainations about
c3a8cbaa
MA
63skeleton muscles. []
64%skeleton. []
eeeb962b 65
704a47c4 66* testsuite
c3a8cbaa
MA
67** tests/pure-parser.at []
68New tests.
0f8d586a
AD
69
70* Debugging parsers
71
72From Greg McGary:
73
74akim demaille <akim.demaille@epita.fr> writes:
75
76> With great pleasure! Nonetheless, things which are debatable
77> (or not, but just `big') should be discuss in `public': something
78> like help- or bug-bison@gnu.org is just fine. Jesse and I are there,
79> but there is also Jim and some other people.
80
81I have no idea whether it qualifies as big or controversial, so I'll
82just summarize for you. I proposed this change years ago and was
83surprised that it was met with utter indifference!
84
85This debug feature is for the programs/grammars one develops with
86bison, not for debugging bison itself. I find that the YYDEBUG
87output comes in a very inconvenient format for my purposes.
88When debugging gcc, for instance, what I want is to see a trace of
89the sequence of reductions and the line#s for the semantic actions
90so I can follow what's happening. Single-step in gdb doesn't cut it
91because to move from one semantic action to the next takes you through
92lots of internal machinery of the parser, which is uninteresting.
93
94The change I made was to the format of the debug output, so that it
95comes out in the format of C error messages, digestible by emacs
96compile mode, like so:
97
98grammar.y:1234: foo: bar(0x123456) baz(0x345678)
99
100where "foo: bar baz" is the reduction rule, whose semantic action
101appears on line 1234 of the bison grammar file grammar.y. The hex
102numbers on the rhs tokens are the parse-stack values associated with
103those tokens. Of course, yytype might be something totally
104incompatible with that representation, but for the most part, yytype
105values are single words (scalars or pointers). In the case of gcc,
106they're most often pointers to tree nodes. Come to think of it, the
107right thing to do is to make the printing of stack values be
108user-definable. It would also be useful to include the filename &
109line# of the file being parsed, but the main filename & line# should
110continue to be that of grammar.y
111
112Anyway, this feature has saved my life on numerous occasions. The way
113I customarily use it is to first run bison with the traces on, isolate
114the sequence of reductions that interests me, put those traces in a
115buffer and force it into compile-mode, then visit each of those lines
116in the grammar and set breakpoints with C-x SPACE. Then, I can run
117again under the control of gdb and stop at each semantic action.
118With the hex addresses of tree nodes, I can inspect the values
119associated with any rhs token.
120
121You like?