researchv10dc/cmd/bison/gram.h - annotate

Return to gram.h CVS log

Up to [Research Unix] / researchv10dc / cmd / bison

Annotation of researchv10dc/cmd/bison/gram.h, revision 1.1.1.1

1.1 root 1: /* Data definitions for internal representation of bison's input,
2: Copyright (C) 1984, 1986 Bob Corbett and Free Software Foundation, Inc.
3:
4: BISON is distributed in the hope that it will be useful, but WITHOUT ANY
5: WARRANTY. No author or distributor accepts responsibility to anyone
6: for the consequences of using it or for whether it serves any
7: particular purpose or works at all, unless he says so in writing.
8: Refer to the BISON General Public License for full details.
9:
10: Everyone is granted permission to copy, modify and redistribute BISON,
11: but only under the conditions described in the BISON General Public
12: License. A copy of this license is supposed to have been given to you
13: along with BISON so you can know your rights and responsibilities. It
14: should be in a file named COPYING. Among other things, the copyright
15: notice and this notice must be preserved on all copies.
16:
17: In other words, you are welcome to use, share and improve this program.
18: You are forbidden to forbid anyone else to use, share and improve
19: what you give them. Help stamp out software-hoarding! */
20:
21: /* representation of the grammar rules:
22:
23: ntokens is the number of tokens, and nvars is the number of variables (nonterminals).
24: nsyms is the total number, ntokens + nvars.
25:
26: Each symbol (either token or variable) receives a symbol number.
27: Numbers 0 to ntokens-1 are for tokens, and ntokens to nsyms-1 are for variables.
28: Symbol number zero is the end-of-input token. This token is counted in ntokens.
29:
30: The rules receive rule numbers 1 to nrules in the order they are written.
31: Actions and guards are accessed via the rule number.
32:
33: The rules themselves are described by three arrays: rrhs, rlhs and ritems.
34: rlhs[r] is the symbol number of the left hand side of rule r.
35: The right hand side is stored as symbol numbers in a portion of ritems.
36: rrhs[r] contains the index in ritems of the beginning of the portion for rule r.
37: The length of the portion is one greater
38: than the number of symbols in the rule's right hand side.
39: The last element in the portion contains minus r, which
40: identifies it as the end of a portion and says which rule it is for.
41:
42: The portions of ritems come in order of increasing rule number and are
43: followed by an element which is zero to mark the end. nitems is the
44: total length of ritems, not counting the final zero. Each element of
45: ritems is called an "item" and its index in ritems is an item number.
46:
47: Item numbers are used in the finite state machine to represent
48: places that parsing can get to.
49:
50: Precedence levels are recorded in the vectors sprec and rprec.
51: sprec records the precedence level of each symbol,
52: rprec the precedence level of each rule.
53:
54: Precedence levels are assigned in increasing order starting with 1
55: so that numerically higher precedence values mean tighter binding
56: as they ought to. Zero as a symbol or rule's precedence means none is assigned.
57:
58: Associativities are recorded similarly in rassoc and sassoc. */
59:
60:
61: #define ISTOKEN(s) ((s) < ntokens)
62: #define ISVAR(s) ((s) >= ntokens)
63:
64:
65: extern int nitems;
66: extern int nrules;
67: extern int nsyms;
68: extern int ntokens;
69: extern int nvars;
70:
71: extern short *ritem;
72: extern short *rlhs;
73: extern short *rrhs;
74: extern short *rprec;
75: extern short *sprec;
76: extern short *rassoc;
77: extern short *sassoc;
78: extern short *rline; /* Source line number of each rule */
79:
80: extern int start_symbol;
81:
82:
83: /* associativity values in elements of rassoc, sassoc. */
84:
85: #define RIGHT_ASSOC 1
86: #define LEFT_ASSOC 2
87: #define NON_ASSOC 3
88:
89: /* token translation table:
90: indexed by a token number as returned by the user's yylex routine,
91: it yields the internal token number used by the parser and throughout bison.
92: If translations is zero, the translation table is not used because
93: the two kinds of token numbers are the same. */
94:
95: extern short *token_translations;
96: extern int translations;
97: extern int max_user_token_number;
98:
99: /* semantic_parser is nonzero if the input file says to use the hairy parser
100: that provides for semantic error recovery. If it is zero, the yacc-compatible
101: simplified parser is used. */
102:
103: extern int semantic_parser;
104:
105: /* pure_parser is nonzero if should generate a parser that is all pure and reentrant. */
106:
107: extern int pure_parser;
108:
109: /* error_token_number is the token number of the error token. */
110:
111: extern int error_token_number;

unix.superglobalmegacorp.com

This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.