]> git.saurik.com Git - bison.git/blob - src/state.h
Build version.c automatically.
[bison.git] / src / state.h
1 /* Type definitions for nondeterministic finite state machine for bison,
2 Copyright (C) 1984, 1989 Free Software Foundation, Inc.
3
4 This file is part of Bison, the GNU Compiler Compiler.
5
6 Bison is free software; you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 2, or (at your option)
9 any later version.
10
11 Bison is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 GNU General Public License for more details.
15
16 You should have received a copy of the GNU General Public License
17 along with Bison; see the file COPYING. If not, write to
18 the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. */
19
20
21 /* These type definitions are used to represent a nondeterministic
22 finite state machine that parses the specified grammar.
23 This information is generated by the function generate_states
24 in the file LR0.
25
26 Each state of the machine is described by a set of items --
27 particular positions in particular rules -- that are the possible
28 places where parsing could continue when the machine is in this state.
29 These symbols at these items are the allowable inputs that can follow now.
30
31 A core represents one state. States are numbered in the number field.
32 When generate_states is finished, the starting state is state 0
33 and nstates is the number of states. (A transition to a state
34 whose state number is nstates indicates termination.) All the cores
35 are chained together and first_state points to the first one (state 0).
36
37 For each state there is a particular symbol which must have been the
38 last thing accepted to reach that state. It is the accessing_symbol
39 of the core.
40
41 Each core contains a vector of nitems items which are the indices
42 in the ritems vector of the items that are selected in this state.
43
44 The link field is used for chaining buckets that hash states by
45 their itemsets. This is for recognizing equivalent states and
46 combining them when the states are generated.
47
48 The two types of transitions are shifts (push the lookahead token
49 and read another) and reductions (combine the last n things on the
50 stack via a rule, replace them with the symbol that the rule derives,
51 and leave the lookahead token alone). When the states are generated,
52 these transitions are represented in two other lists.
53
54 Each shifts structure describes the possible shift transitions out
55 of one state, the state whose number is in the number field.
56 The shifts structures are linked through next and first_shift points to them.
57 Each contains a vector of numbers of the states that shift transitions
58 can go to. The accessing_symbol fields of those states' cores say what kind
59 of input leads to them.
60
61 A shift to state zero should be ignored. Conflict resolution
62 deletes shifts by changing them to zero.
63
64 Each reductions structure describes the possible reductions at the state
65 whose number is in the number field. The data is a list of nreds rules,
66 represented by their rule numbers. first_reduction points to the list
67 of these structures.
68
69 Conflict resolution can decide that certain tokens in certain
70 states should explicitly be errors (for implementing %nonassoc).
71 For each state, the tokens that are errors for this reason
72 are recorded in an errs structure, which has the state number
73 in its number field. The rest of the errs structure is full
74 of token numbers.
75
76 There is at least one shift transition present in state zero.
77 It leads to a next-to-final state whose accessing_symbol is
78 the grammar's start symbol. The next-to-final state has one shift
79 to the final state, whose accessing_symbol is zero (end of input).
80 The final state has one shift, which goes to the termination state
81 (whose number is nstates-1).
82 The reason for the extra state at the end is to placate the parser's
83 strategy of making all decisions one token ahead of its actions. */
84
85
86 typedef
87 struct core
88 {
89 struct core *next;
90 struct core *link;
91 short number;
92 short accessing_symbol;
93 short nitems;
94 short items[1];
95 }
96 core;
97
98
99
100 typedef
101 struct shifts
102 {
103 struct shifts *next;
104 short number;
105 short nshifts;
106 short shifts[1];
107 }
108 shifts;
109
110
111
112 typedef
113 struct errs
114 {
115 short nerrs;
116 short errs[1];
117 }
118 errs;
119
120
121
122 typedef
123 struct reductions
124 {
125 struct reductions *next;
126 short number;
127 short nreds;
128 short rules[1];
129 }
130 reductions;
131
132
133
134 extern int nstates;
135 extern core *first_state;
136 extern shifts *first_shift;
137 extern reductions *first_reduction;