src/state.h

   1 /* Type definitions for nondeterministic finite state machine for bison,
   2    Copyright 1984, 1989, 2000, 2001 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, Inc., 59 Temple Place - Suite 330,
  19    Boston, MA 02111-1307, USA.  */
  20
  21
  22 /* These type definitions are used to represent a nondeterministic
  23    finite state machine that parses the specified grammar.  This
  24    information is generated by the function generate_states in the
  25    file LR0.
  26
  27    Each state of the machine is described by a set of items --
  28    particular positions in particular rules -- that are the possible
  29    places where parsing could continue when the machine is in this
  30    state.  These symbols at these items are the allowable inputs that
  31    can follow now.
  32
  33    A core represents one state.  States are numbered in the NUMBER
  34    field.  When generate_states is finished, the starting state is
  35    state 0 and NSTATES is the number of states.  (FIXME: This sentence
  36    is no longer true: A transition to a state whose state number is
  37    NSTATES indicates termination.)  All the cores are chained together
  38    and FIRST_STATE points to the first one (state 0).
  39
  40    For each state there is a particular symbol which must have been
  41    the last thing accepted to reach that state.  It is the
  42    ACCESSING_SYMBOL of the core.
  43
  44    Each core contains a vector of NITEMS items which are the indices
  45    in the RITEMS vector of the items that are selected in this state.
  46
  47    The two types of actions are shifts/gotos (push the lookahead token
  48    and read another/goto to the state designated by a nterm) and
  49    reductions (combine the last n things on the stack via a rule,
  50    replace them with the symbol that the rule derives, and leave the
  51    lookahead token alone).  When the states are generated, these
  52    actions are represented in two other lists.
  53
  54    Each transition_t structure describes the possible transitions out
  55    of one state, the state whose number is in the number field.  Each
  56    contains a vector of numbers of the states that transitions can go
  57    to.  The accessing_symbol fields of those states' cores say what
  58    kind of input leads to them.
  59
  60    A transition to state zero should be ignored: conflict resolution
  61    deletes transitions by having them point to zero.
  62
  63    Each reductions structure describes the possible reductions at the
  64    state whose number is in the number field.  The data is a list of
  65    nreds rules, represented by their rule numbers.  first_reduction
  66    points to the list of these structures.
  67
  68    Conflict resolution can decide that certain tokens in certain
  69    states should explicitly be errors (for implementing %nonassoc).
  70    For each state, the tokens that are errors for this reason are
  71    recorded in an errs structure, which holds the token numbers.
  72
  73    There is at least one goto transition present in state zero.  It
  74    leads to a next-to-final state whose accessing_symbol is the
  75    grammar's start symbol.  The next-to-final state has one shift to
  76    the final state, whose accessing_symbol is zero (end of input).
  77    The final state has one shift, which goes to the termination state.
  78    The reason for the extra state at the end is to placate the
  79    parser's strategy of making all decisions one token ahead of its
  80    actions.  */
  81
  82 #ifndef STATE_H_
  83 # define STATE_H_
  84
  85 # include "bitsetv.h"
  86
  87
  88 /*-------------------.
  89 | Numbering states.  |
  90 `-------------------*/
  91
  92 typedef short state_number_t;
  93 # define STATE_NUMBER_MAX ((state_number_t) SHRT_MAX)
  94
  95 /* Be ready to map a state_number_t to an int.  */
  96 # define state_number_as_int(Tok) ((int) (Tok))
  97
  98 /*--------------.
  99 | Transitions.  |
 100 `--------------*/
 101
 102 typedef struct transtion_s
 103 {
 104   short num;
 105   state_number_t states[1];
 106 } transitions_t;
 107
 108
 109 /* What is the symbol labelling the transition to
 110    TRANSITIONS->states[Num]?  Can be a token (amongst which the error
 111    token), or non terminals in case of gotos.  */
 112
 113 #define TRANSITION_SYMBOL(Transitions, Num) \
 114   (states[Transitions->states[Num]]->accessing_symbol)
 115
 116 /* Is the TRANSITIONS->states[Num] a shift? (as opposed to gotos).  */
 117
 118 #define TRANSITION_IS_SHIFT(Transitions, Num) \
 119   (ISTOKEN (TRANSITION_SYMBOL (Transitions, Num)))
 120
 121 /* Is the TRANSITIONS->states[Num] a goto?. */
 122
 123 #define TRANSITION_IS_GOTO(Transitions, Num) \
 124   (!TRANSITION_IS_SHIFT (Transitions, Num))
 125
 126 /* Is the TRANSITIONS->states[Num] labelled by the error token?  */
 127
 128 #define TRANSITION_IS_ERROR(Transitions, Num) \
 129   (TRANSITION_SYMBOL (Transitions, Num) == errtoken->number)
 130
 131 /* When resolving a SR conflicts, if the reduction wins, the shift is
 132    disabled.  */
 133
 134 #define TRANSITION_DISABLE(Transitions, Num) \
 135   (Transitions->states[Num] = 0)
 136
 137 #define TRANSITION_IS_DISABLED(Transitions, Num) \
 138   (Transitions->states[Num] == 0)
 139
 140 /* Return the state such these TRANSITIONS contain a shift/goto to it on
 141    SYMBOL.  Aborts if none found.  */
 142 struct state_s;
 143 struct state_s *transitions_to PARAMS ((transitions_t *state,
 144                                         symbol_number_t s));
 145
 146
 147 /*-------.
 148 | Errs.  |
 149 `-------*/
 150
 151 typedef struct errs_s
 152 {
 153   short num;
 154   symbol_number_t symbols[1];
 155 } errs_t;
 156
 157 errs_t *errs_new PARAMS ((int num, symbol_number_t *tokens));
 158
 159
 160 /*-------------.
 161 | Reductions.  |
 162 `-------------*/
 163
 164 typedef struct reductions_s
 165 {
 166   short num;
 167   rule_number_t rules[1];
 168 } reductions_t;
 169
 170
 171
 172 /*----------.
 173 | State_t.  |
 174 `----------*/
 175
 176 typedef struct state_s
 177 {
 178   state_number_t number;
 179   symbol_number_t accessing_symbol;
 180   transitions_t     *transitions;
 181   reductions_t *reductions;
 182   errs_t       *errs;
 183
 184   /* Nonzero if no lookahead is needed to decide what to do in state S.  */
 185   char consistent;
 186
 187   /* Used in LALR, not LR(0).
 188
 189      When a state is not consistent (there is an S/R or R/R conflict),
 190      lookaheads are needed to enable the reductions.  NLOOKAHEADS is
 191      the number of lookahead guarded reductions of the
 192      LOOKAHEADS_RULE.  For each rule LOOKAHEADS_RULE[R], LOOKAHEADS[R]
 193      is the bitset of the lookaheads enabling this reduction.  */
 194   int nlookaheads;
 195   bitsetv lookaheads;
 196   rule_t **lookaheads_rule;
 197
 198   /* If some conflicts were solved thanks to precedence/associativity,
 199      a human readable description of the resolution.  */
 200   const char *solved_conflicts;
 201
 202   /* Its items.  Must be last, since ITEMS can be arbitrarily large.
 203      */
 204   unsigned short nitems;
 205   item_number_t items[1];
 206 } state_t;
 207
 208 extern state_number_t nstates;
 209 extern state_t *final_state;
 210
 211 /* Create a new state with ACCESSING_SYMBOL for those items.  */
 212 state_t *state_new PARAMS ((symbol_number_t accessing_symbol,
 213                             size_t core_size, item_number_t *core));
 214
 215 /* Set the transitions of STATE.  */
 216 void state_transitions_set PARAMS ((state_t *state,
 217                                     int num, state_number_t *transitions));
 218
 219 /* Set the reductions of STATE.  */
 220 void state_reductions_set PARAMS ((state_t *state,
 221                                    int num, rule_number_t *reductions));
 222
 223 /* Set the errs of STATE.  */
 224 void state_errs_set PARAMS ((state_t *state,
 225                              int num, symbol_number_t *errs));
 226
 227 /* Print on OUT all the lookaheads such that this STATE wants to
 228    reduce this RULE.  */
 229 void state_rule_lookaheads_print PARAMS ((state_t *state, rule_t *rule,
 230                                           FILE *out));
 231
 232 /* Create/destroy the states hash table.  */
 233 void state_hash_new PARAMS ((void));
 234 void state_hash_free PARAMS ((void));
 235
 236 /* Find the state associated to the CORE, and return it.  If it does
 237    not exist yet, return NULL.  */
 238 state_t *state_hash_lookup PARAMS ((size_t core_size, item_number_t *core));
 239
 240 /* Insert STATE in the state hash table.  */
 241 void state_hash_insert PARAMS ((state_t *state));
 242
 243 /* All the states, indexed by the state number.  */
 244 extern state_t **states;
 245
 246 /* Free all the states.  */
 247 void states_free PARAMS ((void));
 248 #endif /* !STATE_H_ */