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 transitions are shifts (push the lookahead token
  48    and read another) and reductions (combine the last n things on the
  49    stack via a rule, replace them with the symbol that the rule
  50    derives, and leave the lookahead token alone).  When the states are
  51    generated, these transitions are represented in two other lists.
  52
  53    Each shifts structure describes the possible shift transitions out
  54    of one state, the state whose number is in the number field.  The
  55    shifts structures are linked through next and first_shift points to
  56    them.  Each contains a vector of numbers of the states that shift
  57    transitions can go to.  The accessing_symbol fields of those
  58    states' cores say what kind of input leads to them.
  59
  60    A shift to state zero should be ignored.  Conflict resolution
  61    deletes shifts by changing them 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 has the state number in its
  72    number field.  The rest of the errs structure is full of token
  73    numbers.
  74
  75    There is at least one shift transition present in state zero.  It
  76    leads to a next-to-final state whose accessing_symbol is the
  77    grammar's start symbol.  The next-to-final state has one shift to
  78    the final state, whose accessing_symbol is zero (end of input).
  79    The final state has one shift, which goes to the termination state
  80    (whose number is nstates-1).  The reason for the extra state at the
  81    end is to placate the parser's strategy of making all decisions one
  82    token ahead of its actions.  */
  83
  84 #ifndef STATE_H_
  85 # define STATE_H_
  86
  87 # include "bitsetv.h"
  88
  89
  90 /*-------------------.
  91 | Numbering states.  |
  92 `-------------------*/
  93
  94 typedef short state_number_t;
  95 # define STATE_NUMBER_MAX ((state_number_t) SHRT_MAX)
  96
  97 /* Be ready to map a state_number_t to an int.  */
  98 # define state_number_as_int(Tok) ((int) (Tok))
  99
 100 /*--------------.
 101 | Transitions.  |
 102 `--------------*/
 103
 104 typedef struct transtion_s
 105 {
 106   short num;
 107   state_number_t states[1];
 108 } transitions_t;
 109
 110
 111 /* What is the symbol which is shifted by TRANSITIONS->states[Shift]?  Can
 112    be a token (amongst which the error token), or non terminals in
 113    case of gotos.  */
 114
 115 #define TRANSITION_SYMBOL(Transitions, Shift) \
 116   (states[Transitions->states[Shift]]->accessing_symbol)
 117
 118 /* Is the TRANSITIONS->states[Shift] a real shift? (as opposed to gotos.) */
 119
 120 #define TRANSITION_IS_SHIFT(Transitions, Shift) \
 121   (ISTOKEN (TRANSITION_SYMBOL (Transitions, Shift)))
 122
 123 /* Is the TRANSITIONS->states[Shift] a goto?. */
 124
 125 #define TRANSITION_IS_GOTO(Transitions, Shift) \
 126   (!TRANSITION_IS_SHIFT (Transitions, Shift))
 127
 128 /* Is the TRANSITIONS->states[Shift] then handling of the error token?. */
 129
 130 #define TRANSITION_IS_ERROR(Transitions, Shift) \
 131   (TRANSITION_SYMBOL (Transitions, Shift) == errtoken->number)
 132
 133 /* When resolving a SR conflicts, if the reduction wins, the shift is
 134    disabled.  */
 135
 136 #define TRANSITION_DISABLE(Transitions, Shift) \
 137   (Transitions->states[Shift] = 0)
 138
 139 #define TRANSITION_IS_DISABLED(Transitions, Shift) \
 140   (Transitions->states[Shift] == 0)
 141
 142 /* Return the state such these TRANSITIONS contain a shift/goto to it on
 143    SYMBOL.  Aborts if none found.  */
 144 struct state_s;
 145 struct state_s *transitions_to PARAMS ((transitions_t *state,
 146                                         symbol_number_t s));
 147
 148
 149 /*-------.
 150 | Errs.  |
 151 `-------*/
 152
 153 typedef struct errs_s
 154 {
 155   short num;
 156   symbol_number_t symbols[1];
 157 } errs_t;
 158
 159 errs_t *errs_new PARAMS ((int n));
 160 errs_t *errs_dup PARAMS ((errs_t *src));
 161
 162
 163 /*-------------.
 164 | Reductions.  |
 165 `-------------*/
 166
 167 typedef struct reductions_s
 168 {
 169   short num;
 170   rule_number_t rules[1];
 171 } reductions_t;
 172
 173
 174
 175 /*----------.
 176 | State_t.  |
 177 `----------*/
 178
 179 typedef struct state_s
 180 {
 181   state_number_t number;
 182   symbol_number_t accessing_symbol;
 183   transitions_t     *shifts;
 184   reductions_t *reductions;
 185   errs_t       *errs;
 186
 187   /* Nonzero if no lookahead is needed to decide what to do in state S.  */
 188   char consistent;
 189
 190   /* Used in LALR, not LR(0).
 191
 192      When a state is not consistent (there is an S/R or R/R conflict),
 193      lookaheads are needed to enable the reductions.  NLOOKAHEADS is
 194      the number of lookahead guarded reductions of the
 195      LOOKAHEADS_RULE.  For each rule LOOKAHEADS_RULE[R], LOOKAHEADS[R]
 196      is the bitset of the lookaheads enabling this reduction.  */
 197   int nlookaheads;
 198   bitsetv lookaheads;
 199   rule_t **lookaheads_rule;
 200
 201   /* If some conflicts were solved thanks to precedence/associativity,
 202      a human readable description of the resolution.  */
 203   const char *solved_conflicts;
 204
 205   /* Its items.  Must be last, since ITEMS can be arbitrarily large.
 206      */
 207   unsigned short nitems;
 208   item_number_t items[1];
 209 } state_t;
 210
 211 extern state_number_t nstates;
 212 extern state_t *final_state;
 213
 214 /* Create a new state with ACCESSING_SYMBOL for those items.  */
 215 state_t *state_new PARAMS ((symbol_number_t accessing_symbol,
 216                             size_t core_size, item_number_t *core));
 217
 218 /* Set the shifts of STATE.  */
 219 void state_transitions_set PARAMS ((state_t *state,
 220                                int nshifts, state_number_t *shifts));
 221
 222 /* Set the reductions of STATE.  */
 223 void state_reductions_set PARAMS ((state_t *state,
 224                                    int nreductions, short *reductions));
 225
 226 /* Print on OUT all the lookaheads such that this STATE wants to
 227    reduce this RULE.  */
 228 void state_rule_lookaheads_print PARAMS ((state_t *state, rule_t *rule,
 229                                           FILE *out));
 230
 231 /* Create/destroy the states hash table.  */
 232 void state_hash_new PARAMS ((void));
 233 void state_hash_free PARAMS ((void));
 234
 235 /* Find the state associated to the CORE, and return it.  If it does
 236    not exist yet, return NULL.  */
 237 state_t *state_hash_lookup PARAMS ((size_t core_size, item_number_t *core));
 238
 239 /* Insert STATE in the state hash table.  */
 240 void state_hash_insert PARAMS ((state_t *state));
 241
 242 /* All the states, indexed by the state number.  */
 243 extern state_t **states;
 244
 245 /* Free all the states.  */
 246 void states_free PARAMS ((void));
 247 #endif /* !STATE_H_ */