src/symtab.h

   1 /* Definitions for symtab.c and callers, part of Bison.
   2
   3    Copyright (C) 1984, 1989, 1992, 2000-2002, 2004-2012 Free Software
   4    Foundation, Inc.
   5
   6    This file is part of Bison, the GNU Compiler Compiler.
   7
   8    This program is free software: you can redistribute it and/or modify
   9    it under the terms of the GNU General Public License as published by
  10    the Free Software Foundation, either version 3 of the License, or
  11    (at your option) any later version.
  12
  13    This program is distributed in the hope that it will be useful,
  14    but WITHOUT ANY WARRANTY; without even the implied warranty of
  15    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16    GNU General Public License for more details.
  17
  18    You should have received a copy of the GNU General Public License
  19    along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
  20
  21 /**
  22  * \file symtab.h
  23  * \brief Manipulating ::symbol.
  24  */
  25
  26 #ifndef SYMTAB_H_
  27 # define SYMTAB_H_
  28
  29 # include "assoc.h"
  30 # include "location.h"
  31 # include "scan-code.h"
  32 # include "uniqstr.h"
  33
  34 /*----------.
  35 | Symbols.  |
  36 `----------*/
  37
  38 /** Symbol classes.  */
  39 typedef enum
  40 {
  41   unknown_sym,          /**< Undefined.  */
  42   token_sym,            /**< Terminal. */
  43   nterm_sym             /**< Non-terminal. */
  44 } symbol_class;
  45
  46
  47 /** Internal token numbers. */
  48 typedef int symbol_number;
  49 #define SYMBOL_NUMBER_MAXIMUM INT_MAX
  50
  51
  52 typedef struct symbol symbol;
  53
  54 /* Declaration status of a symbol.
  55
  56    First, it is "undeclared".  Then, if "undeclared" and used in a
  57    %printer/%destructor, it is "used".  If not "declared" by used in a
  58    rule, it is "needed".  Finally, if declared (via a rule for
  59    nonterminals, or %oken), it is "declared".
  60
  61    When status are checked at the end, "declared" symbols are fine,
  62    "used" symbols trigger warnings, otherwise it's an error.
  63 */
  64
  65 typedef enum
  66   {
  67     /** Used in the input file for an unknown reason (error).  */
  68     undeclared,
  69     /** Used by %destructor/%printer but not defined (warning).  */
  70     used,
  71     /** Used in the gramar (rules) but not defined (error).  */
  72     needed,
  73     /** Defined with %type or %token (good).  */
  74     declared,
  75   } status;
  76
  77 typedef enum code_props_type code_props_type;
  78 enum code_props_type
  79   {
  80     destructor = 0,
  81     printer = 1,
  82   };
  83
  84 enum { CODE_PROPS_SIZE = 2 };
  85
  86 /* When extending this structure, be sure to complete
  87    symbol_check_alias_consistency.  */
  88 struct symbol
  89 {
  90   /** The key, name of the symbol.  */
  91   uniqstr tag;
  92   /** The location of its first occurrence.  */
  93   location location;
  94
  95   /** Its \c \%type.
  96
  97       Beware that this is the type_name as was entered by the user,
  98       including silly things such as "]" if she entered "%token <]> t".
  99       Therefore, when outputting type_name to M4, be sure to escape it
 100       into "@}".  See quoted_output for instance.  */
 101   uniqstr type_name;
 102
 103   /** Its \c \%type's location.  */
 104   location type_location;
 105
 106   /** Any \c \%destructor (resp. \%printer) declared specificially for this
 107       symbol.
 108
 109       Access this field only through <tt>symbol</tt>'s interface functions. For
 110       example, if <tt>symbol::destructor = NULL</tt> (resp. <tt>symbol::printer
 111       = NULL</tt>), a default \c \%destructor (resp. \%printer) or a per-type
 112       \c symbol_destructor_printer_get will compute the correct one. */
 113   code_props props[CODE_PROPS_SIZE];
 114
 115   symbol_number number;
 116   location prec_location;
 117   int prec;
 118   assoc assoc;
 119   int user_token_number;
 120
 121   /* Points to the other in the symbol-string pair for an alias.
 122      Special value USER_NUMBER_HAS_STRING_ALIAS in the symbol half of the
 123      symbol-string pair for an alias.  */
 124   symbol *alias;
 125   symbol_class class;
 126   status status;
 127 };
 128
 129 /** Undefined user number.  */
 130 #define USER_NUMBER_UNDEFINED -1
 131
 132 /* `symbol->user_token_number == USER_NUMBER_HAS_STRING_ALIAS' means
 133    this symbol has a literal string alias.  For instance, `%token foo
 134    "foo"' has `"foo"' numbered regularly, and `foo' numbered as
 135    USER_NUMBER_HAS_STRING_ALIAS.  */
 136 #define USER_NUMBER_HAS_STRING_ALIAS -9991
 137
 138 /* Undefined internal token number.  */
 139 #define NUMBER_UNDEFINED (-1)
 140
 141 /** Fetch (or create) the symbol associated to KEY.  */
 142 symbol *symbol_from_uniqstr (const uniqstr key, location loc);
 143
 144 /** Fetch (or create) the symbol associated to KEY.  */
 145 symbol *symbol_get (const char *key, location loc);
 146
 147 /** Generate a dummy nonterminal.
 148
 149    Its name cannot conflict with the user's names.  */
 150 symbol *dummy_symbol_get (location loc);
 151
 152
 153 /*--------------------.
 154 | Methods on symbol.  |
 155 `--------------------*/
 156
 157 /** Print a symbol (for debugging). */
 158 void symbol_print (symbol const *s, FILE *f);
 159
 160 /** Is this a dummy nonterminal?  */
 161 bool symbol_is_dummy (const symbol *sym);
 162
 163 /** The name of the code_props type: "\%destructor" or "\%printer".  */
 164 char const *code_props_type_string (code_props_type kind);
 165
 166 /** The name of the symbol that can be used as an identifier.
 167  ** Consider the alias if needed.
 168  ** Return 0 if there is none (e.g., the symbol is only defined as
 169  ** a string). */
 170 uniqstr symbol_id_get (symbol const *sym);
 171
 172 /**
 173  * Make \c str the literal string alias of \c sym.  Copy token number,
 174  * symbol number, and type from \c sym to \c str.
 175  */
 176 void symbol_make_alias (symbol *sym, symbol *str, location loc);
 177
 178 /** Set the \c type_name associated with \c sym.
 179
 180     Do nothing if passed 0 as \c type_name.  */
 181 void symbol_type_set (symbol *sym, uniqstr type_name, location loc);
 182
 183 /** Set the \c \%destructor or \c \%printer associated with \c sym.  */
 184 void symbol_code_props_set (symbol *sym, code_props_type kind,
 185                             code_props const *destructor);
 186
 187 /** Get the computed \c \%destructor or \c %printer for \c sym, which was
 188     initialized with \c code_props_none_init if there's no \c \%destructor or
 189     \c %printer.  */
 190 code_props *symbol_code_props_get (symbol *sym, code_props_type kind);
 191
 192 /** Set the \c precedence associated with \c sym.
 193
 194     Ensure that \a symbol is a terminal.
 195     Do nothing if invoked with \c undef_assoc as \c assoc.  */
 196 void symbol_precedence_set (symbol *sym, int prec, assoc a, location loc);
 197
 198 /** Set the \c class associated with \c sym.  */
 199 void symbol_class_set (symbol *sym, symbol_class class, location loc,
 200                        bool declaring);
 201
 202 /** Set the \c user_token_number associated with \c sym.  */
 203 void symbol_user_token_number_set (symbol *sym, int user_number, location loc);
 204
 205
 206
 207 /*------------------.
 208 | Special symbols.  |
 209 `------------------*/
 210
 211 /** The error token. */
 212 extern symbol *errtoken;
 213 /** The token for unknown tokens.  */
 214 extern symbol *undeftoken;
 215 /** The end of input token.  */
 216 extern symbol *endtoken;
 217 /** The genuine start symbol.
 218
 219    $accept: start-symbol $end */
 220 extern symbol *accept;
 221
 222 /** The user start symbol. */
 223 extern symbol *startsymbol;
 224 /** The location of the \c \%start declaration.  */
 225 extern location startsymbol_location;
 226
 227
 228 /*-----------------.
 229 | Semantic types.  |
 230 `-----------------*/
 231
 232 /** A semantic type and its associated \c \%destructor and \c \%printer.
 233
 234    Access the fields of this struct only through the interface functions in
 235    this file.  \sa symbol::destructor  */
 236 typedef struct {
 237   /** The key, name of the semantic type.  */
 238   uniqstr tag;
 239
 240   /** The location of its first occurence.  */
 241   location location;
 242
 243   /** Its status : "undeclared", "used" or "declared".
 244       It cannot be "needed".  */
 245   status status;
 246
 247   /** Any \c %destructor and %printer declared for this
 248       semantic type.  */
 249   code_props props[CODE_PROPS_SIZE];
 250
 251 } semantic_type;
 252
 253 /** Fetch (or create) the semantic type associated to KEY.  */
 254 semantic_type *semantic_type_from_uniqstr (const uniqstr key,
 255                                            const location *loc);
 256
 257 /** Fetch (or create) the semantic type associated to KEY.  */
 258 semantic_type *semantic_type_get (const char *key, const location *loc);
 259
 260 /** Set the \c destructor or \c printer associated with \c type.  */
 261 void semantic_type_code_props_set (semantic_type *type,
 262                                    code_props_type kind,
 263                                    code_props const *code);
 264
 265 /*----------------------------------.
 266 | Symbol and semantic type tables.  |
 267 `----------------------------------*/
 268
 269 /** Create the symbol and semantic type tables.  */
 270 void symbols_new (void);
 271
 272 /** Free all the memory allocated for symbols and semantic types.  */
 273 void symbols_free (void);
 274
 275 /** Check that all the symbols are defined.
 276
 277     Report any undefined symbols and consider them nonterminals.  */
 278 void symbols_check_defined (void);
 279
 280 /** Sanity checks and #token_translations construction.
 281
 282    Perform various sanity checks, assign symbol numbers, and set up
 283    #token_translations.  */
 284 void symbols_pack (void);
 285
 286 #endif /* !SYMTAB_H_ */