c-exp.y

早期freebsd实现

  switch (c = *tokstart)
    {
    case 0:
      return 0;

    case ' ':
    case '\t':
    case '\n':
      lexptr++;
      goto retry;

    case '\'':
      /* We either have a character constant ('0' or '\177' for example)
	 or we have a quoted symbol reference ('foo(int,int)' in C++
	 for example). */
      lexptr++;
      c = *lexptr++;
      if (c == '\\')
	c = parse_escape (&lexptr);
      yylval.lval = c;
      c = *lexptr++;
      if (c != '\'')
	{
	  namelen = skip_quoted (tokstart) - tokstart;
	  if (namelen > 2)
	    {
	      lexptr = tokstart + namelen;
	      namelen -= 2;
	      tokstart++;
	      goto tryname;
	    }
	  error ("Invalid character constant.");
	}
      return CHAR;

    case '(':
      paren_depth++;
      lexptr++;
      return c;

    case ')':
      if (paren_depth == 0)
	return 0;
      paren_depth--;
      lexptr++;
      return c;

    case ',':
      if (comma_terminates && paren_depth == 0)
	return 0;
      lexptr++;
      return c;

    case '.':
      /* Might be a floating point number.  */
      if (lexptr[1] < '0' || lexptr[1] > '9')
	goto symbol;		/* Nope, must be a symbol. */
      /* FALL THRU into number case.  */

    case '0':
    case '1':
    case '2':
    case '3':
    case '4':
    case '5':
    case '6':
    case '7':
    case '8':
    case '9':
      {
	/* It's a number.  */
	int got_dot = 0, got_e = 0, toktype;
	register char *p = tokstart;
	int hex = input_radix > 10;

	if (c == '0' && (p[1] == 'x' || p[1] == 'X'))
	  {
	    p += 2;
	    hex = 1;
	  }
	else if (c == '0' && (p[1]=='t' || p[1]=='T' || p[1]=='d' || p[1]=='D'))
	  {
	    p += 2;
	    hex = 0;
	  }

	for (;; ++p)
	  {
	    if (!hex && !got_e && (*p == 'e' || *p == 'E'))
	      got_dot = got_e = 1;
	    else if (!hex && !got_dot && *p == '.')
	      got_dot = 1;
	    else if (got_e && (p[-1] == 'e' || p[-1] == 'E')
		     && (*p == '-' || *p == '+'))
	      /* This is the sign of the exponent, not the end of the
		 number.  */
	      continue;
	    /* We will take any letters or digits.  parse_number will
	       complain if past the radix, or if L or U are not final.  */
	    else if ((*p < '0' || *p > '9')
		     && ((*p < 'a' || *p > 'z')
				  && (*p < 'A' || *p > 'Z')))
	      break;
	  }
	toktype = parse_number (tokstart, p - tokstart, got_dot|got_e, &yylval);
        if (toktype == ERROR)
	  {
	    char *err_copy = (char *) alloca (p - tokstart + 1);

	    memcpy (err_copy, tokstart, p - tokstart);
	    err_copy[p - tokstart] = 0;
	    error ("Invalid number \"%s\".", err_copy);
	  }
	lexptr = p;
	return toktype;
      }

    case '+':
    case '-':
    case '*':
    case '/':
    case '%':
    case '|':
    case '&':
    case '^':
    case '~':
    case '!':
    case '@':
    case '<':
    case '>':
    case '[':
    case ']':
    case '?':
    case ':':
    case '=':
    case '{':
    case '}':
    symbol:
      lexptr++;
      return c;

    case '"':
      for (namelen = 1; (c = tokstart[namelen]) != '"'; namelen++)
	if (c == '\\')
	  {
	    c = tokstart[++namelen];
	    if (c >= '0' && c <= '9')
	      {
		c = tokstart[++namelen];
		if (c >= '0' && c <= '9')
		  c = tokstart[++namelen];
	      }
	  }
      yylval.sval.ptr = tokstart + 1;
      yylval.sval.length = namelen - 1;
      lexptr += namelen + 1;
      return STRING;
    }

  if (!(c == '_' || c == '$'
	|| (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z')))
    /* We must have come across a bad character (e.g. ';').  */
    error ("Invalid character '%c' in expression.", c);

  /* It's a name.  See how long it is.  */
  namelen = 0;
  for (c = tokstart[namelen];
       (c == '_' || c == '$' || (c >= '0' && c <= '9')
	|| (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z'));
       c = tokstart[++namelen])
    ;

  /* The token "if" terminates the expression and is NOT 
     removed from the input stream.  */
  if (namelen == 2 && tokstart[0] == 'i' && tokstart[1] == 'f')
    {
      return 0;
    }

  lexptr += namelen;

  /* Handle the tokens $digits; also $ (short for $0) and $$ (short for $$1)
     and $$digits (equivalent to $<-digits> if you could type that).
     Make token type LAST, and put the number (the digits) in yylval.  */

  tryname:
  if (*tokstart == '$')
    {
      register int negate = 0;
      c = 1;
      /* Double dollar means negate the number and add -1 as well.
	 Thus $$ alone means -1.  */
      if (namelen >= 2 && tokstart[1] == '$')
	{
	  negate = 1;
	  c = 2;
	}
      if (c == namelen)
	{
	  /* Just dollars (one or two) */
	  yylval.lval = - negate;
	  return LAST;
	}
      /* Is the rest of the token digits?  */
      for (; c < namelen; c++)
	if (!(tokstart[c] >= '0' && tokstart[c] <= '9'))
	  break;
      if (c == namelen)
	{
	  yylval.lval = atoi (tokstart + 1 + negate);
	  if (negate)
	    yylval.lval = - yylval.lval;
	  return LAST;
	}
    }

  /* Handle tokens that refer to machine registers:
     $ followed by a register name.  */

  if (*tokstart == '$') {
    for (c = 0; c < NUM_REGS; c++)
      if (namelen - 1 == strlen (reg_names[c])
	  && !strncmp (tokstart + 1, reg_names[c], namelen - 1))
	{
	  yylval.lval = c;
	  return REGNAME;
	}
    for (c = 0; c < num_std_regs; c++)
     if (namelen - 1 == strlen (std_regs[c].name)
	 && !strncmp (tokstart + 1, std_regs[c].name, namelen - 1))
       {
	 yylval.lval = std_regs[c].regnum;
	 return REGNAME;
       }
  }
  /* Catch specific keywords.  Should be done with a data structure.  */
  switch (namelen)
    {
    case 8:
      if (!strncmp (tokstart, "unsigned", 8))
	return UNSIGNED;
      if (current_language->la_language == language_cplus
	  && !strncmp (tokstart, "template", 8))
	return TEMPLATE;
      if (!strncmp (tokstart, "volatile", 8))
	return VOLATILE_KEYWORD;
      break;
    case 6:
      if (!strncmp (tokstart, "struct", 6))
	return STRUCT;
      if (!strncmp (tokstart, "signed", 6))
	return SIGNED_KEYWORD;
      if (!strncmp (tokstart, "sizeof", 6))      
	return SIZEOF;
      break;
    case 5:
      if (current_language->la_language == language_cplus
	  && !strncmp (tokstart, "class", 5))
	return CLASS;
      if (!strncmp (tokstart, "union", 5))
	return UNION;
      if (!strncmp (tokstart, "short", 5))
	return SHORT;
      if (!strncmp (tokstart, "const", 5))
	return CONST_KEYWORD;
      break;
    case 4:
      if (!strncmp (tokstart, "enum", 4))
	return ENUM;
      if (!strncmp (tokstart, "long", 4))
	return LONG;
      if (current_language->la_language == language_cplus
	  && !strncmp (tokstart, "this", 4))
	{
	  static const char this_name[] =
				 { CPLUS_MARKER, 't', 'h', 'i', 's', '\0' };

	  if (lookup_symbol (this_name, expression_context_block,
			     VAR_NAMESPACE, 0, NULL))
	    return THIS;
	}
      break;
    case 3:
      if (!strncmp (tokstart, "int", 3))
	return INT_KEYWORD;
      break;
    default:
      break;
    }

  yylval.sval.ptr = tokstart;
  yylval.sval.length = namelen;

  /* Any other names starting in $ are debugger internal variables.  */

  if (*tokstart == '$')
    {
      yylval.ivar =  lookup_internalvar (copy_name (yylval.sval) + 1);
      return VARIABLE;
    }

  /* Use token-type BLOCKNAME for symbols that happen to be defined as
     functions or symtabs.  If this is not so, then ...
     Use token-type TYPENAME for symbols that happen to be defined
     currently as names of types; NAME for other symbols.
     The caller is not constrained to care about the distinction.  */
  {
    char *tmp = copy_name (yylval.sval);
    struct symbol *sym;
    int is_a_field_of_this = 0;
    int hextype;

    sym = lookup_symbol (tmp, expression_context_block,
			 VAR_NAMESPACE,
			 current_language->la_language == language_cplus
			 ? &is_a_field_of_this : NULL,
			 NULL);
    if ((sym && SYMBOL_CLASS (sym) == LOC_BLOCK) ||
        lookup_partial_symtab (tmp))
      {
	yylval.ssym.sym = sym;
	yylval.ssym.is_a_field_of_this = is_a_field_of_this;
	return BLOCKNAME;
      }
    if (sym && SYMBOL_CLASS (sym) == LOC_TYPEDEF)
        {
	  yylval.tsym.type = SYMBOL_TYPE (sym);
	  return TYPENAME;
        }
    if ((yylval.tsym.type = lookup_primitive_typename (tmp)) != 0)
	return TYPENAME;

    /* Input names that aren't symbols but ARE valid hex numbers,
       when the input radix permits them, can be names or numbers
       depending on the parse.  Note we support radixes > 16 here.  */
    if (!sym && 
        ((tokstart[0] >= 'a' && tokstart[0] < 'a' + input_radix - 10) ||
         (tokstart[0] >= 'A' && tokstart[0] < 'A' + input_radix - 10)))
      {
 	YYSTYPE newlval;	/* Its value is ignored.  */
	hextype = parse_number (tokstart, namelen, 0, &newlval);
	if (hextype == INT)
	  {
	    yylval.ssym.sym = sym;
	    yylval.ssym.is_a_field_of_this = is_a_field_of_this;
	    return NAME_OR_INT;
	  }
	if (hextype == UINT)
	  {
	    yylval.ssym.sym = sym;
	    yylval.ssym.is_a_field_of_this = is_a_field_of_this;
	    return NAME_OR_UINT;
	  }
      }

    /* Any other kind of symbol */
    yylval.ssym.sym = sym;
    yylval.ssym.is_a_field_of_this = is_a_field_of_this;
    return NAME;
  }
}

void
yyerror (msg)
     char *msg;
{
  error (msg ? msg : "Invalid syntax in expression.");
}

/* Table mapping opcodes into strings for printing operators
   and precedences of the operators.  */

const static struct op_print c_op_print_tab[] =
  {
    {",",  BINOP_COMMA, PREC_COMMA, 0},
    {"=",  BINOP_ASSIGN, PREC_ASSIGN, 1},
    {"||", BINOP_OR, PREC_OR, 0},
    {"&&", BINOP_AND, PREC_AND, 0},
    {"|",  BINOP_LOGIOR, PREC_LOGIOR, 0},
    {"&",  BINOP_LOGAND, PREC_LOGAND, 0},
    {"^",  BINOP_LOGXOR, PREC_LOGXOR, 0},
    {"==", BINOP_EQUAL, PREC_EQUAL, 0},
    {"!=", BINOP_NOTEQUAL, PREC_EQUAL, 0},
    {"<=", BINOP_LEQ, PREC_ORDER, 0},
    {">=", BINOP_GEQ, PREC_ORDER, 0},
    {">",  BINOP_GTR, PREC_ORDER, 0},
    {"<",  BINOP_LESS, PREC_ORDER, 0},
    {">>", BINOP_RSH, PREC_SHIFT, 0},
    {"<<", BINOP_LSH, PREC_SHIFT, 0},
    {"+",  BINOP_ADD, PREC_ADD, 0},
    {"-",  BINOP_SUB, PREC_ADD, 0},
    {"*",  BINOP_MUL, PREC_MUL, 0},
    {"/",  BINOP_DIV, PREC_MUL, 0},
    {"%",  BINOP_REM, PREC_MUL, 0},
    {"@",  BINOP_REPEAT, PREC_REPEAT, 0},
    {"-",  UNOP_NEG, PREC_PREFIX, 0},
    {"!",  UNOP_ZEROP, PREC_PREFIX, 0},
    {"~",  UNOP_LOGNOT, PREC_PREFIX, 0},
    {"*",  UNOP_IND, PREC_PREFIX, 0},
    {"&",  UNOP_ADDR, PREC_PREFIX, 0},
    {"sizeof ", UNOP_SIZEOF, PREC_PREFIX, 0},
    {"++", UNOP_PREINCREMENT, PREC_PREFIX, 0},
    {"--", UNOP_PREDECREMENT, PREC_PREFIX, 0},
    /* C++  */
    {"::", BINOP_SCOPE, PREC_PREFIX, 0},
};

/* These variables point to the objects
   representing the predefined C data types.  */

struct type *builtin_type_void;
struct type *builtin_type_char;
struct type *builtin_type_short;
struct type *builtin_type_int;
struct type *builtin_type_long;
struct type *builtin_type_long_long;
struct type *builtin_type_signed_char;
struct type *builtin_type_unsigned_char;
struct type *builtin_type_unsigned_short;
struct type *builtin_type_unsigned_int;
struct type *builtin_type_unsigned_long;
struct type *builtin_type_unsigned_long_long;
struct type *builtin_type_float;
struct type *builtin_type_double;
struct type *builtin_type_long_double;
struct type *builtin_type_complex;
struct type *builtin_type_double_complex;

struct type ** const (c_builtin_types[]) = 
{
  &builtin_type_int,
  &builtin_type_long,
  &builtin_type_short,
  &builtin_type_char,
  &builtin_type_float,
  &builtin_type_double,
  &builtin_type_void,
  &builtin_type_long_long,
  &builtin_type_signed_char,
  &builtin_type_unsigned_char,
  &builtin_type_unsigned_short,
  &builtin_type_unsigned_int,
  &builtin_type_unsigned_long,
  &builtin_type_unsigned_long_long,
  &builtin_type_long_double,
  &builtin_type_complex,
  &builtin_type_double_complex,
  0
};

const struct language_defn c_language_defn = {
  "c",				/* Language name */
  language_c,
  c_builtin_types,
  range_check_off,
  type_check_off,
  c_parse,
  c_error,
  &BUILTIN_TYPE_LONGEST,	 /* longest signed   integral type */
  &BUILTIN_TYPE_UNSIGNED_LONGEST,/* longest unsigned integral type */
  &builtin_type_double,		/* longest floating point type */ /*FIXME*/
  "0x%x", "0x%", "x",		/* Hex   format, prefix, suffix */
  "0%o",  "0%",  "o",		/* Octal format, prefix, suffix */
  c_op_print_tab,		/* expression operators for printing */
  LANG_MAGIC
};

const struct language_defn cplus_language_defn = {
  "c++",				/* Language name */
  language_cplus,
  c_builtin_types,
  range_check_off,
  type_check_off,
  c_parse,
  c_error,
  &BUILTIN_TYPE_LONGEST,	 /* longest signed   integral type */
  &BUILTIN_TYPE_UNSIGNED_LONGEST,/* longest unsigned integral type */
  &builtin_type_double,		/* longest floating point type */ /*FIXME*/
  "0x%x", "0x%", "x",		/* Hex   format, prefix, suffix */
  "0%o",  "0%",  "o",		/* Octal format, prefix, suffix */
  c_op_print_tab,		/* expression operators for printing */
  LANG_MAGIC
};

void
_initialize_c_exp ()
{
  builtin_type_void =
    init_type (TYPE_CODE_VOID, 1,
	       0,
	       "void", (struct objfile *) NULL);
  builtin_type_char =
    init_type (TYPE_CODE_INT, TARGET_CHAR_BIT / TARGET_CHAR_BIT,
	       0,
	       "char", (struct objfile *) NULL);
  builtin_type_signed_char =
    init_type (TYPE_CODE_INT, TARGET_CHAR_BIT / TARGET_CHAR_BIT,
	       TYPE_FLAG_SIGNED,
	       "signed char", (struct objfile *) NULL);
  builtin_type_unsigned_char =
    init_type (TYPE_CODE_INT, TARGET_CHAR_BIT / TARGET_CHAR_BIT,
	       TYPE_FLAG_UNSIGNED,
	       "unsigned char", (struct objfile *) NULL);
  builtin_type_short =
    init_type (TYPE_CODE_INT, TARGET_SHORT_BIT / TARGET_CHAR_BIT,
	       0,
	       "short", (struct objfile *) NULL);
  builtin_type_unsigned_short =
    init_type (TYPE_CODE_INT, TARGET_SHORT_BIT / TARGET_CHAR_BIT,
	       TYPE_FLAG_UNSIGNED,
	       "unsigned short", (struct objfile *) NULL);
  builtin_type_int =
    init_type (TYPE_CODE_INT, TARGET_INT_BIT / TARGET_CHAR_BIT,
	       0,
	       "int", (struct objfile *) NULL);
  builtin_type_unsigned_int =
    init_type (TYPE_CODE_INT, TARGET_INT_BIT / TARGET_CHAR_BIT,
	       TYPE_FLAG_UNSIGNED,
	       "unsigned int", (struct objfile *) NULL);
  builtin_type_long =
    init_type (TYPE_CODE_INT, TARGET_LONG_BIT / TARGET_CHAR_BIT,
	       0,
	       "long", (struct objfile *) NULL);
  builtin_type_unsigned_long =
    init_type (TYPE_CODE_INT, TARGET_LONG_BIT / TARGET_CHAR_BIT,
	       TYPE_FLAG_UNSIGNED,
	       "unsigned long", (struct objfile *) NULL);
  builtin_type_long_long =
    init_type (TYPE_CODE_INT, TARGET_LONG_LONG_BIT / TARGET_CHAR_BIT,
	       0,
	       "long long", (struct objfile *) NULL);
  builtin_type_unsigned_long_long = 
    init_type (TYPE_CODE_INT, TARGET_LONG_LONG_BIT / TARGET_CHAR_BIT,
	       TYPE_FLAG_UNSIGNED,
	       "unsigned long long", (struct objfile *) NULL);
  builtin_type_float =
    init_type (TYPE_CODE_FLT, TARGET_FLOAT_BIT / TARGET_CHAR_BIT,
	       0,
	       "float", (struct objfile *) NULL);
  builtin_type_double =
    init_type (TYPE_CODE_FLT, TARGET_DOUBLE_BIT / TARGET_CHAR_BIT,
	       0,
	       "double", (struct objfile *) NULL);
  builtin_type_long_double =
    init_type (TYPE_CODE_FLT, TARGET_LONG_DOUBLE_BIT / TARGET_CHAR_BIT,
	       0,
	       "long double", (struct objfile *) NULL);
  builtin_type_complex =
    init_type (TYPE_CODE_FLT, TARGET_COMPLEX_BIT / TARGET_CHAR_BIT,
	       0,
	       "complex", (struct objfile *) NULL);
  builtin_type_double_complex =
    init_type (TYPE_CODE_FLT, TARGET_DOUBLE_COMPLEX_BIT / TARGET_CHAR_BIT,
	       0,
	       "double complex", (struct objfile *) NULL);

  add_language (&c_language_defn);
  add_language (&cplus_language_defn);
}