cexp.y

GUN开源阻止下的编译器GCC

/* Parse C expressions for CCCP.
   Copyright (C) 1987, 1992, 1994, 1995 Free Software Foundation.

This program is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
Free Software Foundation; either version 2, or (at your option) any
later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA.

 In other words, you are welcome to use, share and improve this program.
 You are forbidden to forbid anyone else to use, share and improve
 what you give them.   Help stamp out software-hoarding!

 Adapted from expread.y of GDB by Paul Rubin, July 1986.  */

/* Parse a C expression from text in a string  */
   
%{
#include "config.h"
#include <setjmp.h>
/* #define YYDEBUG 1 */

#ifdef MULTIBYTE_CHARS
#include <stdlib.h>
#include <locale.h>
#endif

#include <stdio.h>

typedef unsigned char U_CHAR;

/* This is used for communicating lists of keywords with cccp.c.  */
struct arglist {
  struct arglist *next;
  U_CHAR *name;
  int length;
  int argno;
};

/* Define a generic NULL if one hasn't already been defined.  */

#ifndef NULL
#define NULL 0
#endif

#ifndef GENERIC_PTR
#if defined (USE_PROTOTYPES) ? USE_PROTOTYPES : defined (__STDC__)
#define GENERIC_PTR void *
#else
#define GENERIC_PTR char *
#endif
#endif

/* Find the largest host integer type and set its size and type.  */

#ifndef HOST_BITS_PER_WIDE_INT

#if HOST_BITS_PER_LONG > HOST_BITS_PER_INT
#define HOST_BITS_PER_WIDE_INT HOST_BITS_PER_LONG
#define HOST_WIDE_INT long
#else
#define HOST_BITS_PER_WIDE_INT HOST_BITS_PER_INT
#define HOST_WIDE_INT int
#endif

#endif

#ifndef NULL_PTR
#define NULL_PTR ((GENERIC_PTR)0)
#endif

int yylex ();
void yyerror ();
HOST_WIDE_INT expression_value;

static jmp_buf parse_return_error;

/* Nonzero means count most punctuation as part of a name.  */
static int keyword_parsing = 0;

/* Nonzero means do not evaluate this expression.
   This is a count, since unevaluated expressions can nest.  */
static int skip_evaluation;

/* some external tables of character types */
extern unsigned char is_idstart[], is_idchar[], is_hor_space[];

extern char *xmalloc ();

/* Flag for -pedantic.  */
extern int pedantic;

/* Flag for -traditional.  */
extern int traditional;

#ifndef CHAR_TYPE_SIZE
#define CHAR_TYPE_SIZE BITS_PER_UNIT
#endif

#ifndef INT_TYPE_SIZE
#define INT_TYPE_SIZE BITS_PER_WORD
#endif

#ifndef LONG_TYPE_SIZE
#define LONG_TYPE_SIZE BITS_PER_WORD
#endif

#ifndef WCHAR_TYPE_SIZE
#define WCHAR_TYPE_SIZE INT_TYPE_SIZE
#endif

#ifndef MAX_CHAR_TYPE_SIZE
#define MAX_CHAR_TYPE_SIZE CHAR_TYPE_SIZE
#endif

#ifndef MAX_INT_TYPE_SIZE
#define MAX_INT_TYPE_SIZE INT_TYPE_SIZE
#endif

#ifndef MAX_LONG_TYPE_SIZE
#define MAX_LONG_TYPE_SIZE LONG_TYPE_SIZE
#endif

#ifndef MAX_WCHAR_TYPE_SIZE
#define MAX_WCHAR_TYPE_SIZE WCHAR_TYPE_SIZE
#endif

/* Yield nonzero if adding two numbers with A's and B's signs can yield a
   number with SUM's sign, where A, B, and SUM are all C integers.  */
#define possible_sum_sign(a, b, sum) ((((a) ^ (b)) | ~ ((a) ^ (sum))) < 0)

static void integer_overflow ();
static long left_shift ();
static long right_shift ();
%}

%union {
  struct constant {long value; int unsignedp;} integer;
  struct name {U_CHAR *address; int length;} name;
  struct arglist *keywords;
}

%type <integer> exp exp1 start
%type <keywords> keywords
%token <integer> INT CHAR
%token <name> NAME
%token <integer> ERROR

%right '?' ':'
%left ','
%left OR
%left AND
%left '|'
%left '^'
%left '&'
%left EQUAL NOTEQUAL
%left '<' '>' LEQ GEQ
%left LSH RSH
%left '+' '-'
%left '*' '/' '%'
%right UNARY

/* %expect 40 */

%%

start   :	exp1
		{ expression_value = $1.value; }
	;

/* Expressions, including the comma operator.  */
exp1	:	exp
	|	exp1 ',' exp
			{ if (pedantic)
			    pedwarn ("comma operator in operand of `#if'");
			  $$ = $3; }
	;

/* Expressions, not including the comma operator.  */
exp	:	'-' exp    %prec UNARY
			{ $$.value = - $2.value;
			  if (($$.value & $2.value) < 0 && ! $2.unsignedp)
			    integer_overflow ();
			  $$.unsignedp = $2.unsignedp; }
	|	'!' exp    %prec UNARY
			{ $$.value = ! $2.value;
			  $$.unsignedp = 0; }
	|	'+' exp    %prec UNARY
			{ $$ = $2; }
	|	'~' exp    %prec UNARY
			{ $$.value = ~ $2.value;
			  $$.unsignedp = $2.unsignedp; }
	|	'#' NAME
  			{ $$.value = check_assertion ($2.address, $2.length,
						      0, NULL_PTR);
			  $$.unsignedp = 0; }
	|	'#' NAME
			{ keyword_parsing = 1; }
		'(' keywords ')'
  			{ $$.value = check_assertion ($2.address, $2.length,
						      1, $5);
			  keyword_parsing = 0;
			  $$.unsignedp = 0; }
	|	'(' exp1 ')'
			{ $$ = $2; }
	;

/* Binary operators in order of decreasing precedence.  */
exp	:	exp '*' exp
			{ $$.unsignedp = $1.unsignedp || $3.unsignedp;
			  if ($$.unsignedp)
			    $$.value = (unsigned long) $1.value * $3.value;
			  else
			    {
			      $$.value = $1.value * $3.value;
			      if ($1.value
				  && ($$.value / $1.value != $3.value
				      || ($$.value & $1.value & $3.value) < 0))
				integer_overflow ();
			    } }
	|	exp '/' exp
			{ if ($3.value == 0)
			    {
			      if (!skip_evaluation)
				error ("division by zero in #if");
			      $3.value = 1;
			    }
			  $$.unsignedp = $1.unsignedp || $3.unsignedp;
			  if ($$.unsignedp)
			    $$.value = (unsigned long) $1.value / $3.value;
			  else
			    {
			      $$.value = $1.value / $3.value;
			      if (($$.value & $1.value & $3.value) < 0)
				integer_overflow ();
			    } }
	|	exp '%' exp
			{ if ($3.value == 0)
			    {
			      if (!skip_evaluation)
				error ("division by zero in #if");
			      $3.value = 1;
			    }
			  $$.unsignedp = $1.unsignedp || $3.unsignedp;
			  if ($$.unsignedp)
			    $$.value = (unsigned long) $1.value % $3.value;
			  else
			    $$.value = $1.value % $3.value; }
	|	exp '+' exp
			{ $$.value = $1.value + $3.value;
			  $$.unsignedp = $1.unsignedp || $3.unsignedp;
			  if (! $$.unsignedp
			      && ! possible_sum_sign ($1.value, $3.value,
						      $$.value))
			    integer_overflow (); }
	|	exp '-' exp
			{ $$.value = $1.value - $3.value;
			  $$.unsignedp = $1.unsignedp || $3.unsignedp;
			  if (! $$.unsignedp
			      && ! possible_sum_sign ($$.value, $3.value,
						      $1.value))
			    integer_overflow (); }
	|	exp LSH exp
			{ $$.unsignedp = $1.unsignedp;
			  if ($3.value < 0 && ! $3.unsignedp)
			    $$.value = right_shift (&$1, -$3.value);
			  else
			    $$.value = left_shift (&$1, $3.value); }
	|	exp RSH exp
			{ $$.unsignedp = $1.unsignedp;
			  if ($3.value < 0 && ! $3.unsignedp)
			    $$.value = left_shift (&$1, -$3.value);
			  else
			    $$.value = right_shift (&$1, $3.value); }
	|	exp EQUAL exp
			{ $$.value = ($1.value == $3.value);
			  $$.unsignedp = 0; }
	|	exp NOTEQUAL exp
			{ $$.value = ($1.value != $3.value);
			  $$.unsignedp = 0; }
	|	exp LEQ exp
			{ $$.unsignedp = 0;
			  if ($1.unsignedp || $3.unsignedp)
			    $$.value = (unsigned long) $1.value <= $3.value;
			  else
			    $$.value = $1.value <= $3.value; }
	|	exp GEQ exp
			{ $$.unsignedp = 0;
			  if ($1.unsignedp || $3.unsignedp)
			    $$.value = (unsigned long) $1.value >= $3.value;
			  else
			    $$.value = $1.value >= $3.value; }
	|	exp '<' exp
			{ $$.unsignedp = 0;
			  if ($1.unsignedp || $3.unsignedp)
			    $$.value = (unsigned long) $1.value < $3.value;
			  else
			    $$.value = $1.value < $3.value; }
	|	exp '>' exp
			{ $$.unsignedp = 0;
			  if ($1.unsignedp || $3.unsignedp)
			    $$.value = (unsigned long) $1.value > $3.value;
			  else
			    $$.value = $1.value > $3.value; }
	|	exp '&' exp
			{ $$.value = $1.value & $3.value;
			  $$.unsignedp = $1.unsignedp || $3.unsignedp; }
	|	exp '^' exp
			{ $$.value = $1.value ^ $3.value;
			  $$.unsignedp = $1.unsignedp || $3.unsignedp; }
	|	exp '|' exp
			{ $$.value = $1.value | $3.value;
			  $$.unsignedp = $1.unsignedp || $3.unsignedp; }
	|	exp AND
			{ skip_evaluation += !$1.value; }
		exp
			{ skip_evaluation -= !$1.value;
			  $$.value = ($1.value && $4.value);
			  $$.unsignedp = 0; }
	|	exp OR
			{ skip_evaluation += !!$1.value; }
		exp
			{ skip_evaluation -= !!$1.value;
			  $$.value = ($1.value || $4.value);
			  $$.unsignedp = 0; }
	|	exp '?'
			{ skip_evaluation += !$1.value; }
	        exp ':'
			{ skip_evaluation += !!$1.value - !$1.value; }
		exp
			{ skip_evaluation -= !!$1.value;
			  $$.value = $1.value ? $4.value : $7.value;
			  $$.unsignedp = $4.unsignedp || $7.unsignedp; }
	|	INT
			{ $$ = yylval.integer; }
	|	CHAR
			{ $$ = yylval.integer; }
	|	NAME
			{ $$.value = 0;
			  $$.unsignedp = 0; }
	;

keywords :
			{ $$ = 0; } 
	|	'(' keywords ')' keywords
			{ struct arglist *temp;
			  $$ = (struct arglist *) xmalloc (sizeof (struct arglist));
			  $$->next = $2;
			  $$->name = (U_CHAR *) "(";
			  $$->length = 1;
			  temp = $$;
			  while (temp != 0 && temp->next != 0)
			    temp = temp->next;
			  temp->next = (struct arglist *) xmalloc (sizeof (struct arglist));
			  temp->next->next = $4;
			  temp->next->name = (U_CHAR *) ")";
			  temp->next->length = 1; }
	|	NAME keywords
			{ $$ = (struct arglist *) xmalloc (sizeof (struct arglist));
			  $$->name = $1.address;
			  $$->length = $1.length;
			  $$->next = $2; } 
	;
%%

/* During parsing of a C expression, the pointer to the next character
   is in this variable.  */

static char *lexptr;

/* Take care of parsing a number (anything that starts with a digit).
   Set yylval and return the token type; update lexptr.
   LEN is the number of characters in it.  */

/* maybe needs to actually deal with floating point numbers */

int
parse_number (olen)
     int olen;
{
  register char *p = lexptr;
  register int c;
  register unsigned long n = 0, nd, ULONG_MAX_over_base;
  register int base = 10;
  register int len = olen;
  register int overflow = 0;
  register int digit, largest_digit = 0;
  int spec_long = 0;

  for (c = 0; c < len; c++)
    if (p[c] == '.') {
      /* It's a float since it contains a point.  */
      yyerror ("floating point numbers not allowed in #if expressions");
      return ERROR;
    }

  yylval.integer.unsignedp = 0;

  if (len >= 3 && (!strncmp (p, "0x", 2) || !strncmp (p, "0X", 2))) {
    p += 2;
    base = 16;
    len -= 2;
  }
  else if (*p == '0')
    base = 8;

  ULONG_MAX_over_base = (unsigned long) -1 / base;

  for (; len > 0; len--) {
    c = *p++;

    if (c >= '0' && c <= '9')
      digit = c - '0';
    else if (base == 16 && c >= 'a' && c <= 'f')
      digit = c - 'a' + 10;
    else if (base == 16 && c >= 'A' && c <= 'F')
      digit = c - 'A' + 10;
    else {
      /* `l' means long, and `u' means unsigned.  */
      while (1) {
	if (c == 'l' || c == 'L')
	  {
	    if (spec_long)
	      yyerror ("two `l's in integer constant");
	    spec_long = 1;
	  }
	else if (c == 'u' || c == 'U')
	  {
	    if (yylval.integer.unsignedp)
	      yyerror ("two `u's in integer constant");
	    yylval.integer.unsignedp = 1;
	  }
	else
	  break;

	if (--len == 0)
	  break;
	c = *p++;
      }
      /* Don't look for any more digits after the suffixes.  */
      break;
    }
    if (largest_digit < digit)
      largest_digit = digit;
    nd = n * base + digit;
    overflow |= ULONG_MAX_over_base < n | nd < n;
    n = nd;
  }

  if (len != 0) {
    yyerror ("Invalid number in #if expression");
    return ERROR;
  }

  if (base <= largest_digit)
    warning ("integer constant contains digits beyond the radix");

  if (overflow)
    warning ("integer constant out of range");

  /* If too big to be signed, consider it unsigned.  */
  if ((long) n < 0 && ! yylval.integer.unsignedp)
    {
      if (base == 10)
	warning ("integer constant is so large that it is unsigned");
      yylval.integer.unsignedp = 1;
    }

  lexptr = p;
  yylval.integer.value = n;
  return INT;
}

struct token {
  char *operator;
  int token;
};

static struct token tokentab2[] = {
  {"&&", AND},
  {"||", OR},
  {"<<", LSH},
  {">>", RSH},
  {"==", EQUAL},
  {"!=", NOTEQUAL},
  {"<=", LEQ},
  {">=", GEQ},
  {"++", ERROR},
  {"--", ERROR},
  {NULL, ERROR}
};

/* Read one token, getting characters through lexptr.  */

int
yylex ()
{
  register int c;
  register int namelen;
  register unsigned char *tokstart;
  register struct token *toktab;
  int wide_flag;

 retry:

  tokstart = (unsigned char *) lexptr;
  c = *tokstart;
  /* See if it is a special token of length 2.  */
  if (! keyword_parsing)
    for (toktab = tokentab2; toktab->operator != NULL; toktab++)
      if (c == *toktab->operator && tokstart[1] == toktab->operator[1]) {
	lexptr += 2;
	if (toktab->token == ERROR)
	  {
	    char *buf = (char *) alloca (40);