lex.c

LCC4.2 C编译器源码

			}
			if (rcp[0] == 'n'
			&&  rcp[1] == 'u'
			&&  rcp[2] == 'm'
			&& !(map[rcp[3]]&(DIGIT|LETTER))) {
				cp = rcp + 3;
				return ENUM;
			}
			if (rcp[0] == 'x'
			&&  rcp[1] == 't'
			&&  rcp[2] == 'e'
			&&  rcp[3] == 'r'
			&&  rcp[4] == 'n'
			&& !(map[rcp[5]]&(DIGIT|LETTER))) {
				cp = rcp + 5;
				return EXTERN;
			}
			goto id;
		case 'f':
			if (rcp[0] == 'l'
			&&  rcp[1] == 'o'
			&&  rcp[2] == 'a'
			&&  rcp[3] == 't'
			&& !(map[rcp[4]]&(DIGIT|LETTER))) {
				cp = rcp + 4;
				tsym = floattype->u.sym;
				return FLOAT;
			}
			if (rcp[0] == 'o'
			&&  rcp[1] == 'r'
			&& !(map[rcp[2]]&(DIGIT|LETTER))) {
				cp = rcp + 2;
				return FOR;
			}
			goto id;
		case 'g':
			if (rcp[0] == 'o'
			&&  rcp[1] == 't'
			&&  rcp[2] == 'o'
			&& !(map[rcp[3]]&(DIGIT|LETTER))) {
				cp = rcp + 3;
				return GOTO;
			}
			goto id;
		case 'l':
			if (rcp[0] == 'o'
			&&  rcp[1] == 'n'
			&&  rcp[2] == 'g'
			&& !(map[rcp[3]]&(DIGIT|LETTER))) {
				cp = rcp + 3;
				return LONG;
			}
			goto id;
		case 'r':
			if (rcp[0] == 'e'
			&&  rcp[1] == 'g'
			&&  rcp[2] == 'i'
			&&  rcp[3] == 's'
			&&  rcp[4] == 't'
			&&  rcp[5] == 'e'
			&&  rcp[6] == 'r'
			&& !(map[rcp[7]]&(DIGIT|LETTER))) {
				cp = rcp + 7;
				return REGISTER;
			}
			if (rcp[0] == 'e'
			&&  rcp[1] == 't'
			&&  rcp[2] == 'u'
			&&  rcp[3] == 'r'
			&&  rcp[4] == 'n'
			&& !(map[rcp[5]]&(DIGIT|LETTER))) {
				cp = rcp + 5;
				return RETURN;
			}
			goto id;
		case 's':
			if (rcp[0] == 'h'
			&&  rcp[1] == 'o'
			&&  rcp[2] == 'r'
			&&  rcp[3] == 't'
			&& !(map[rcp[4]]&(DIGIT|LETTER))) {
				cp = rcp + 4;
				return SHORT;
			}
			if (rcp[0] == 'i'
			&&  rcp[1] == 'g'
			&&  rcp[2] == 'n'
			&&  rcp[3] == 'e'
			&&  rcp[4] == 'd'
			&& !(map[rcp[5]]&(DIGIT|LETTER))) {
				cp = rcp + 5;
				return SIGNED;
			}
			if (rcp[0] == 'i'
			&&  rcp[1] == 'z'
			&&  rcp[2] == 'e'
			&&  rcp[3] == 'o'
			&&  rcp[4] == 'f'
			&& !(map[rcp[5]]&(DIGIT|LETTER))) {
				cp = rcp + 5;
				return SIZEOF;
			}
			if (rcp[0] == 't'
			&&  rcp[1] == 'a'
			&&  rcp[2] == 't'
			&&  rcp[3] == 'i'
			&&  rcp[4] == 'c'
			&& !(map[rcp[5]]&(DIGIT|LETTER))) {
				cp = rcp + 5;
				return STATIC;
			}
			if (rcp[0] == 't'
			&&  rcp[1] == 'r'
			&&  rcp[2] == 'u'
			&&  rcp[3] == 'c'
			&&  rcp[4] == 't'
			&& !(map[rcp[5]]&(DIGIT|LETTER))) {
				cp = rcp + 5;
				return STRUCT;
			}
			if (rcp[0] == 'w'
			&&  rcp[1] == 'i'
			&&  rcp[2] == 't'
			&&  rcp[3] == 'c'
			&&  rcp[4] == 'h'
			&& !(map[rcp[5]]&(DIGIT|LETTER))) {
				cp = rcp + 5;
				return SWITCH;
			}
			goto id;
		case 't':
			if (rcp[0] == 'y'
			&&  rcp[1] == 'p'
			&&  rcp[2] == 'e'
			&&  rcp[3] == 'd'
			&&  rcp[4] == 'e'
			&&  rcp[5] == 'f'
			&& !(map[rcp[6]]&(DIGIT|LETTER))) {
				cp = rcp + 6;
				return TYPEDEF;
			}
			goto id;
		case 'u':
			if (rcp[0] == 'n'
			&&  rcp[1] == 'i'
			&&  rcp[2] == 'o'
			&&  rcp[3] == 'n'
			&& !(map[rcp[4]]&(DIGIT|LETTER))) {
				cp = rcp + 4;
				return UNION;
			}
			if (rcp[0] == 'n'
			&&  rcp[1] == 's'
			&&  rcp[2] == 'i'
			&&  rcp[3] == 'g'
			&&  rcp[4] == 'n'
			&&  rcp[5] == 'e'
			&&  rcp[6] == 'd'
			&& !(map[rcp[7]]&(DIGIT|LETTER))) {
				cp = rcp + 7;
				return UNSIGNED;
			}
			goto id;
		case 'v':
			if (rcp[0] == 'o'
			&&  rcp[1] == 'i'
			&&  rcp[2] == 'd'
			&& !(map[rcp[3]]&(DIGIT|LETTER))) {
				cp = rcp + 3;
				tsym = voidtype->u.sym;
				return VOID;
			}
			if (rcp[0] == 'o'
			&&  rcp[1] == 'l'
			&&  rcp[2] == 'a'
			&&  rcp[3] == 't'
			&&  rcp[4] == 'i'
			&&  rcp[5] == 'l'
			&&  rcp[6] == 'e'
			&& !(map[rcp[7]]&(DIGIT|LETTER))) {
				cp = rcp + 7;
				return VOLATILE;
			}
			goto id;
		case 'w':
			if (rcp[0] == 'h'
			&&  rcp[1] == 'i'
			&&  rcp[2] == 'l'
			&&  rcp[3] == 'e'
			&& !(map[rcp[4]]&(DIGIT|LETTER))) {
				cp = rcp + 4;
				return WHILE;
			}
			goto id;
		case '_':
			if (rcp[0] == '_'
			&&  rcp[1] == 't'
			&&  rcp[2] == 'y'
			&&  rcp[3] == 'p'
			&&  rcp[4] == 'e'
			&&  rcp[5] == 'c'
			&&  rcp[6] == 'o'
			&&  rcp[7] == 'd'
			&&  rcp[8] == 'e'
			&& !(map[rcp[9]]&(DIGIT|LETTER))) {
				cp = rcp + 9;
				return TYPECODE;
			}
			if (rcp[0] == '_'
			&&  rcp[1] == 'f'
			&&  rcp[2] == 'i'
			&&  rcp[3] == 'r'
			&&  rcp[4] == 's'
			&&  rcp[5] == 't'
			&&  rcp[6] == 'a'
			&&  rcp[7] == 'r'
			&&  rcp[8] == 'g'
			&& !(map[rcp[9]]&(DIGIT|LETTER))) {
				cp = rcp + 9;
				return FIRSTARG;
			}
			goto id;
		default:
			if ((map[cp[-1]]&BLANK) == 0)
				if (cp[-1] < ' ' || cp[-1] >= 0177)
					error("illegal character `\\0%o'\n", cp[-1]);
				else
					error("illegal character `%c'\n", cp[-1]);
		}
	}
}
static Symbol icon(unsigned long n, int overflow, int base) {
	if ((*cp=='u'||*cp=='U') && (cp[1]=='l'||cp[1]=='L')
	||  (*cp=='l'||*cp=='L') && (cp[1]=='u'||cp[1]=='U')) {
		tval.type = unsignedlong;
		cp += 2;
	} else if (*cp == 'u' || *cp == 'U') {
		if (overflow || n > unsignedtype->u.sym->u.limits.max.i)
			tval.type = unsignedlong;
		else
			tval.type = unsignedtype;
		cp += 1;
	} else if (*cp == 'l' || *cp == 'L') {
		if (overflow || n > longtype->u.sym->u.limits.max.i)
			tval.type = unsignedlong;
		else
			tval.type = longtype;
		cp += 1;
	} else if (overflow || n > longtype->u.sym->u.limits.max.i)
		tval.type = unsignedlong;
	else if (n > inttype->u.sym->u.limits.max.i)
		tval.type = longtype;
	else if (base != 10 && n > inttype->u.sym->u.limits.max.i)
		tval.type = unsignedtype;
	else
		tval.type = inttype;
	switch (tval.type->op) {
	case INT:
		if (overflow || n > tval.type->u.sym->u.limits.max.i) {
			warning("overflow in constant `%S'\n", token,
				(char*)cp - token);
			tval.u.c.v.i = tval.type->u.sym->u.limits.max.i;
		} else
			tval.u.c.v.i = n;
		break;
	case UNSIGNED:
		if (overflow || n > tval.type->u.sym->u.limits.max.u) {
			warning("overflow in constant `%S'\n", token,
				(char*)cp - token);
			tval.u.c.v.u = tval.type->u.sym->u.limits.max.u;
		} else
			tval.u.c.v.u = n;
		break;
	default: assert(0);
	}
	ppnumber("integer");
	return &tval;
}
static void ppnumber(char *which) {
	unsigned char *rcp = cp--;

	for ( ; (map[*cp]&(DIGIT|LETTER)) || *cp == '.'; cp++)
		if ((cp[0] == 'E' || cp[0] == 'e')
		&&  (cp[1] == '-' || cp[1] == '+'))
			cp++;
	if (cp > rcp)
		error("`%S' is a preprocessing number but an invalid %s constant\n", token,

			(char*)cp-token, which);
}
static Symbol fcon(void) {
	if (*cp == '.')
		do
			cp++;
		while (map[*cp]&DIGIT);
	if (*cp == 'e' || *cp == 'E') {
		if (*++cp == '-' || *cp == '+')
			cp++;
		if (map[*cp]&DIGIT)
			do
				cp++;
			while (map[*cp]&DIGIT);
		else
			error("invalid floating constant `%S'\n", token,
				(char*)cp - token);
	}

	errno = 0;
	tval.u.c.v.d = strtod(token, NULL);
	if (errno == ERANGE)
		warning("overflow in floating constant `%S'\n", token,
			(char*)cp - token);
	if (*cp == 'f' || *cp == 'F') {
		++cp;
		if (tval.u.c.v.d > floattype->u.sym->u.limits.max.d)
			warning("overflow in floating constant `%S'\n", token,
				(char*)cp - token);
		tval.type = floattype;
	} else if (*cp == 'l' || *cp == 'L') {
		cp++;
		tval.type = longdouble;
	} else {
		if (tval.u.c.v.d > doubletype->u.sym->u.limits.max.d)
			warning("overflow in floating constant `%S'\n", token,
				(char*)cp - token);
		tval.type = doubletype;
	}
	ppnumber("floating");
	return &tval;
}

static void *cput(int c, void *cl) {
	char *s = cl;

	if (c < 0 || c > 255)
		warning("overflow in escape sequence with resulting value `%d'\n", c);
	*s++ = c;
	return s;
}

static void *wcput(int c, void *cl) {
	unsigned int *s = cl;

	*s++ = c;
	return s;
}

static void *scon(int q, void *put(int c, void *cl), void *cl) {
	int n = 0, nbad = 0;

	do {
		cp++;
		while (*cp != q) {
			int c;
			if (map[*cp]&NEWLINE) {
				if (cp < limit)
					break;
				cp++;
				nextline();
				if (cp == limit)
					break;
				continue;
			}
			c = *cp++;
			if (c == '\\') {
				if (map[*cp]&NEWLINE) {
					if (cp++ < limit)
						continue;
					nextline();
				}
				if (limit - cp < MAXTOKEN)
					fillbuf();
				c = backslash(q);
			} else if (c < 0 || c > 255 || map[c] == 0)
				nbad++;
			if (n++ < BUFSIZE)
				cl = put(c, cl);
		}
		if (*cp == q)
			cp++;
		else
			error("missing %c\n", q);
		if (q == '"' && put == wcput && getchr() == 'L') {
			if (limit - cp < 2)
				fillbuf();
			if (cp[1] == '"')
				cp++;
		}
	} while (q == '"' && getchr() == '"');
	cl = put(0, cl);
	if (n >= BUFSIZE)
		error("%s literal too long\n", q == '"' ? "string" : "character");
	if (Aflag >= 2 && q == '"' && n > 509)
		warning("more than 509 characters in a string literal\n");
	if (Aflag >= 2 && nbad > 0)
		warning("%s literal contains non-portable characters\n",
			q == '"' ? "string" : "character");
	return cl;
}
int getchr(void) {
	for (;;) {
		while (map[*cp]&BLANK)
			cp++;
		if (!(map[*cp]&NEWLINE))
			return *cp;
		cp++;
		nextline();
		if (cp == limit)
			return EOI;
	}
}
static int backslash(int q) {
	unsigned int c;

	switch (*cp++) {
	case 'a': return 7;
	case 'b': return '\b';
	case 'f': return '\f';
	case 'n': return '\n';
	case 'r': return '\r';
	case 't': return '\t';
	case 'v': return '\v';
	case '\'': case '"': case '\\': case '\?': break;
	case 'x': {
		int overflow = 0;
		if ((map[*cp]&(DIGIT|HEX)) == 0) {
			if (*cp < ' ' || *cp == 0177)
				error("ill-formed hexadecimal escape sequence\n");
			else
				error("ill-formed hexadecimal escape sequence `\\x%c'\n", *cp);
			if (*cp != q)
				cp++;
			return 0;
		}
		for (c = 0; map[*cp]&(DIGIT|HEX); cp++) {
			if (c >> (8*widechar->size - 4))
				overflow = 1;
			if (map[*cp]&DIGIT)
				c = (c<<4) + *cp - '0';
			else
				c = (c<<4) + (*cp&~040) - 'A' + 10;
		}
		if (overflow)
			warning("overflow in hexadecimal escape sequence\n");
		return c&ones(8*widechar->size);
		}
	case '0': case '1': case '2': case '3':
	case '4': case '5': case '6': case '7':
		c = *(cp-1) - '0';
		if (*cp >= '0' && *cp <= '7') {
			c = (c<<3) + *cp++ - '0';
			if (*cp >= '0' && *cp <= '7')
				c = (c<<3) + *cp++ - '0';
		}
		return c;
	default:
		if (cp[-1] < ' ' || cp[-1] >= 0177)
			warning("unrecognized character escape sequence\n");
		else
			warning("unrecognized character escape sequence `\\%c'\n", cp[-1]);
	}
	return cp[-1];
}