subr_scanf.c

来自「基于组件方式开发操作系统的OSKIT源代码」· C语言 代码 · 共 794 行 · 第 1/2 页

			/* size_t is unsigned, hence this optimisation */
			if (--width > sizeof(buf) - 2)
				width = sizeof(buf) - 2;
			width++;
#endif
			flags |= SIGNOK | NDIGITS | NZDIGITS;
			for (p = buf; width; width--) {
				c = *inp;
				/*
				 * Switch on the character; `goto ok'
				 * if we accept it as a part of number.
				 */
				switch (c) {

				/*
				 * The digit 0 is always legal, but is
				 * special.  For %i conversions, if no
				 * digits (zero or nonzero) have been
				 * scanned (only signs), we will have
				 * base==0.  In that case, we should set
				 * it to 8 and enable 0x prefixing.
				 * Also, if we have not scanned zero digits
				 * before this, do not turn off prefixing
				 * (someone else will turn it off if we
				 * have scanned any nonzero digits).
				 */
				case '0':
					if (base == 0) {
						base = 8;
						flags |= PFXOK;
					}
					if (flags & NZDIGITS)
					    flags &= ~(SIGNOK|NZDIGITS|NDIGITS);
					else
					    flags &= ~(SIGNOK|PFXOK|NDIGITS);
					goto ok;

				/* 1 through 7 always legal */
				case '1': case '2': case '3':
				case '4': case '5': case '6': case '7':
					base = basefix[base];
					flags &= ~(SIGNOK | PFXOK | NDIGITS);
					goto ok;

				/* digits 8 and 9 ok iff decimal or hex */
				case '8': case '9':
					base = basefix[base];
					if (base <= 8)
						break;	/* not legal here */
					flags &= ~(SIGNOK | PFXOK | NDIGITS);
					goto ok;

				/* letters ok iff hex */
				case 'A': case 'B': case 'C':
				case 'D': case 'E': case 'F':
				case 'a': case 'b': case 'c':
				case 'd': case 'e': case 'f':
					/* no need to fix base here */
					if (base <= 10)
						break;	/* not legal here */
					flags &= ~(SIGNOK | PFXOK | NDIGITS);
					goto ok;

				/* sign ok only as first character */
				case '+': case '-':
					if (flags & SIGNOK) {
						flags &= ~SIGNOK;
						goto ok;
					}
					break;

				/* x ok iff flag still set & 2nd char */
				case 'x': case 'X':
					if (flags & PFXOK && p == buf + 1) {
						base = 16;	/* if %i */
						flags &= ~PFXOK;
						goto ok;
					}
					break;
				}

				/*
				 * If we got here, c is not a legal character
				 * for a number.  Stop accumulating digits.
				 */
				break;
		ok:
				/*
				 * c is legal: store it and look at the next.
				 */
				*p++ = c;
				if (--inr > 0)
					inp++;
				else 
					break;		/* end of input */
			}
			/*
			 * If we had only a sign, it is no good; push
			 * back the sign.  If the number ends in `x',
			 * it was [sign] '0' 'x', so push back the x
			 * and treat it as [sign] '0'.
			 */
			if (flags & NDIGITS) {
				if (p > buf) {
					inp--;
					inr++;
				}
				goto match_failure;
			}
			c = ((u_char *)p)[-1];
			if (c == 'x' || c == 'X') {
				--p;
				inp--;
				inr++;
			}
			if ((flags & SUPPRESS) == 0) {
				u_quad_t res;

				*p = 0;
				res = (*ccfn)(buf, (char **)NULL, base);
				if (flags & POINTER)
					*va_arg(ap, void **) =
						(void *)(u_long)res;
				else if (flags & SHORT)
					*va_arg(ap, short *) = res;
				else if (flags & LONG)
					*va_arg(ap, long *) = res;
				else if (flags & QUAD)
					*va_arg(ap, quad_t *) = res;
				else
					*va_arg(ap, int *) = res;
				nassigned++;
			}
			nread += p - buf;
			nconversions++;
			break;

		}
	}
input_failure:
	return (nconversions != 0 ? nassigned : -1);
match_failure:
	return (nassigned);
}

/*
 * Fill in the given table from the scanset at the given format
 * (just after `[').  Return a pointer to the character past the
 * closing `]'.  The table has a 1 wherever characters should be
 * considered part of the scanset.
 */
static u_char *
__sccl(char *tab, u_char *fmt)
{
	int c, n, v;

	/* first `clear' the whole table */
	c = *fmt++;		/* first char hat => negated scanset */
	if (c == '^') {
		v = 1;		/* default => accept */
		c = *fmt++;	/* get new first char */
	} else
		v = 0;		/* default => reject */

	/* XXX: Will not work if sizeof(tab*) > sizeof(char) */
	for (n = 0; n < 256; n++)
		     tab[n] = v;	/* memset(tab, v, 256) */

	if (c == 0)
		return (fmt - 1);/* format ended before closing ] */

	/*
	 * Now set the entries corresponding to the actual scanset
	 * to the opposite of the above.
	 *
	 * The first character may be ']' (or '-') without being special;
	 * the last character may be '-'.
	 */
	v = 1 - v;
	for (;;) {
		tab[c] = v;		/* take character c */
doswitch:
		n = *fmt++;		/* and examine the next */
		switch (n) {

		case 0:			/* format ended too soon */
			return (fmt - 1);

		case '-':
			/*
			 * A scanset of the form
			 *	[01+-]
			 * is defined as `the digit 0, the digit 1,
			 * the character +, the character -', but
			 * the effect of a scanset such as
			 *	[a-zA-Z0-9]
			 * is implementation defined.  The V7 Unix
			 * scanf treats `a-z' as `the letters a through
			 * z', but treats `a-a' as `the letter a, the
			 * character -, and the letter a'.
			 *
			 * For compatibility, the `-' is not considerd
			 * to define a range if the character following
			 * it is either a close bracket (required by ANSI)
			 * or is not numerically greater than the character
			 * we just stored in the table (c).
			 */
			n = *fmt;
			if (n == ']' || n < c) {
				c = '-';
				break;	/* resume the for(;;) */
			}
			fmt++;
			/* fill in the range */
			do {
			    tab[++c] = v;
			} while (c < n);
			c = n;
			/*
			 * Alas, the V7 Unix scanf also treats formats
			 * such as [a-c-e] as `the letters a through e'.
			 * This too is permitted by the standard....
			 */
			goto doswitch;
			break;

		case ']':		/* end of scanset */
			return (fmt);

		default:		/* just another character */
			c = n;
			break;
		}
	}
	/* NOTREACHED */
}

/*
 * Convert a string to an unsigned quad integer.
 *
 * Ignores `locale' stuff.  Assumes that the upper and lower case
 * alphabets and digits are each contiguous.
 */
u_quad_t
strtouq(const char *nptr, char **endptr, int base)
{
	const char *s = nptr;
	u_quad_t acc;
	unsigned char c;
	u_quad_t qbase, cutoff;
	int neg, any, cutlim;

	/*
	 * See strtoq for comments as to the logic used.
	 */
	s = nptr;
	do {
		c = *s++;
	} while (isspace(c));
	if (c == '-') {
		neg = 1;
		c = *s++;
	} else {
		neg = 0;
		if (c == '+')
			c = *s++;
	}
	if ((base == 0 || base == 16) &&
	    c == '0' && (*s == 'x' || *s == 'X')) {
		c = s[1];
		s += 2;
		base = 16;
	}
	if (base == 0)
		base = c == '0' ? 8 : 10;
	qbase = (unsigned)base;
	cutoff = (u_quad_t)UQUAD_MAX / qbase;
	cutlim = (u_quad_t)UQUAD_MAX % qbase;
	for (acc = 0, any = 0;; c = *s++) {
		if (!isascii(c))
			break;
		if (isdigit(c))
			c -= '0';
		else if (isalpha(c))
			c -= isupper(c) ? 'A' - 10 : 'a' - 10;
		else
			break;
		if (c >= base)
			break;
		if (any < 0 || acc > cutoff || (acc == cutoff && c > cutlim))
			any = -1;
		else {
			any = 1;
			acc *= qbase;
			acc += c;
		}
	}
	if (any < 0) {
		acc = UQUAD_MAX;
	} else if (neg)
		acc = -acc;
	if (endptr != 0)
		*endptr = (char *)(any ? s - 1 : nptr);
	return (acc);
}

/*
 * Convert a string to a quad integer.
 *
 * Ignores `locale' stuff.  Assumes that the upper and lower case
 * alphabets and digits are each contiguous.
 */
quad_t
strtoq(const char *nptr, char **endptr, int base)
{
	const char *s;
	u_quad_t acc;
	unsigned char c;
	u_quad_t qbase, cutoff;
	int neg, any, cutlim;

	/*
	 * Skip white space and pick up leading +/- sign if any.
	 * If base is 0, allow 0x for hex and 0 for octal, else
	 * assume decimal; if base is already 16, allow 0x.
	 */
	s = nptr;
	do {
		c = *s++;
	} while (isspace(c));
	if (c == '-') {
		neg = 1;
		c = *s++;
	} else {
		neg = 0;
		if (c == '+')
			c = *s++;
	}
	if ((base == 0 || base == 16) &&
	    c == '0' && (*s == 'x' || *s == 'X')) {
		c = s[1];
		s += 2;
		base = 16;
	}
	if (base == 0)
		base = c == '0' ? 8 : 10;

	/*
	 * Compute the cutoff value between legal numbers and illegal
	 * numbers.  That is the largest legal value, divided by the
	 * base.  An input number that is greater than this value, if
	 * followed by a legal input character, is too big.  One that
	 * is equal to this value may be valid or not; the limit
	 * between valid and invalid numbers is then based on the last
	 * digit.  For instance, if the range for quads is
	 * [-9223372036854775808..9223372036854775807] and the input base
	 * is 10, cutoff will be set to 922337203685477580 and cutlim to
	 * either 7 (neg==0) or 8 (neg==1), meaning that if we have
	 * accumulated a value > 922337203685477580, or equal but the
	 * next digit is > 7 (or 8), the number is too big, and we will
	 * return a range error.
	 *
	 * Set any if any `digits' consumed; make it negative to indicate
	 * overflow.
	 */
	qbase = (unsigned)base;
	cutoff = neg ? (u_quad_t)-(QUAD_MIN + QUAD_MAX) + QUAD_MAX : QUAD_MAX;
	cutlim = cutoff % qbase;
	cutoff /= qbase;
	for (acc = 0, any = 0;; c = *s++) {
		if (!isascii(c))
			break;
		if (isdigit(c))
			c -= '0';
		else if (isalpha(c))
			c -= isupper(c) ? 'A' - 10 : 'a' - 10;
		else
			break;
		if (c >= base)
			break;
		if (any < 0 || acc > cutoff || (acc == cutoff && c > cutlim))
			any = -1;
		else {
			any = 1;
			acc *= qbase;
			acc += c;
		}
	}
	if (any < 0) {
		acc = neg ? QUAD_MIN : QUAD_MAX;
	} else if (neg)
		acc = -acc;
	if (endptr != 0)
		*endptr = (char *)(any ? s - 1 : nptr);
	return (acc);
}