vfscanf.c

在x86平台上运行不可信任代码的sandbox。

							goto input_failure;
						break;
					}
				}
				n = p - p0;
				if (n == 0)
					goto match_failure;
				*p = 0;
				nassigned++;
			}
			nread += n;
			break;

		case CT_STRING:
			/* like CCL, but zero-length string OK, & no NOSKIP */
			if (width == 0)
				width = (size_t)~0;
			if (flags & SUPPRESS) {
				n = 0;
				while (!isspace(*fp->_p)) {
					n++, fp->_r--, fp->_p++;
					if (--width == 0)
						break;
					if (fp->_r <= 0 && __srefill(fp))
						break;
				}
				nread += n;
			} else {
				p0 = p = va_arg(ap, char *);
				while (!isspace(*fp->_p)) {
					fp->_r--;
					*p++ = *fp->_p++;
					if (--width == 0)
						break;
					if (fp->_r <= 0 && __srefill(fp))
						break;
				}
				*p = 0;
				nread += p - p0;
				nassigned++;
			}
			continue;

		case CT_INT:
			/* scan an integer as if by strtoq/strtouq */
#ifdef hardway
			if (width == 0 || width > sizeof(buf) - 1)
				width = sizeof(buf) - 1;
#else
			/* 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 = *fp->_p;
				/*
				 * 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 (--fp->_r > 0)
					fp->_p++;
				else if (__srefill(fp))
					break;		/* EOF */
			}
			/*
			 * 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)
					(void) ungetc(*(u_char *)--p, fp);
				goto match_failure;
			}
			c = ((u_char *)p)[-1];
			if (c == 'x' || c == 'X') {
				--p;
				(void) ungetc(c, fp);
			}
			if ((flags & SUPPRESS) == 0) {
				u_quad_t res;

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

#ifdef FLOATING_POINT
		case CT_FLOAT:
			/* scan a floating point number as if by strtod */
#ifdef hardway
			if (width == 0 || width > sizeof(buf) - 1)
				width = sizeof(buf) - 1;
#else
			/* size_t is unsigned, hence this optimisation */
			if (--width > sizeof(buf) - 2)
				width = sizeof(buf) - 2;
			width++;
#endif
			flags |= SIGNOK | NDIGITS | DPTOK | EXPOK;
			for (p = buf; width; width--) {
				c = *fp->_p;
				/*
				 * This code mimicks the integer conversion
				 * code, but is much simpler.
				 */
				switch (c) {

				case '0': case '1': case '2': case '3':
				case '4': case '5': case '6': case '7':
				case '8': case '9':
					flags &= ~(SIGNOK | NDIGITS);
					goto fok;

				case '+': case '-':
					if (flags & SIGNOK) {
						flags &= ~SIGNOK;
						goto fok;
					}
					break;
				case '.':
					if (flags & DPTOK) {
						flags &= ~(SIGNOK | DPTOK);
						goto fok;
					}
					break;
				case 'e': case 'E':
					/* no exponent without some digits */
					if ((flags&(NDIGITS|EXPOK)) == EXPOK) {
						flags =
						    (flags & ~(EXPOK|DPTOK)) |
						    SIGNOK | NDIGITS;
						goto fok;
					}
					break;
				}
				break;
		fok:
				*p++ = c;
				if (--fp->_r > 0)
					fp->_p++;
				else if (__srefill(fp))
					break;	/* EOF */
			}
			/*
			 * If no digits, might be missing exponent digits
			 * (just give back the exponent) or might be missing
			 * regular digits, but had sign and/or decimal point.
			 */
			if (flags & NDIGITS) {
				if (flags & EXPOK) {
					/* no digits at all */
					while (p > buf)
						ungetc(*(u_char *)--p, fp);
					goto match_failure;
				}
				/* just a bad exponent (e and maybe sign) */
				c = *(u_char *)--p;
				if (c != 'e' && c != 'E') {
					(void) ungetc(c, fp);/* sign */
					c = *(u_char *)--p;
				}
				(void) ungetc(c, fp);
			}
			if ((flags & SUPPRESS) == 0) {
				double res;

				*p = 0;
				res = strtod(buf, (char **) NULL);
				if (flags & LONGDBL)
					*va_arg(ap, long double *) = res;
				else if (flags & LONG)
					*va_arg(ap, double *) = res;
				else
					*va_arg(ap, float *) = res;
				nassigned++;
			}
			nread += p - buf;
			break;
#endif /* FLOATING_POINT */
		}
	}
input_failure:
	return (nassigned ? 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(tab, fmt)
	register char *tab;
	register u_char *fmt;
{
	register 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 */
	/* should probably use memset here */
	for (n = 0; n < 256; n++)
		tab[n] = v;
	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++;
			do {		/* fill in the range */
				tab[++c] = v;
			} while (c < n);
#if 1	/* XXX another disgusting compatibility hack */
			/*
			 * 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;
#else
			c = *fmt++;
			if (c == 0)
				return (fmt - 1);
			if (c == ']')
				return (fmt);
#endif
			break;

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

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