Date: Thu, 7 Sep 2023 20:28:45 GMT From: Dag-Erling =?utf-8?Q?Sm=C3=B8rgrav?= <des@FreeBSD.org> To: src-committers@FreeBSD.org, dev-commits-src-all@FreeBSD.org, dev-commits-src-branches@FreeBSD.org Subject: git: bae932999e0f - stable/14 - libc: Implement N2630. Message-ID: <202309072028.387KSjrr064522@gitrepo.freebsd.org>
next in thread | raw e-mail | index | archive | help
The branch stable/14 has been updated by des: URL: https://cgit.FreeBSD.org/src/commit/?id=bae932999e0fa9e9a08d84ca992332c57e5b42be commit bae932999e0fa9e9a08d84ca992332c57e5b42be Author: Dag-Erling Smørgrav <des@FreeBSD.org> AuthorDate: 2023-08-28 15:32:23 +0000 Commit: Dag-Erling Smørgrav <des@FreeBSD.org> CommitDate: 2023-09-07 20:28:30 +0000 libc: Implement N2630. This adds formatted input/output of binary integer numbers to the printf(), scanf(), and strtol() families, including their wide-character counterparts. Reviewed by: imp, emaste Differential Revision: https://reviews.freebsd.org/D41511 (cherry picked from commit d9dc1603d6e48cca84cad3ebe859129131b8387c) libc: Add unit tests for N2630 and possible collateral damage. Reviewed by: imp, emaste Differential Revision: https://reviews.freebsd.org/D41512 (cherry picked from commit b9385720f34b536ef2568a642e8b1fad0450056f) libc: Document support for binary integers. Reviewed by: debdrup, emaste Differential Revision: https://reviews.freebsd.org/D41522 (cherry picked from commit 76edfabbecdec686a570b8e009d5ea4112f943e0) libc: Fix fixed-width case in the new integer parser. Fixes: d9dc1603d6e4 Differential Revision: https://reviews.freebsd.org/D41622 (cherry picked from commit aca3bd1602577591e5cd237c4bb0bb71b3be0c75) libc: Add a wide version of snprintf_test. Reviewed by: imp, emaste Differential Revision: https://reviews.freebsd.org/D41726 (cherry picked from commit 4ec9ee9912765ac4ca57353999caa92a23283d8e) libc: Suppress format checks on printf() / scanf() tests. Reviewed by: jrtc27, markj, emaste Differential Revision: https://reviews.freebsd.org/D41727 (cherry picked from commit 294bd2827e61a78041f6613f4b82235fcc454157) Approved by: re (gjb) --- contrib/netbsd-tests/lib/libc/stdlib/t_strtol.c | 6 + lib/libc/iconv/_strtol.h | 7 + lib/libc/iconv/_strtoul.h | 7 + lib/libc/locale/wcstoimax.c | 7 + lib/libc/locale/wcstol.c | 7 + lib/libc/locale/wcstoll.c | 7 + lib/libc/locale/wcstoul.c | 7 + lib/libc/locale/wcstoull.c | 7 + lib/libc/locale/wcstoumax.c | 7 + lib/libc/stdio/printf.3 | 34 ++- lib/libc/stdio/printfcommon.h | 14 ++ lib/libc/stdio/scanf.3 | 29 ++- lib/libc/stdio/vfprintf.c | 13 ++ lib/libc/stdio/vfscanf.c | 267 +++++++++++++----------- lib/libc/stdio/vfwprintf.c | 13 ++ lib/libc/stdio/vfwscanf.c | 263 +++++++++++++---------- lib/libc/stdlib/strtoimax.c | 7 + lib/libc/stdlib/strtol.3 | 4 +- lib/libc/stdlib/strtol.c | 7 + lib/libc/stdlib/strtoll.c | 12 +- lib/libc/stdlib/strtoul.3 | 4 +- lib/libc/stdlib/strtoul.c | 7 + lib/libc/stdlib/strtoull.c | 7 + lib/libc/stdlib/strtoumax.c | 7 + lib/libc/tests/stdio/Makefile | 12 +- lib/libc/tests/stdio/snprintf_test.c | 139 ++++++++++++ lib/libc/tests/stdio/sscanf_test.c | 266 +++++++++++++++++++++++ lib/libc/tests/stdio/swprintf_test.c | 140 +++++++++++++ lib/libc/tests/stdio/swscanf_test.c | 267 ++++++++++++++++++++++++ 29 files changed, 1315 insertions(+), 259 deletions(-) diff --git a/contrib/netbsd-tests/lib/libc/stdlib/t_strtol.c b/contrib/netbsd-tests/lib/libc/stdlib/t_strtol.c index 54e190760656..d1027fcc7bb1 100644 --- a/contrib/netbsd-tests/lib/libc/stdlib/t_strtol.c +++ b/contrib/netbsd-tests/lib/libc/stdlib/t_strtol.c @@ -94,6 +94,12 @@ ATF_TC_BODY(strtol_base, tc) { "01234567", 342391, 0, NULL }, { "0123456789", 123456789, 10, NULL }, { "0x75bcd15", 123456789, 0, NULL }, +#ifdef __FreeBSD__ + { "0x", 0, 0, "x" }, + { "0b111010110111100110100010101", 123456789, 0, NULL }, + { "0b0123", 1, 0, "23" }, + { "0b", 0, 0, "b" }, +#endif }; long long int lli; diff --git a/lib/libc/iconv/_strtol.h b/lib/libc/iconv/_strtol.h index d183edbe8c3a..94a13c56db98 100644 --- a/lib/libc/iconv/_strtol.h +++ b/lib/libc/iconv/_strtol.h @@ -91,6 +91,13 @@ _FUNCNAME(const char *nptr, char **endptr, int base) s += 2; base = 16; } + if ((base == 0 || base == 2) && + c == '0' && (*s == 'b' || *s == 'B') && + (s[1] >= '0' && s[1] <= '1')) { + c = s[1]; + s += 2; + base = 2; + } if (base == 0) base = (c == '0' ? 8 : 10); diff --git a/lib/libc/iconv/_strtoul.h b/lib/libc/iconv/_strtoul.h index eade72e9c2e6..4944e1fb06e0 100644 --- a/lib/libc/iconv/_strtoul.h +++ b/lib/libc/iconv/_strtoul.h @@ -87,6 +87,13 @@ _FUNCNAME(const char *nptr, char **endptr, int base) s += 2; base = 16; } + if ((base == 0 || base == 2) && + c == '0' && (*s == 'b' || *s == 'B') && + (s[1] >= '0' && s[1] <= '1')) { + c = s[1]; + s += 2; + base = 2; + } if (base == 0) base = (c == '0' ? 8 : 10); diff --git a/lib/libc/locale/wcstoimax.c b/lib/libc/locale/wcstoimax.c index 259faa2b011c..5ed949cd0531 100644 --- a/lib/libc/locale/wcstoimax.c +++ b/lib/libc/locale/wcstoimax.c @@ -86,6 +86,13 @@ wcstoimax_l(const wchar_t * __restrict nptr, wchar_t ** __restrict endptr, s += 2; base = 16; } + if ((base == 0 || base == 2) && + c == L'0' && (*s == L'b' || *s == L'B') && + (s[1] >= L'0' && s[1] <= L'1')) { + c = s[1]; + s += 2; + base = 2; + } if (base == 0) base = c == L'0' ? 8 : 10; acc = any = 0; diff --git a/lib/libc/locale/wcstol.c b/lib/libc/locale/wcstol.c index b0b787384f39..1678b615ca1c 100644 --- a/lib/libc/locale/wcstol.c +++ b/lib/libc/locale/wcstol.c @@ -80,6 +80,13 @@ wcstol_l(const wchar_t * __restrict nptr, wchar_t ** __restrict endptr, int s += 2; base = 16; } + if ((base == 0 || base == 2) && + c == L'0' && (*s == L'b' || *s == L'B') && + (s[1] >= L'0' && s[1] <= L'1')) { + c = s[1]; + s += 2; + base = 2; + } if (base == 0) base = c == L'0' ? 8 : 10; acc = any = 0; diff --git a/lib/libc/locale/wcstoll.c b/lib/libc/locale/wcstoll.c index ac07d6c6adbf..ef1e6ef58861 100644 --- a/lib/libc/locale/wcstoll.c +++ b/lib/libc/locale/wcstoll.c @@ -86,6 +86,13 @@ wcstoll_l(const wchar_t * __restrict nptr, wchar_t ** __restrict endptr, s += 2; base = 16; } + if ((base == 0 || base == 2) && + c == L'0' && (*s == L'b' || *s == L'B') && + (s[1] >= L'0' && s[1] <= L'1')) { + c = s[1]; + s += 2; + base = 2; + } if (base == 0) base = c == L'0' ? 8 : 10; acc = any = 0; diff --git a/lib/libc/locale/wcstoul.c b/lib/libc/locale/wcstoul.c index 9f58db799c0e..2c9c8820b1f6 100644 --- a/lib/libc/locale/wcstoul.c +++ b/lib/libc/locale/wcstoul.c @@ -80,6 +80,13 @@ wcstoul_l(const wchar_t * __restrict nptr, wchar_t ** __restrict endptr, s += 2; base = 16; } + if ((base == 0 || base == 2) && + c == L'0' && (*s == L'b' || *s == L'B') && + (s[1] >= L'0' && s[1] <= L'1')) { + c = s[1]; + s += 2; + base = 2; + } if (base == 0) base = c == L'0' ? 8 : 10; acc = any = 0; diff --git a/lib/libc/locale/wcstoull.c b/lib/libc/locale/wcstoull.c index cbc7253f884d..692eb90eef6b 100644 --- a/lib/libc/locale/wcstoull.c +++ b/lib/libc/locale/wcstoull.c @@ -86,6 +86,13 @@ wcstoull_l(const wchar_t * __restrict nptr, wchar_t ** __restrict endptr, s += 2; base = 16; } + if ((base == 0 || base == 2) && + c == L'0' && (*s == L'b' || *s == L'B') && + (s[1] >= L'0' && s[1] <= L'1')) { + c = s[1]; + s += 2; + base = 2; + } if (base == 0) base = c == L'0' ? 8 : 10; acc = any = 0; diff --git a/lib/libc/locale/wcstoumax.c b/lib/libc/locale/wcstoumax.c index 4380cccf2424..c4f2ec3aaf41 100644 --- a/lib/libc/locale/wcstoumax.c +++ b/lib/libc/locale/wcstoumax.c @@ -86,6 +86,13 @@ wcstoumax_l(const wchar_t * __restrict nptr, wchar_t ** __restrict endptr, s += 2; base = 16; } + if ((base == 0 || base == 2) && + c == L'0' && (*s == L'b' || *s == L'B') && + (s[1] >= L'0' && s[1] <= L'1')) { + c = s[1]; + s += 2; + base = 2; + } if (base == 0) base = c == L'0' ? 8 : 10; acc = any = 0; diff --git a/lib/libc/stdio/printf.3 b/lib/libc/stdio/printf.3 index 3e5c6ca23511..110851e2a421 100644 --- a/lib/libc/stdio/printf.3 +++ b/lib/libc/stdio/printf.3 @@ -31,7 +31,7 @@ .\" .\" @(#)printf.3 8.1 (Berkeley) 6/4/93 .\" -.Dd May 22, 2018 +.Dd August 21, 2023 .Dt PRINTF 3 .Os .Sh NAME @@ -212,6 +212,17 @@ and .Cm u conversions, this option has no effect. For +.Cm b +and +.Cm B +conversions, a non-zero result has the string +.Ql 0b +(or +.Ql 0B +for +.Cm B +conversions) prepended to it. +For .Cm o conversions, the precision of the number is increased to force the first character of the output string to a zero. @@ -245,7 +256,7 @@ For all conversions except .Cm n , the converted value is padded on the left with zeros rather than blanks. If a precision is given with a numeric conversion -.Cm ( d , i , o , u , i , x , +.Cm ( b , B , d , i , o , u , i , x , and .Cm X ) , the @@ -301,7 +312,7 @@ followed by an optional digit string. If the digit string is omitted, the precision is taken as zero. This gives the minimum number of digits to appear for -.Cm d , i , o , u , x , +.Cm b , B , d , i , o , u , x , and .Cm X conversions, the number of digits to appear after the decimal-point for @@ -319,12 +330,12 @@ conversions. .It An optional length modifier, that specifies the size of the argument. The following length modifiers are valid for the -.Cm d , i , n , o , u , x , +.Cm b , B , d , i , n , o , u , x , or .Cm X conversion: .Bl -column ".Cm q Em (deprecated)" ".Vt signed char" ".Vt unsigned long long" ".Vt long long *" -.It Sy Modifier Ta Cm d , i Ta Cm o , u , x , X Ta Cm n +.It Sy Modifier Ta Cm d , i Ta Cm b , B , o , u , x , X Ta Cm n .It Cm hh Ta Vt "signed char" Ta Vt "unsigned char" Ta Vt "signed char *" .It Cm h Ta Vt short Ta Vt "unsigned short" Ta Vt "short *" .It Cm l No (ell) Ta Vt long Ta Vt "unsigned long" Ta Vt "long *" @@ -339,7 +350,7 @@ Note: the .Cm t modifier, when applied to a -.Cm o , u , x , +.Cm b , B , o , u , x , or .Cm X conversion, indicates that the argument is of an unsigned type @@ -403,11 +414,16 @@ If a single format directive mixes positional and non-positional arguments, the results are undefined. .Pp The conversion specifiers and their meanings are: -.Bl -tag -width ".Cm diouxX" -.It Cm diouxX +.Bl -tag -width ".Cm bBdiouxX" +.It Cm bBdiouxX The .Vt int -(or appropriate variant) argument is converted to signed decimal +(or appropriate variant) argument is converted to +unsigned binary +.Cm ( b +and +.Cm B ) , +signed decimal .Cm ( d and .Cm i ) , diff --git a/lib/libc/stdio/printfcommon.h b/lib/libc/stdio/printfcommon.h index ac5aed0a5fcd..411b778dc234 100644 --- a/lib/libc/stdio/printfcommon.h +++ b/lib/libc/stdio/printfcommon.h @@ -194,6 +194,13 @@ __ultoa(u_long val, CHAR *endp, int base, int octzero, const char *xdigs) } while (sval != 0); break; + case 2: + do { + *--cp = to_char(val & 1); + val >>= 1; + } while (val); + break; + case 8: do { *--cp = to_char(val & 7); @@ -244,6 +251,13 @@ __ujtoa(uintmax_t val, CHAR *endp, int base, int octzero, const char *xdigs) } while (sval != 0); break; + case 2: + do { + *--cp = to_char(val & 1); + val >>= 1; + } while (val); + break; + case 8: do { *--cp = to_char(val & 7); diff --git a/lib/libc/stdio/scanf.3 b/lib/libc/stdio/scanf.3 index b1c50e10a795..6cefdb133983 100644 --- a/lib/libc/stdio/scanf.3 +++ b/lib/libc/stdio/scanf.3 @@ -31,7 +31,7 @@ .\" .\" @(#)scanf.3 8.2 (Berkeley) 12/11/93 .\" -.Dd April 2, 2022 +.Dd August 21, 2023 .Dt SCANF 3 .Os .Sh NAME @@ -141,7 +141,7 @@ The conversion that follows occurs as usual, but no pointer is used; the result of the conversion is simply discarded. .It Cm hh Indicates that the conversion will be one of -.Cm dioux +.Cm bdioux or .Cm n and the next pointer is a pointer to a @@ -150,7 +150,7 @@ and the next pointer is a pointer to a .Vt int ) . .It Cm h Indicates that the conversion will be one of -.Cm dioux +.Cm bdioux or .Cm n and the next pointer is a pointer to a @@ -159,7 +159,7 @@ and the next pointer is a pointer to a .Vt int ) . .It Cm l No (ell) Indicates that the conversion will be one of -.Cm dioux +.Cm bdioux or .Cm n and the next pointer is a pointer to a @@ -185,7 +185,7 @@ and the next pointer is a pointer to an array of .Vt char ) . .It Cm ll No (ell ell) Indicates that the conversion will be one of -.Cm dioux +.Cm bdioux or .Cm n and the next pointer is a pointer to a @@ -201,7 +201,7 @@ and the next pointer is a pointer to .Vt "long double" . .It Cm j Indicates that the conversion will be one of -.Cm dioux +.Cm bdioux or .Cm n and the next pointer is a pointer to a @@ -210,7 +210,7 @@ and the next pointer is a pointer to a .Vt int ) . .It Cm t Indicates that the conversion will be one of -.Cm dioux +.Cm bdioux or .Cm n and the next pointer is a pointer to a @@ -219,7 +219,7 @@ and the next pointer is a pointer to a .Vt int ) . .It Cm z Indicates that the conversion will be one of -.Cm dioux +.Cm bdioux or .Cm n and the next pointer is a pointer to a @@ -229,7 +229,7 @@ and the next pointer is a pointer to a .It Cm q (deprecated.) Indicates that the conversion will be one of -.Cm dioux +.Cm bdioux or .Cm n and the next pointer is a pointer to a @@ -273,6 +273,10 @@ matches a single input .Ql % character. No conversion is done, and assignment does not occur. +.It Cm b , B +Matches an optionally signed binary integer; +the next pointer must be a pointer to +.Vt "unsigned int" . .It Cm d Matches an optionally signed decimal integer; the next pointer must be a pointer to @@ -281,7 +285,12 @@ the next pointer must be a pointer to Matches an optionally signed integer; the next pointer must be a pointer to .Vt int . -The integer is read in base 16 if it begins +The integer is read +in base 2 if it begins with +.Ql 0b +or +.Ql 0B , +in base 16 if it begins with .Ql 0x or diff --git a/lib/libc/stdio/vfprintf.c b/lib/libc/stdio/vfprintf.c index ad655c5d78d4..5e5a9b5e31c1 100644 --- a/lib/libc/stdio/vfprintf.c +++ b/lib/libc/stdio/vfprintf.c @@ -613,6 +613,19 @@ reswitch: switch (ch) { case 'z': flags |= SIZET; goto rflag; + case 'B': + case 'b': + if (flags & INTMAX_SIZE) + ujval = UJARG(); + else + ulval = UARG(); + base = 2; + /* leading 0b/B only if non-zero */ + if (flags & ALT && + (flags & INTMAX_SIZE ? ujval != 0 : ulval != 0)) + ox[1] = ch; + goto nosign; + break; case 'C': flags |= LONGINT; /*FALLTHROUGH*/ diff --git a/lib/libc/stdio/vfscanf.c b/lib/libc/stdio/vfscanf.c index cc2e1e428321..9727c9e70c34 100644 --- a/lib/libc/stdio/vfscanf.c +++ b/lib/libc/stdio/vfscanf.c @@ -6,6 +6,8 @@ * * Copyright (c) 2011 The FreeBSD Foundation * + * Copyright (c) 2023 Dag-Erling Smørgrav + * * Portions of this software were developed by David Chisnall * under sponsorship from the FreeBSD Foundation. * @@ -80,16 +82,6 @@ static char sccsid[] = "@(#)vfscanf.c 8.1 (Berkeley) 6/4/93"; #define SHORTSHORT 0x4000 /* hh: char */ #define UNSIGNED 0x8000 /* %[oupxX] conversions */ -/* - * The following are used in integral conversions only: - * SIGNOK, NDIGITS, PFXOK, and NZDIGITS - */ -#define SIGNOK 0x40 /* +/- is (still) legal */ -#define NDIGITS 0x80 /* no digits detected */ -#define PFXOK 0x100 /* 0x prefix is (still) legal */ -#define NZDIGITS 0x200 /* no zero digits detected */ -#define HAVESIGN 0x10000 /* sign detected */ - /* * Conversion types. */ @@ -307,129 +299,160 @@ convert_wstring(FILE *fp, wchar_t *wcp, int width, locale_t locale) return (n); } +enum parseint_state { + begin, + havesign, + havezero, + haveprefix, + any, +}; + +static __inline int +parseint_fsm(int c, enum parseint_state *state, int *base) +{ + switch (c) { + case '+': + case '-': + if (*state == begin) { + *state = havesign; + return 1; + } + break; + case '0': + if (*state == begin || *state == havesign) { + *state = havezero; + } else { + *state = any; + } + return 1; + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + if (*state == havezero && *base == 0) { + *base = 8; + } + /* FALL THROUGH */ + case '8': + case '9': + if (*state == begin || + *state == havesign) { + if (*base == 0) { + *base = 10; + } + } + if (*state == begin || + *state == havesign || + *state == havezero || + *state == haveprefix || + *state == any) { + if (*base > c - '0') { + *state = any; + return 1; + } + } + break; + case 'b': + if (*state == havezero) { + if (*base == 0 || *base == 2) { + *state = haveprefix; + *base = 2; + return 1; + } + } + /* FALL THROUGH */ + case 'a': + case 'c': + case 'd': + case 'e': + case 'f': + if (*state == begin || + *state == havesign || + *state == havezero || + *state == haveprefix || + *state == any) { + if (*base > c - 'a' + 10) { + *state = any; + return 1; + } + } + break; + case 'B': + if (*state == havezero) { + if (*base == 0 || *base == 2) { + *state = haveprefix; + *base = 2; + return 1; + } + } + /* FALL THROUGH */ + case 'A': + case 'C': + case 'D': + case 'E': + case 'F': + if (*state == begin || + *state == havesign || + *state == havezero || + *state == haveprefix || + *state == any) { + if (*base > c - 'A' + 10) { + *state = any; + return 1; + } + } + break; + case 'x': + case 'X': + if (*state == havezero) { + if (*base == 0 || *base == 16) { + *state = haveprefix; + *base = 16; + return 1; + } + } + break; + } + return 0; +} + /* - * Read an integer, storing it in buf. The only relevant bit in the - * flags argument is PFXOK. + * Read an integer, storing it in buf. * * Return 0 on a match failure, and the number of characters read * otherwise. */ static __inline int -parseint(FILE *fp, char * __restrict buf, int width, int base, int flags) +parseint(FILE *fp, char * __restrict buf, int width, int base) { - /* `basefix' is used to avoid `if' tests */ - static const short basefix[17] = - { 10, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 }; + enum parseint_state state = begin; char *p; int c; - 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; - flags |= HAVESIGN; - goto ok; - } + c = __sgetc(fp); + if (c == EOF) break; - - /* - * x ok iff flag still set & 2nd char (or 3rd char if - * we have a sign). - */ - case 'x': case 'X': - if (flags & PFXOK && p == - buf + 1 + !!(flags & HAVESIGN)) { - base = 16; /* if %i */ - flags &= ~PFXOK; - goto ok; - } + if (!parseint_fsm(c, &state, &base)) 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 we only had a sign, push it back. If we only had a 0b or 0x + * prefix (possibly preceded by a sign), we view it as "0" and + * push back the letter. In all other cases, if we stopped + * because we read a non-number character, push it back. */ - if (flags & NDIGITS) { - if (p > buf) - (void) __ungetc(*(u_char *)--p, fp); - return (0); - } - c = ((u_char *)p)[-1]; - if (c == 'x' || c == 'X') { - --p; + if (state == havesign) { + p--; + (void) __ungetc(*(u_char *)p, fp); + } else if (state == haveprefix) { + p--; + (void) __ungetc(c, fp); + } else if (width && c != EOF) { (void) __ungetc(c, fp); } return (p - buf); @@ -554,6 +577,13 @@ literal: /* * Conversions. */ + case 'B': + case 'b': + c = CT_INT; + flags |= UNSIGNED; + base = 2; + break; + case 'd': c = CT_INT; base = 10; @@ -578,7 +608,6 @@ literal: case 'X': case 'x': - flags |= PFXOK; /* enable 0x prefixing */ c = CT_INT; flags |= UNSIGNED; base = 16; @@ -613,7 +642,7 @@ literal: break; case 'p': /* pointer format is like hex */ - flags |= POINTER | PFXOK; + flags |= POINTER; c = CT_INT; /* assumes sizeof(uintmax_t) */ flags |= UNSIGNED; /* >= sizeof(uintptr_t) */ base = 16; @@ -738,7 +767,7 @@ literal: width = sizeof(buf) - 2; width++; #endif - nr = parseint(fp, buf, width, base, flags); + nr = parseint(fp, buf, width, base); if (nr == 0) goto match_failure; if ((flags & SUPPRESS) == 0) { diff --git a/lib/libc/stdio/vfwprintf.c b/lib/libc/stdio/vfwprintf.c index fc681e8d0575..259a86467ea7 100644 --- a/lib/libc/stdio/vfwprintf.c +++ b/lib/libc/stdio/vfwprintf.c @@ -684,6 +684,19 @@ reswitch: switch (ch) { case 'z': flags |= SIZET; goto rflag; + case 'B': + case 'b': + if (flags & INTMAX_SIZE) + ujval = UJARG(); + else + ulval = UARG(); + base = 2; + /* leading 0b/B only if non-zero */ + if (flags & ALT && + (flags & INTMAX_SIZE ? ujval != 0 : ulval != 0)) + ox[1] = ch; + goto nosign; + break; case 'C': flags |= LONGINT; /*FALLTHROUGH*/ diff --git a/lib/libc/stdio/vfwscanf.c b/lib/libc/stdio/vfwscanf.c index 1a28ff665247..b03c9dba0699 100644 --- a/lib/libc/stdio/vfwscanf.c +++ b/lib/libc/stdio/vfwscanf.c @@ -9,6 +9,8 @@ * * Copyright (c) 2011 The FreeBSD Foundation * + * Copyright (c) 2023 Dag-Erling Smørgrav + * * Portions of this software were developed by David Chisnall * under sponsorship from the FreeBSD Foundation. * @@ -78,16 +80,6 @@ static char sccsid[] = "@(#)vfscanf.c 8.1 (Berkeley) 6/4/93"; #define SHORTSHORT 0x4000 /* hh: char */ #define UNSIGNED 0x8000 /* %[oupxX] conversions */ -/* - * The following are used in integral conversions only: - * SIGNOK, NDIGITS, PFXOK, and NZDIGITS - */ -#define SIGNOK 0x40 /* +/- is (still) legal */ -#define NDIGITS 0x80 /* no digits detected */ -#define PFXOK 0x100 /* 0x prefix is (still) legal */ -#define NZDIGITS 0x200 /* no zero digits detected */ -#define HAVESIGN 0x10000 /* sign detected */ - /* * Conversion types. */ @@ -289,128 +281,161 @@ convert_wstring(FILE *fp, wchar_t *wcp, int width, locale_t locale) return (nread); } +enum parseint_state { + begin, + havesign, + havezero, + haveprefix, + any, +}; + +static __inline int +parseint_fsm(wchar_t c, enum parseint_state *state, int *base) +{ + switch (c) { + case '+': + case '-': + if (*state == begin) { + *state = havesign; + return 1; + } + break; + case '0': + if (*state == begin || *state == havesign) { + *state = havezero; + } else { + *state = any; + } + return 1; + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + if (*state == havezero && *base == 0) { + *base = 8; + } + /* FALL THROUGH */ + case '8': + case '9': + if (*state == begin || + *state == havesign) { + if (*base == 0) { + *base = 10; + } + } + if (*state == begin || + *state == havesign || + *state == havezero || + *state == haveprefix || + *state == any) { + if (*base > c - '0') { + *state = any; + return 1; + } + } + break; + case 'b': + if (*state == havezero) { + if (*base == 0 || *base == 2) { + *state = haveprefix; + *base = 2; + return 1; + } + } + /* FALL THROUGH */ + case 'a': + case 'c': + case 'd': + case 'e': + case 'f': + if (*state == begin || + *state == havesign || + *state == havezero || + *state == haveprefix || + *state == any) { + if (*base > c - 'a' + 10) { + *state = any; + return 1; + } + } + break; + case 'B': + if (*state == havezero) { + if (*base == 0 || *base == 2) { + *state = haveprefix; *** 1248 LINES SKIPPED ***
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?202309072028.387KSjrr064522>