stdlib/strtol.c

   1 /* strtol - Convert string representation of a number into an integer value.
   2    Copyright (C) 1991, 92, 94, 95, 96 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4
   5    The GNU C Library is free software; you can redistribute it and/or
   6    modify it under the terms of the GNU Library General Public License as
   7    published by the Free Software Foundation; either version 2 of the
   8    License, or (at your option) any later version.
   9
  10    The GNU C Library is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13    Library General Public License for more details.
  14
  15    You should have received a copy of the GNU Library General Public
  16    License along with the GNU C Library; see the file COPYING.LIB.  If not,
  17    write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  18    Boston, MA 02111-1307, USA.  */
  19
  20 #if HAVE_CONFIG_H
  21 # include <config.h>
  22 #endif
  23
  24 #ifdef _LIBC
  25 # define USE_NUMBER_GROUPING
  26 # define STDC_HEADERS
  27 # define HAVE_LIMITS_H
  28 #endif
  29
  30 #include <ctype.h>
  31 #include <errno.h>
  32 #ifndef errno
  33 extern int errno;
  34 #endif
  35 #ifndef __set_errno
  36 # define __set_errno(Val) errno = (Val)
  37 #endif
  38
  39 #ifdef HAVE_LIMITS_H
  40 # include <limits.h>
  41 #endif
  42
  43 #ifdef STDC_HEADERS
  44 # include <stddef.h>
  45 # include <stdlib.h>
  46 # include <string.h>
  47 #else
  48 # ifndef NULL
  49 #  define NULL 0
  50 # endif
  51 #endif
  52
  53 #ifdef USE_NUMBER_GROUPING
  54 # include "../locale/localeinfo.h"
  55 #endif
  56
  57 /* Nonzero if we are defining `strtoul' or `strtouq', operating on
  58    unsigned integers.  */
  59 #ifndef UNSIGNED
  60 # define UNSIGNED 0
  61 # define INT LONG int
  62 #else
  63 # define INT unsigned LONG int
  64 #endif
  65
  66 /* Determine the name.  */
  67 #if UNSIGNED
  68 # ifdef USE_WIDE_CHAR
  69 #  ifdef QUAD
  70 #   define strtol wcstouq
  71 #  else
  72 #   define strtol wcstoul
  73 #  endif
  74 # else
  75 #  ifdef QUAD
  76 #   define strtol strtouq
  77 #  else
  78 #   define strtol strtoul
  79 #  endif
  80 # endif
  81 #else
  82 # ifdef USE_WIDE_CHAR
  83 #  ifdef QUAD
  84 #   define strtol wcstoq
  85 #  else
  86 #   define strtol wcstol
  87 #  endif
  88 # else
  89 #  ifdef QUAD
  90 #   define strtol strtoq
  91 #  endif
  92 # endif
  93 #endif
  94
  95 /* If QUAD is defined, we are defining `strtoq' or `strtouq',
  96    operating on `long long int's.  */
  97 #ifdef QUAD
  98 # define LONG long long
  99 # undef LONG_MIN
 100 # define LONG_MIN LONG_LONG_MIN
 101 # undef LONG_MAX
 102 # define LONG_MAX LONG_LONG_MAX
 103 # undef ULONG_MAX
 104 # define ULONG_MAX ULONG_LONG_MAX
 105 # if __GNUC__ == 2 && __GNUC_MINOR__ < 7
 106    /* Work around gcc bug with using this constant.  */
 107    static const unsigned long long int maxquad = ULONG_LONG_MAX;
 108 #  undef ULONG_MAX
 109 #  define ULONG_MAX maxquad
 110 # endif
 111 #else
 112 # define LONG long
 113
 114 #ifndef ULONG_MAX
 115 # define ULONG_MAX ((unsigned long) ~(unsigned long) 0)
 116 #endif
 117 #ifndef LONG_MAX
 118 # define LONG_MAX ((long int) (ULONG_MAX >> 1))
 119 #endif
 120 #endif
 121
 122 #ifdef USE_WIDE_CHAR
 123 # include <wchar.h>
 124 # include <wctype.h>
 125 # define L_(Ch) L##Ch
 126 # define UCHAR_TYPE wint_t
 127 # define STRING_TYPE wchar_t
 128 # define ISSPACE(Ch) iswspace (Ch)
 129 # define ISALPHA(Ch) iswalpha (Ch)
 130 # define TOUPPER(Ch) towupper (Ch)
 131 #else
 132 # define L_(Ch) Ch
 133 # define UCHAR_TYPE unsigned char
 134 # define STRING_TYPE char
 135 # define ISSPACE(Ch) isspace (Ch)
 136 # define ISALPHA(Ch) isalpha (Ch)
 137 # define TOUPPER(Ch) toupper (Ch)
 138 #endif
 139
 140 #ifdef __STDC__
 141 # define INTERNAL(X) INTERNAL1(X)
 142 # define INTERNAL1(X) __##X##_internal
 143 # define WEAKNAME(X) WEAKNAME1(X)
 144 #else
 145 # define INTERNAL(X) __/**/X/**/_internal
 146 #endif
 147
 148 #ifdef USE_NUMBER_GROUPING
 149 /* This file defines a function to check for correct grouping.  */
 150 # include "grouping.h"
 151 #endif
 152
 153
 154 /* Convert NPTR to an `unsigned long int' or `long int' in base BASE.
 155    If BASE is 0 the base is determined by the presence of a leading
 156    zero, indicating octal or a leading "0x" or "0X", indicating hexadecimal.
 157    If BASE is < 2 or > 36, it is reset to 10.
 158    If ENDPTR is not NULL, a pointer to the character after the last
 159    one converted is stored in *ENDPTR.  */
 160
 161 INT
 162 INTERNAL (strtol) (nptr, endptr, base, group)
 163      const STRING_TYPE *nptr;
 164      STRING_TYPE **endptr;
 165      int base;
 166      int group;
 167 {
 168   int negative;
 169   register unsigned LONG int cutoff;
 170   register unsigned int cutlim;
 171   register unsigned LONG int i;
 172   register const STRING_TYPE *s;
 173   register UCHAR_TYPE c;
 174   const STRING_TYPE *save, *end;
 175   int overflow;
 176
 177 #ifdef USE_NUMBER_GROUPING
 178   /* The thousands character of the current locale.  */
 179   wchar_t thousands;
 180   /* The numeric grouping specification of the current locale,
 181      in the format described in <locale.h>.  */
 182   const char *grouping;
 183
 184   if (group)
 185     {
 186       grouping = _NL_CURRENT (LC_NUMERIC, GROUPING);
 187       if (*grouping <= 0 || *grouping == CHAR_MAX)
 188         grouping = NULL;
 189       else
 190         {
 191           /* Figure out the thousands separator character.  */
 192           if (mbtowc (&thousands, _NL_CURRENT (LC_NUMERIC, THOUSANDS_SEP),
 193                       strlen (_NL_CURRENT (LC_NUMERIC, THOUSANDS_SEP))) <= 0)
 194             thousands = (wchar_t) *_NL_CURRENT (LC_NUMERIC, THOUSANDS_SEP);
 195           if (thousands == L'\0')
 196             grouping = NULL;
 197         }
 198     }
 199   else
 200     grouping = NULL;
 201 #endif
 202
 203   if (base < 0 || base == 1 || base > 36)
 204     {
 205       __set_errno (EINVAL);
 206       return 0;
 207     }
 208
 209   save = s = nptr;
 210
 211   /* Skip white space.  */
 212   while (ISSPACE (*s))
 213     ++s;
 214   if (*s == L_('\0'))
 215     goto noconv;
 216
 217   /* Check for a sign.  */
 218   if (*s == L_('-'))
 219     {
 220 #if UNSIGNED
 221       __set_errno (EINVAL);
 222       return 0;
 223 #else
 224       negative = 1;
 225       ++s;
 226 #endif
 227     }
 228   else if (*s == L_('+'))
 229     {
 230       negative = 0;
 231       ++s;
 232     }
 233   else
 234     negative = 0;
 235
 236   if (base == 16 && s[0] == L_('0') && TOUPPER (s[1]) == L_('X'))
 237     s += 2;
 238
 239   /* If BASE is zero, figure it out ourselves.  */
 240   if (base == 0)
 241     if (*s == L_('0'))
 242       {
 243         if (TOUPPER (s[1]) == L_('X'))
 244           {
 245             s += 2;
 246             base = 16;
 247           }
 248         else
 249           base = 8;
 250       }
 251     else
 252       base = 10;
 253
 254   /* Save the pointer so we can check later if anything happened.  */
 255   save = s;
 256
 257 #ifdef USE_NUMBER_GROUPING
 258   if (group)
 259     {
 260       /* Find the end of the digit string and check its grouping.  */
 261       end = s;
 262       for (c = *end; c != L_('\0'); c = *++end)
 263         if ((wchar_t) c != thousands
 264             && ((wchar_t) c < L_('0') || (wchar_t) c > L_('9'))
 265             && (!ISALPHA (c) || (int) (TOUPPER (c) - L_('A') + 10) >= base))
 266           break;
 267       if (*s == thousands)
 268         end = s;
 269       else
 270         end = correctly_grouped_prefix (s, end, thousands, grouping);
 271     }
 272   else
 273 #endif
 274     end = NULL;
 275
 276   cutoff = ULONG_MAX / (unsigned LONG int) base;
 277   cutlim = ULONG_MAX % (unsigned LONG int) base;
 278
 279   overflow = 0;
 280   i = 0;
 281   for (c = *s; c != L_('\0'); c = *++s)
 282     {
 283       if (s == end)
 284         break;
 285       if (c >= L_('0') && c <= L_('9'))
 286         c -= L_('0');
 287       else if (ISALPHA (c))
 288         c = TOUPPER (c) - L_('A') + 10;
 289       else
 290         break;
 291       if ((int) c >= base)
 292         break;
 293       /* Check for overflow.  */
 294       if (i > cutoff || (i == cutoff && c > cutlim))
 295         overflow = 1;
 296       else
 297         {
 298           i *= (unsigned LONG int) base;
 299           i += c;
 300         }
 301     }
 302
 303   /* Check if anything actually happened.  */
 304   if (s == save)
 305     goto noconv;
 306
 307   /* Store in ENDPTR the address of one character
 308      past the last character we converted.  */
 309   if (endptr != NULL)
 310     *endptr = (STRING_TYPE *) s;
 311
 312 #if !UNSIGNED
 313   /* Check for a value that is within the range of
 314      `unsigned LONG int', but outside the range of `LONG int'.  */
 315   if (overflow == 0
 316       && i > (negative
 317               ? -((unsigned LONG int) (LONG_MIN + 1)) + 1
 318               : (unsigned LONG int) LONG_MAX))
 319     overflow = 1;
 320 #endif
 321
 322   if (overflow)
 323     {
 324       __set_errno (ERANGE);
 325 #if UNSIGNED
 326       return ULONG_MAX;
 327 #else
 328       return negative ? LONG_MIN : LONG_MAX;
 329 #endif
 330     }
 331
 332   /* Return the result of the appropriate sign.  */
 333   return (negative ? -i : i);
 334
 335 noconv:
 336   /* We must handle a special case here: the base is 0 or 16 and the
 337      first two characters are '0' and 'x', but the rest are no
 338      hexadecimal digits.  This is no error case.  We return 0 and
 339      ENDPTR points to the `x`.  */
 340   if (endptr != NULL)
 341     if (save - nptr >= 2 && TOUPPER (save[-1]) == L_('X')
 342         && save[-2] == L_('0'))
 343       *endptr = (STRING_TYPE *) &save[-1];
 344     else
 345       /*  There was no number to convert.  */
 346       *endptr = (STRING_TYPE *) nptr;
 347
 348   return 0L;
 349 }
 350 \f
 351 /* External user entry point.  */
 352
 353 #if _LIBC - 0 == 0
 354 # undef PARAMS
 355 # if defined (__STDC__) && __STDC__
 356 #  define PARAMS(Args) Args
 357 # else
 358 #  define PARAMS(Args) ()
 359 # endif
 360
 361 /* Prototype.  */
 362 INT strtol PARAMS ((const STRING_TYPE *nptr, STRING_TYPE **endptr, int base));
 363 #endif
 364
 365
 366 INT
 367 #ifdef weak_function
 368 weak_function
 369 #endif
 370 strtol (nptr, endptr, base)
 371      const STRING_TYPE *nptr;
 372      STRING_TYPE **endptr;
 373      int base;
 374 {
 375   return INTERNAL (strtol) (nptr, endptr, base, 0);
 376 }