mystrtoul.c 8.15 KB
Newer Older
Guido van Rossum's avatar
Guido van Rossum committed
1

Guido van Rossum's avatar
Guido van Rossum committed
2
#include "Python.h"
3

4 5 6 7
#if defined(__sgi) && defined(WITH_THREAD) && !defined(_SGI_MP_SOURCE)
#define _SGI_MP_SOURCE
#endif

8 9
/* strtol and strtoul, renamed to avoid conflicts */

10 11

#include <ctype.h>
12
#ifdef HAVE_ERRNO_H
13 14 15 16 17 18 19
#include <errno.h>
#endif

/* Static overflow check values for bases 2 through 36.
 * smallmax[base] is the largest unsigned long i such that
 * i * base doesn't overflow unsigned long.
 */
20
static const unsigned long smallmax[] = {
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
    0, /* bases 0 and 1 are invalid */
    0,
    ULONG_MAX / 2,
    ULONG_MAX / 3,
    ULONG_MAX / 4,
    ULONG_MAX / 5,
    ULONG_MAX / 6,
    ULONG_MAX / 7,
    ULONG_MAX / 8,
    ULONG_MAX / 9,
    ULONG_MAX / 10,
    ULONG_MAX / 11,
    ULONG_MAX / 12,
    ULONG_MAX / 13,
    ULONG_MAX / 14,
    ULONG_MAX / 15,
    ULONG_MAX / 16,
    ULONG_MAX / 17,
    ULONG_MAX / 18,
    ULONG_MAX / 19,
    ULONG_MAX / 20,
    ULONG_MAX / 21,
    ULONG_MAX / 22,
    ULONG_MAX / 23,
    ULONG_MAX / 24,
    ULONG_MAX / 25,
    ULONG_MAX / 26,
    ULONG_MAX / 27,
    ULONG_MAX / 28,
    ULONG_MAX / 29,
    ULONG_MAX / 30,
    ULONG_MAX / 31,
    ULONG_MAX / 32,
    ULONG_MAX / 33,
    ULONG_MAX / 34,
    ULONG_MAX / 35,
    ULONG_MAX / 36,
58 59 60 61 62 63
};

/* maximum digits that can't ever overflow for bases 2 through 36,
 * calculated by [int(math.floor(math.log(2**32, i))) for i in range(2, 37)].
 * Note that this is pessimistic if sizeof(long) > 4.
 */
64
#if SIZEOF_LONG == 4
65
static const int digitlimit[] = {
66 67 68 69
    0,  0, 32, 20, 16, 13, 12, 11, 10, 10,  /*  0 -  9 */
    9,  9,  8,  8,  8,  8,  8,  7,  7,  7,  /* 10 - 19 */
    7,  7,  7,  7,  6,  6,  6,  6,  6,  6,  /* 20 - 29 */
    6,  6,  6,  6,  6,  6,  6};             /* 30 - 36 */
70 71
#elif SIZEOF_LONG == 8
/* [int(math.floor(math.log(2**64, i))) for i in range(2, 37)] */
72
static const int digitlimit[] = {
73 74 75 76
         0,   0, 64, 40, 32, 27, 24, 22, 21, 20,  /*  0 -  9 */
    19,  18, 17, 17, 16, 16, 16, 15, 15, 15,  /* 10 - 19 */
    14,  14, 14, 14, 13, 13, 13, 13, 13, 13,  /* 20 - 29 */
    13,  12, 12, 12, 12, 12, 12};             /* 30 - 36 */
77 78 79
#else
#error "Need table for SIZEOF_LONG"
#endif
80

Guido van Rossum's avatar
Guido van Rossum committed
81
/*
82 83 84 85 86 87 88 89 90 91 92
**      strtoul
**              This is a general purpose routine for converting
**              an ascii string to an integer in an arbitrary base.
**              Leading white space is ignored.  If 'base' is zero
**              it looks for a leading 0b, 0o or 0x to tell which
**              base.  If these are absent it defaults to 10.
**              Base must be 0 or between 2 and 36 (inclusive).
**              If 'ptr' is non-NULL it will contain a pointer to
**              the end of the scan.
**              Errors due to bad pointers will probably result in
**              exceptions - we don't check for them.
Guido van Rossum's avatar
Guido van Rossum committed
93 94
*/
unsigned long
95
PyOS_strtoul(const char *str, char **ptr, int base)
Guido van Rossum's avatar
Guido van Rossum committed
96
{
97 98 99
    unsigned long result = 0; /* return value of the function */
    int c;             /* current input character */
    int ovlimit;       /* required digits to overflow */
100

101
    /* skip leading white space */
102
    while (*str && Py_ISSPACE(Py_CHARMASK(*str)))
103
        ++str;
Guido van Rossum's avatar
Guido van Rossum committed
104

105 106 107 108 109 110 111 112 113
    /* check for leading 0b, 0o or 0x for auto-base or base 16 */
    switch (base) {
    case 0:             /* look for leading 0b, 0o or 0x */
        if (*str == '0') {
            ++str;
            if (*str == 'x' || *str == 'X') {
                /* there must be at least one digit after 0x */
                if (_PyLong_DigitValue[Py_CHARMASK(str[1])] >= 16) {
                    if (ptr)
114
                        *ptr = (char *)str;
115 116 117 118 119 120 121 122
                    return 0;
                }
                ++str;
                base = 16;
            } else if (*str == 'o' || *str == 'O') {
                /* there must be at least one digit after 0o */
                if (_PyLong_DigitValue[Py_CHARMASK(str[1])] >= 8) {
                    if (ptr)
123
                        *ptr = (char *)str;
124 125 126 127 128 129 130 131
                    return 0;
                }
                ++str;
                base = 8;
            } else if (*str == 'b' || *str == 'B') {
                /* there must be at least one digit after 0b */
                if (_PyLong_DigitValue[Py_CHARMASK(str[1])] >= 2) {
                    if (ptr)
132
                        *ptr = (char *)str;
133 134 135 136 137 138 139 140
                    return 0;
                }
                ++str;
                base = 2;
            } else {
                /* skip all zeroes... */
                while (*str == '0')
                    ++str;
141
                while (Py_ISSPACE(Py_CHARMASK(*str)))
142 143
                    ++str;
                if (ptr)
144
                    *ptr = (char *)str;
145 146 147 148 149 150
                return 0;
            }
        }
        else
            base = 10;
        break;
151

152 153 154 155 156 157 158 159
    /* even with explicit base, skip leading 0? prefix */
    case 16:
        if (*str == '0') {
            ++str;
            if (*str == 'x' || *str == 'X') {
                /* there must be at least one digit after 0x */
                if (_PyLong_DigitValue[Py_CHARMASK(str[1])] >= 16) {
                    if (ptr)
160
                        *ptr = (char *)str;
161 162 163 164 165 166 167 168 169 170 171 172 173
                    return 0;
                }
                ++str;
            }
        }
        break;
    case 8:
        if (*str == '0') {
            ++str;
            if (*str == 'o' || *str == 'O') {
                /* there must be at least one digit after 0o */
                if (_PyLong_DigitValue[Py_CHARMASK(str[1])] >= 8) {
                    if (ptr)
174
                        *ptr = (char *)str;
175 176 177 178 179 180 181 182 183 184 185 186 187
                    return 0;
                }
                ++str;
            }
        }
        break;
    case 2:
        if(*str == '0') {
            ++str;
            if (*str == 'b' || *str == 'B') {
                /* there must be at least one digit after 0b */
                if (_PyLong_DigitValue[Py_CHARMASK(str[1])] >= 2) {
                    if (ptr)
188
                        *ptr = (char *)str;
189 190 191 192 193 194 195
                    return 0;
                }
                ++str;
            }
        }
        break;
    }
196

197 198 199
    /* catch silly bases */
    if (base < 2 || base > 36) {
        if (ptr)
200
            *ptr = (char *)str;
201 202
        return 0;
    }
203

204 205 206
    /* skip leading zeroes */
    while (*str == '0')
        ++str;
207

208 209
    /* base is guaranteed to be in [2, 36] at this point */
    ovlimit = digitlimit[base];
210

211 212 213 214 215
    /* do the conversion until non-digit character encountered */
    while ((c = _PyLong_DigitValue[Py_CHARMASK(*str)]) < base) {
        if (ovlimit > 0) /* no overflow check required */
            result = result * base + c;
        else { /* requires overflow check */
216
            unsigned long temp_result;
217

218 219
            if (ovlimit < 0) /* guaranteed overflow */
                goto overflowed;
220

221 222 223 224
            /* there could be an overflow */
            /* check overflow just from shifting */
            if (result > smallmax[base])
                goto overflowed;
225

226
            result *= base;
227

228 229 230 231
            /* check overflow from the digit's value */
            temp_result = result + c;
            if (temp_result < result)
                goto overflowed;
232

233 234
            result = temp_result;
        }
235

236 237 238
        ++str;
        --ovlimit;
    }
239

240 241
    /* set pointer to point to the last character scanned */
    if (ptr)
242
        *ptr = (char *)str;
243

244
    return result;
245 246

overflowed:
247 248 249 250
    if (ptr) {
        /* spool through remaining digit characters */
        while (_PyLong_DigitValue[Py_CHARMASK(*str)] < base)
            ++str;
251
        *ptr = (char *)str;
252 253 254
    }
    errno = ERANGE;
    return (unsigned long)-1;
Guido van Rossum's avatar
Guido van Rossum committed
255 256
}

257 258
/* Checking for overflow in PyOS_strtol is a PITA; see comments
 * about PY_ABS_LONG_MIN in longobject.c.
259
 */
260
#define PY_ABS_LONG_MIN         (0-(unsigned long)LONG_MIN)
261

Guido van Rossum's avatar
Guido van Rossum committed
262
long
263
PyOS_strtol(const char *str, char **ptr, int base)
Guido van Rossum's avatar
Guido van Rossum committed
264
{
265 266 267
    long result;
    unsigned long uresult;
    char sign;
268

269
    while (*str && Py_ISSPACE(Py_CHARMASK(*str)))
270
        str++;
271

272 273 274
    sign = *str;
    if (sign == '+' || sign == '-')
        str++;
275

276
    uresult = PyOS_strtoul(str, ptr, base);
277

278 279 280 281 282 283 284 285 286 287 288 289 290
    if (uresult <= (unsigned long)LONG_MAX) {
        result = (long)uresult;
        if (sign == '-')
            result = -result;
    }
    else if (sign == '-' && uresult == PY_ABS_LONG_MIN) {
        result = LONG_MIN;
    }
    else {
        errno = ERANGE;
        result = LONG_MAX;
    }
    return result;
Guido van Rossum's avatar
Guido van Rossum committed
291
}