summaryrefslogtreecommitdiffstats
path: root/fluent-bit/src/flb_strptime.c
diff options
context:
space:
mode:
Diffstat (limited to 'fluent-bit/src/flb_strptime.c')
-rw-r--r--fluent-bit/src/flb_strptime.c750
1 files changed, 750 insertions, 0 deletions
diff --git a/fluent-bit/src/flb_strptime.c b/fluent-bit/src/flb_strptime.c
new file mode 100644
index 000000000..492a6ee74
--- /dev/null
+++ b/fluent-bit/src/flb_strptime.c
@@ -0,0 +1,750 @@
+/* $OpenBSD: strptime.c,v 1.30 2019/05/12 12:49:52 schwarze Exp $ */
+/* $NetBSD: strptime.c,v 1.12 1998/01/20 21:39:40 mycroft Exp $ */
+/*-
+ * Copyright (c) 1997, 1998, 2005, 2008 The NetBSD Foundation, Inc.
+ * All rights reserved.
+ *
+ * This code was contributed to The NetBSD Foundation by Klaus Klein.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ * 1. Redistributions of source code must retain the above copyright
+ * notice, this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ * notice, this list of conditions and the following disclaimer in the
+ * documentation and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
+ * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
+ * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
+ * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+/*
+ * This file provides a portable implementation of strptime(2), based
+ * on the work of OpenSBD project. Since various platforms implement
+ * strptime differently, this one should work as a fallback.
+ */
+
+#include <ctype.h>
+#include <locale.h>
+#include <stdint.h>
+#include <string.h>
+
+#include <fluent-bit/flb_compat.h>
+#include <fluent-bit/flb_langinfo.h>
+#include <fluent-bit/flb_time.h>
+
+#define _ctloc(x) (nl_langinfo(x))
+
+/*
+ * We do not implement alternate representations. However, we always
+ * check whether a given modifier is allowed for a certain conversion.
+ */
+#define _ALT_E 0x01
+#define _ALT_O 0x02
+#define _LEGAL_ALT(x) { if (alt_format & ~(x)) return (0); }
+
+/*
+ * Copied from libc/time/private.h and libc/time/tzfile.h
+ */
+#define TM_YEAR_BASE 1900
+#define DAYSPERNYEAR 365
+#define DAYSPERLYEAR 366
+#define DAYSPERWEEK 7
+#define MONSPERYEAR 12
+#define EPOCH_YEAR 1970
+#define EPOCH_WDAY 4 /* Thursday */
+#define SECSPERHOUR 3600
+#define SECSPERMIN 60
+
+#define isleap(y) (((y) % 4) == 0 && (((y) % 100) != 0 || ((y) % 400) == 0))
+
+/*
+ * We keep track of some of the fields we set in order to compute missing ones.
+ */
+#define FIELD_TM_MON (1 << 0)
+#define FIELD_TM_MDAY (1 << 1)
+#define FIELD_TM_WDAY (1 << 2)
+#define FIELD_TM_YDAY (1 << 3)
+#define FIELD_TM_YEAR (1 << 4)
+
+static char gmt[] = { "GMT" };
+static char utc[] = { "UTC" };
+/* RFC-822/RFC-2822 */
+static const char * const nast[5] = {
+ "EST", "CST", "MST", "PST", "\0\0\0"
+};
+static const char * const nadt[5] = {
+ "EDT", "CDT", "MDT", "PDT", "\0\0\0"
+};
+
+static const int mon_lengths[2][MONSPERYEAR] = {
+ { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 },
+ { 31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 }
+};
+
+static nl_item day[] = {
+ DAY_1, DAY_2, DAY_3, DAY_4, DAY_5, DAY_6, DAY_7
+};
+
+static nl_item mon[] = {
+ MON_1, MON_2, MON_3, MON_4, MON_5, MON_6, MON_7, MON_8, MON_9,
+ MON_10, MON_11, MON_12
+};
+
+static nl_item abday[] = {
+ ABDAY_1, ABDAY_2, ABDAY_3, ABDAY_4, ABDAY_5, ABDAY_6, ABDAY_7
+};
+
+static nl_item abmon[] = {
+ ABMON_1, ABMON_2, ABMON_3, ABMON_4, ABMON_5, ABMON_6, ABMON_7,
+ ABMON_8, ABMON_9, ABMON_10, ABMON_11, ABMON_12
+};
+
+static int _conv_num64(const unsigned char **, int64_t *, int64_t, int64_t);
+static int _conv_num(const unsigned char **, int *, int, int);
+static int leaps_thru_end_of(const int y);
+static char *_flb_strptime(const char *, const char *, struct flb_tm *, int);
+static const u_char *_find_string(const u_char *, int *, const char * const *,
+ const char * const *, int);
+
+/*
+ * FreeBSD does not support `timezone` in time.h.
+ * https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=24590
+ */
+#ifdef __FreeBSD__
+int flb_timezone(void)
+{
+ struct tm tm;
+ time_t t = 0;
+ tzset();
+ localtime_r(&t, &tm);
+ return -(tm.tm_gmtoff);
+}
+#define timezone (flb_timezone())
+#endif
+
+char *
+flb_strptime(const char *buf, const char *fmt, struct flb_tm *tm)
+{
+ return(_flb_strptime(buf, fmt, tm, 1));
+}
+
+static char *
+_flb_strptime(const char *buf, const char *fmt, struct flb_tm *tm, int initialize)
+{
+ unsigned char c;
+ const unsigned char *bp, *ep;
+ size_t len = 0;
+ int alt_format, i, offs;
+ int neg = 0;
+ static int century, relyear, fields;
+
+ if (initialize) {
+ century = TM_YEAR_BASE;
+ relyear = -1;
+ fields = 0;
+ }
+
+ bp = (const unsigned char *)buf;
+ while ((c = *fmt) != '\0') {
+ /* Clear `alternate' modifier prior to new conversion. */
+ alt_format = 0;
+
+ /* Eat up white-space. */
+ if (isspace(c)) {
+ while (isspace(*bp))
+ bp++;
+
+ fmt++;
+ continue;
+ }
+
+ /*
+ * Having increased bp we need to ensure we are not
+ * moving beyond bounds.
+ */
+ if (*bp == '\0')
+ return (NULL);
+
+ if ((c = *fmt++) != '%')
+ goto literal;
+
+
+again: switch (c = *fmt++) {
+ case '%': /* "%%" is converted to "%". */
+literal:
+ if (c != *bp++)
+ return (NULL);
+
+ break;
+
+ /*
+ * "Alternative" modifiers. Just set the appropriate flag
+ * and start over again.
+ */
+ case 'E': /* "%E?" alternative conversion modifier. */
+ _LEGAL_ALT(0);
+ alt_format |= _ALT_E;
+ goto again;
+
+ case 'O': /* "%O?" alternative conversion modifier. */
+ _LEGAL_ALT(0);
+ alt_format |= _ALT_O;
+ goto again;
+
+ /*
+ * "Complex" conversion rules, implemented through recursion.
+ */
+ case 'c': /* Date and time, using the locale's format. */
+ _LEGAL_ALT(_ALT_E);
+ if (!(bp = (const unsigned char *)_flb_strptime((const char *)bp, _ctloc(D_T_FMT), tm, 0)))
+ return (NULL);
+ break;
+
+ case 'D': /* The date as "%m/%d/%y". */
+ _LEGAL_ALT(0);
+ if (!(bp = (const unsigned char *)_flb_strptime((const char *)bp, "%m/%d/%y", tm, 0)))
+ return (NULL);
+ break;
+
+ case 'F': /* The date as "%Y-%m-%d". */
+ _LEGAL_ALT(0);
+ if (!(bp = (const unsigned char *)_flb_strptime((const char *)bp, "%Y-%m-%d", tm, 0)))
+ return (NULL);
+ continue;
+
+ case 'R': /* The time as "%H:%M". */
+ _LEGAL_ALT(0);
+ if (!(bp = (const unsigned char *)_flb_strptime((const char *)bp, "%H:%M", tm, 0)))
+ return (NULL);
+ break;
+
+ case 'r': /* The time as "%I:%M:%S %p". */
+ _LEGAL_ALT(0);
+ if (!(bp = (const unsigned char *)_flb_strptime((const char *)bp, "%I:%M:%S %p", tm, 0)))
+ return (NULL);
+ break;
+
+ case 'T': /* The time as "%H:%M:%S". */
+ _LEGAL_ALT(0);
+ if (!(bp = (const unsigned char *)_flb_strptime((const char *)bp, "%H:%M:%S", tm, 0)))
+ return (NULL);
+ break;
+
+ case 'X': /* The time, using the locale's format. */
+ _LEGAL_ALT(_ALT_E);
+ if (!(bp = (const unsigned char *)_flb_strptime((const char *)bp, _ctloc(T_FMT), tm, 0)))
+ return (NULL);
+ break;
+
+ case 'x': /* The date, using the locale's format. */
+ _LEGAL_ALT(_ALT_E);
+ if (!(bp = (const unsigned char *)_flb_strptime((const char *)bp, _ctloc(D_FMT), tm, 0)))
+ return (NULL);
+ break;
+
+ /*
+ * "Elementary" conversion rules.
+ */
+ case 'A': /* The day of week, using the locale's form. */
+ case 'a':
+ _LEGAL_ALT(0);
+ for (i = 0; i < 7; i++) {
+ /* Full name. */
+ len = strlen(_ctloc(day[i]));
+ if (strncasecmp(_ctloc(day[i]), (const char *)bp, len) == 0)
+ break;
+
+ /* Abbreviated name. */
+ len = strlen(_ctloc(abday[i]));
+ if (strncasecmp(_ctloc(abday[i]), (const char *)bp, len) == 0)
+ break;
+ }
+
+ /* Nothing matched. */
+ if (i == 7)
+ return (NULL);
+
+ tm->tm.tm_wday = i;
+ bp += len;
+ fields |= FIELD_TM_WDAY;
+ break;
+
+ case 'B': /* The month, using the locale's form. */
+ case 'b':
+ case 'h':
+ _LEGAL_ALT(0);
+ for (i = 0; i < 12; i++) {
+ /* Full name. */
+ len = strlen(_ctloc(mon[i]));
+ if (strncasecmp(_ctloc(mon[i]), (const char *)bp, len) == 0)
+ break;
+
+ /* Abbreviated name. */
+ len = strlen(_ctloc(abmon[i]));
+ if (strncasecmp(_ctloc(abmon[i]), (const char *)bp, len) == 0)
+ break;
+ }
+
+ /* Nothing matched. */
+ if (i == 12)
+ return (NULL);
+
+ tm->tm.tm_mon = i;
+ bp += len;
+ fields |= FIELD_TM_MON;
+ break;
+
+ case 'C': /* The century number. */
+ _LEGAL_ALT(_ALT_E);
+ if (!(_conv_num(&bp, &i, 0, 99)))
+ return (NULL);
+
+ century = i * 100;
+ break;
+
+ case 'e': /* The day of month. */
+ if (isspace(*bp))
+ bp++;
+ /* FALLTHROUGH */
+ case 'd':
+ _LEGAL_ALT(_ALT_O);
+ if (!(_conv_num(&bp, &tm->tm.tm_mday, 1, 31)))
+ return (NULL);
+ fields |= FIELD_TM_MDAY;
+ break;
+
+ case 'k': /* The hour (24-hour clock representation). */
+ _LEGAL_ALT(0);
+ /* FALLTHROUGH */
+ case 'H':
+ _LEGAL_ALT(_ALT_O);
+ if (!(_conv_num(&bp, &tm->tm.tm_hour, 0, 23)))
+ return (NULL);
+ break;
+
+ case 'l': /* The hour (12-hour clock representation). */
+ _LEGAL_ALT(0);
+ /* FALLTHROUGH */
+ case 'I':
+ _LEGAL_ALT(_ALT_O);
+ if (!(_conv_num(&bp, &tm->tm.tm_hour, 1, 12)))
+ return (NULL);
+ break;
+
+ case 'j': /* The day of year. */
+ _LEGAL_ALT(0);
+ if (!(_conv_num(&bp, &tm->tm.tm_yday, 1, 366)))
+ return (NULL);
+ tm->tm.tm_yday--;
+ fields |= FIELD_TM_YDAY;
+ break;
+
+ case 'M': /* The minute. */
+ _LEGAL_ALT(_ALT_O);
+ if (!(_conv_num(&bp, &tm->tm.tm_min, 0, 59)))
+ return (NULL);
+ break;
+
+ case 'm': /* The month. */
+ _LEGAL_ALT(_ALT_O);
+ if (!(_conv_num(&bp, &tm->tm.tm_mon, 1, 12)))
+ return (NULL);
+ tm->tm.tm_mon--;
+ fields |= FIELD_TM_MON;
+ break;
+
+ case 'p': /* The locale's equivalent of AM/PM. */
+ _LEGAL_ALT(0);
+ /* AM? */
+ len = strlen(_ctloc(AM_STR));
+ if (strncasecmp(_ctloc(AM_STR), (const char *)bp, len) == 0) {
+ if (tm->tm.tm_hour > 12) /* i.e., 13:00 AM ?! */
+ return (NULL);
+ else if (tm->tm.tm_hour == 12)
+ tm->tm.tm_hour = 0;
+
+ bp += len;
+ break;
+ }
+ /* PM? */
+ len = strlen(_ctloc(PM_STR));
+ if (strncasecmp(_ctloc(PM_STR), (const char *)bp, len) == 0) {
+ if (tm->tm.tm_hour > 12) /* i.e., 13:00 PM ?! */
+ return (NULL);
+ else if (tm->tm.tm_hour < 12)
+ tm->tm.tm_hour += 12;
+
+ bp += len;
+ break;
+ }
+
+ /* Nothing matched. */
+ return (NULL);
+
+ case 'S': /* The seconds. */
+ _LEGAL_ALT(_ALT_O);
+ if (!(_conv_num(&bp, &tm->tm.tm_sec, 0, 60)))
+ return (NULL);
+ break;
+ case 's': /* Seconds since epoch */
+ {
+ int64_t i64;
+ if (!(_conv_num64(&bp, &i64, 0, INT64_MAX)))
+ return (NULL);
+ if (!gmtime_r((const time_t *) &i64, &tm->tm))
+ return (NULL);
+ fields = 0xffff; /* everything */
+ }
+ break;
+ case 'U': /* The week of year, beginning on sunday. */
+ case 'W': /* The week of year, beginning on monday. */
+ _LEGAL_ALT(_ALT_O);
+ /*
+ * XXX This is bogus, as we can not assume any valid
+ * information present in the tm structure at this
+ * point to calculate a real value, so just check the
+ * range for now.
+ */
+ if (!(_conv_num(&bp, &i, 0, 53)))
+ return (NULL);
+ break;
+
+ case 'w': /* The day of week, beginning on sunday. */
+ _LEGAL_ALT(_ALT_O);
+ if (!(_conv_num(&bp, &tm->tm.tm_wday, 0, 6)))
+ return (NULL);
+ fields |= FIELD_TM_WDAY;
+ break;
+
+ case 'u': /* The day of week, monday = 1. */
+ _LEGAL_ALT(_ALT_O);
+ if (!(_conv_num(&bp, &i, 1, 7)))
+ return (NULL);
+ tm->tm.tm_wday = i % 7;
+ fields |= FIELD_TM_WDAY;
+ continue;
+
+ case 'g': /* The year corresponding to the ISO week
+ * number but without the century.
+ */
+ if (!(_conv_num(&bp, &i, 0, 99)))
+ return (NULL);
+ continue;
+
+ case 'G': /* The year corresponding to the ISO week
+ * number with century.
+ */
+ do
+ bp++;
+ while (isdigit(*bp));
+ continue;
+
+ case 'V': /* The ISO 8601:1988 week number as decimal */
+ if (!(_conv_num(&bp, &i, 0, 53)))
+ return (NULL);
+ continue;
+
+ case 'Y': /* The year. */
+ _LEGAL_ALT(_ALT_E);
+ if (!(_conv_num(&bp, &i, 0, 9999)))
+ return (NULL);
+
+ relyear = -1;
+ tm->tm.tm_year = i - TM_YEAR_BASE;
+ fields |= FIELD_TM_YEAR;
+ break;
+
+ case 'y': /* The year within the century (2 digits). */
+ _LEGAL_ALT(_ALT_E | _ALT_O);
+ if (!(_conv_num(&bp, &relyear, 0, 99)))
+ return (NULL);
+ break;
+
+ case 'Z':
+ tzset();
+ if (strncmp((const char *)bp, gmt, 3) == 0) {
+ tm->tm.tm_isdst = 0;
+ flb_tm_gmtoff(tm) = 0;
+#ifdef FLB_HAVE_ZONE
+ tm->tm.tm_zone = gmt;
+#endif
+ bp += 3;
+ } else if (strncmp((const char *)bp, utc, 3) == 0) {
+ tm->tm.tm_isdst = 0;
+ flb_tm_gmtoff(tm) = 0;
+#ifdef FLB_HAVE_ZONE
+ tm->tm.tm_zone = utc;
+#endif
+ bp += 3;
+ } else {
+ ep = _find_string(bp, &i,
+ (const char * const *)tzname,
+ NULL, 2);
+ if (ep == NULL)
+ return (NULL);
+
+ tm->tm.tm_isdst = i;
+ flb_tm_gmtoff(tm) = -(timezone);
+#ifdef FLB_HAVE_ZONE
+ tm->tm.tm_zone = tzname[i];
+#endif
+ bp = ep;
+ }
+ continue;
+
+ case 'z':
+ /*
+ * We recognize all ISO 8601 formats:
+ * Z = Zulu time/UTC
+ * [+-]hhmm
+ * [+-]hh:mm
+ * [+-]hh
+ * We recognize all RFC-822/RFC-2822 formats:
+ * UT|GMT
+ * North American : UTC offsets
+ * E[DS]T = Eastern : -4 | -5
+ * C[DS]T = Central : -5 | -6
+ * M[DS]T = Mountain: -6 | -7
+ * P[DS]T = Pacific : -7 | -8
+ */
+ while (isspace(*bp))
+ bp++;
+
+ switch (*bp++) {
+ case 'G':
+ if (*bp++ != 'M')
+ return NULL;
+ /*FALLTHROUGH*/
+ case 'U':
+ if (*bp++ != 'T')
+ return NULL;
+ /*FALLTHROUGH*/
+ case 'Z':
+ tm->tm.tm_isdst = 0;
+ flb_tm_gmtoff(tm) = 0;
+#ifdef FLB_HAVE_ZONE
+ tm->tm.tm_zone = utc;
+#endif
+ continue;
+ case '+':
+ neg = 0;
+ break;
+ case '-':
+ neg = 1;
+ break;
+ default:
+ --bp;
+ ep = _find_string(bp, &i, nast, NULL, 4);
+ if (ep != NULL) {
+ flb_tm_gmtoff(tm) = (-5 - i) * SECSPERHOUR;
+#ifdef FLB_HAVE_ZONE
+ tm->tm.tm_zone = (char *)nast[i];
+#endif
+ bp = ep;
+ continue;
+ }
+ ep = _find_string(bp, &i, nadt, NULL, 4);
+ if (ep != NULL) {
+ tm->tm.tm_isdst = 1;
+ flb_tm_gmtoff(tm) = (-4 - i) * SECSPERHOUR;
+#ifdef FLB_HAVE_ZONE
+ tm->tm.tm_zone = (char *)nadt[i];
+#endif
+ bp = ep;
+ continue;
+ }
+ return NULL;
+ }
+ if (!isdigit(bp[0]) || !isdigit(bp[1]))
+ return NULL;
+ offs = ((bp[0]-'0') * 10 + (bp[1]-'0')) * SECSPERHOUR;
+ bp += 2;
+ if (*bp == ':')
+ bp++;
+ if (isdigit(*bp)) {
+ offs += (*bp++ - '0') * 10 * SECSPERMIN;
+ if (!isdigit(*bp))
+ return NULL;
+ offs += (*bp++ - '0') * SECSPERMIN;
+ }
+ if (neg)
+ offs = -offs;
+ tm->tm.tm_isdst = 0; /* XXX */
+ flb_tm_gmtoff(tm) = offs;
+#ifdef FLB_HAVE_ZONE
+ tm->tm.tm_zone = NULL; /* XXX */
+#endif
+ continue;
+
+ /*
+ * Miscellaneous conversions.
+ */
+ case 'n': /* Any kind of white-space. */
+ case 't':
+ _LEGAL_ALT(0);
+ while (isspace(*bp))
+ bp++;
+ break;
+
+
+ default: /* Unknown/unsupported conversion. */
+ return (NULL);
+ }
+
+
+ }
+
+ /*
+ * We need to evaluate the two digit year spec (%y)
+ * last as we can get a century spec (%C) at any time.
+ */
+ if (relyear != -1) {
+ if (century == TM_YEAR_BASE) {
+ if (relyear <= 68)
+ tm->tm.tm_year = relyear + 2000 - TM_YEAR_BASE;
+ else
+ tm->tm.tm_year = relyear + 1900 - TM_YEAR_BASE;
+ } else {
+ tm->tm.tm_year = relyear + century - TM_YEAR_BASE;
+ }
+ fields |= FIELD_TM_YEAR;
+ }
+
+ /* Compute some missing values when possible. */
+ if (fields & FIELD_TM_YEAR) {
+ const int year = (unsigned int)tm->tm.tm_year + (unsigned int)TM_YEAR_BASE;
+ const int *mon_lens = mon_lengths[isleap(year)];
+ if (!(fields & FIELD_TM_YDAY) &&
+ (fields & FIELD_TM_MON) && (fields & FIELD_TM_MDAY)) {
+ tm->tm.tm_yday = tm->tm.tm_mday - 1;
+ for (i = 0; i < tm->tm.tm_mon; i++)
+ tm->tm.tm_yday += mon_lens[i];
+ fields |= FIELD_TM_YDAY;
+ }
+ if (fields & FIELD_TM_YDAY) {
+ int days = tm->tm.tm_yday;
+ if (!(fields & FIELD_TM_WDAY)) {
+ tm->tm.tm_wday = EPOCH_WDAY +
+ ((year - EPOCH_YEAR) % DAYSPERWEEK) *
+ (DAYSPERNYEAR % DAYSPERWEEK) +
+ leaps_thru_end_of(year - 1) -
+ leaps_thru_end_of(EPOCH_YEAR - 1) +
+ tm->tm.tm_yday;
+ tm->tm.tm_wday %= DAYSPERWEEK;
+ if (tm->tm.tm_wday < 0)
+ tm->tm.tm_wday += DAYSPERWEEK;
+ }
+ if (!(fields & FIELD_TM_MON)) {
+ tm->tm.tm_mon = 0;
+ while (tm->tm.tm_mon < MONSPERYEAR && days >= mon_lens[tm->tm.tm_mon])
+ days -= mon_lens[tm->tm.tm_mon++];
+ }
+ if (!(fields & FIELD_TM_MDAY))
+ tm->tm.tm_mday = days + 1;
+ }
+ }
+
+ return ((char *)bp);
+}
+
+
+static int
+_conv_num(const unsigned char **buf, int *dest, int llim, int ulim)
+{
+ int result = 0;
+ int rulim = ulim;
+
+ if (**buf < '0' || **buf > '9')
+ return (0);
+
+ /* we use rulim to break out of the loop when we run out of digits */
+ do {
+ result *= 10;
+ result += *(*buf)++ - '0';
+ rulim /= 10;
+ } while ((result * 10 <= ulim) && rulim && **buf >= '0' && **buf <= '9');
+
+ if (result < llim || result > ulim)
+ return (0);
+
+ *dest = result;
+ return (1);
+}
+
+static int
+_conv_num64(const unsigned char **buf, int64_t *dest, int64_t llim, int64_t ulim)
+{
+ int64_t result = 0;
+ int64_t rulim = ulim;
+
+ if (**buf < '0' || **buf > '9')
+ return (0);
+
+ /* we use rulim to break out of the loop when we run out of digits */
+ do {
+ /* Avoid overflow: result > ((2**64)/2.0) / 10.0 */
+ if (result > 922337203685477580) {
+ return (0);
+ }
+ result *= 10;
+
+ /* Avoid overflow: result > ((2**64)/2.0) - 48 */
+ if (result > 9223372036854775760) {
+ return (0);
+ }
+ result += *(*buf)++ - '0';
+ rulim /= 10;
+ /* watch out for overflows. If value gets above
+ * ((2**64)/2.0)/10.0 then we will overflow. So instead
+ * we return 0 */
+ if (result >= 922337203685477580) {
+ return (0);
+ }
+ } while ((result * 10 <= ulim) && rulim && **buf >= '0' && **buf <= '9');
+
+ if (result < llim || result > ulim)
+ return (0);
+
+ *dest = result;
+ return (1);
+}
+
+static const u_char *
+_find_string(const u_char *bp, int *tgt, const char * const *n1,
+ const char * const *n2, int c)
+{
+ int i;
+ unsigned int len;
+
+ /* check full name - then abbreviated ones */
+ for (; n1 != NULL; n1 = n2, n2 = NULL) {
+ for (i = 0; i < c; i++, n1++) {
+ len = strlen(*n1);
+ if (strncasecmp(*n1, (const char *)bp, len) == 0) {
+ *tgt = i;
+ return bp + len;
+ }
+ }
+ }
+
+ /* Nothing matched */
+ return NULL;
+}
+
+static int
+leaps_thru_end_of(const int y)
+{
+ return (y >= 0) ? (y / 4 - y / 100 + y / 400) :
+ -(leaps_thru_end_of(-(y + 1)) + 1);
+}