summaryrefslogtreecommitdiffstats
path: root/src/gl/parse-datetime.c
diff options
context:
space:
mode:
Diffstat (limited to '')
-rw-r--r--src/gl/parse-datetime.c3992
1 files changed, 3992 insertions, 0 deletions
diff --git a/src/gl/parse-datetime.c b/src/gl/parse-datetime.c
new file mode 100644
index 0000000..f385e12
--- /dev/null
+++ b/src/gl/parse-datetime.c
@@ -0,0 +1,3992 @@
+/* A Bison parser, made by GNU Bison 3.8.2. */
+
+/* Bison implementation for Yacc-like parsers in C
+
+ Copyright (C) 1984, 1989-1990, 2000-2015, 2018-2021 Free Software Foundation,
+ Inc.
+
+ This program is free software: you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation, either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <https://www.gnu.org/licenses/>. */
+
+/* As a special exception, you may create a larger work that contains
+ part or all of the Bison parser skeleton and distribute that work
+ under terms of your choice, so long as that work isn't itself a
+ parser generator using the skeleton or a modified version thereof
+ as a parser skeleton. Alternatively, if you modify or redistribute
+ the parser skeleton itself, you may (at your option) remove this
+ special exception, which will cause the skeleton and the resulting
+ Bison output files to be licensed under the GNU General Public
+ License without this special exception.
+
+ This special exception was added by the Free Software Foundation in
+ version 2.2 of Bison. */
+
+/* C LALR(1) parser skeleton written by Richard Stallman, by
+ simplifying the original so-called "semantic" parser. */
+
+/* DO NOT RELY ON FEATURES THAT ARE NOT DOCUMENTED in the manual,
+ especially those whose name start with YY_ or yy_. They are
+ private implementation details that can be changed or removed. */
+
+/* All symbols defined below should begin with yy or YY, to avoid
+ infringing on user name space. This should be done even for local
+ variables, as they might otherwise be expanded by user macros.
+ There are some unavoidable exceptions within include files to
+ define necessary library symbols; they are noted "INFRINGES ON
+ USER NAME SPACE" below. */
+
+/* Identify Bison output, and Bison version. */
+#define YYBISON 30802
+
+/* Bison version string. */
+#define YYBISON_VERSION "3.8.2"
+
+/* Skeleton name. */
+#define YYSKELETON_NAME "yacc.c"
+
+/* Pure parsers. */
+#define YYPURE 1
+
+/* Push parsers. */
+#define YYPUSH 0
+
+/* Pull parsers. */
+#define YYPULL 1
+
+
+
+
+/* First part of user prologue. */
+#line 1 "parse-datetime.y"
+
+/* Parse a string into an internal timestamp.
+
+ Copyright (C) 1999-2000, 2002-2021 Free Software Foundation, Inc.
+
+ This program is free software: you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see <https://www.gnu.org/licenses/>. */
+
+/* Originally written by Steven M. Bellovin <smb@research.att.com> while
+ at the University of North Carolina at Chapel Hill. Later tweaked by
+ a couple of people on Usenet. Completely overhauled by Rich $alz
+ <rsalz@bbn.com> and Jim Berets <jberets@bbn.com> in August, 1990.
+
+ Modified by Assaf Gordon <assafgordon@gmail.com> in 2016 to add
+ debug output.
+
+ Modified by Paul Eggert <eggert@twinsun.com> in 1999 to do the
+ right thing about local DST. Also modified by Paul Eggert
+ <eggert@cs.ucla.edu> in 2004 to support nanosecond-resolution
+ timestamps, in 2004 to support TZ strings in dates, and in 2017 and 2020 to
+ check for integer overflow and to support longer-than-'long'
+ 'time_t' and 'tv_nsec'. */
+
+#include <config.h>
+
+#include "parse-datetime.h"
+
+#include "idx.h"
+#include "intprops.h"
+#include "timespec.h"
+#include "verify.h"
+#include "strftime.h"
+
+/* There's no need to extend the stack, so there's no need to involve
+ alloca. */
+#define YYSTACK_USE_ALLOCA 0
+
+/* Tell Bison how much stack space is needed. 20 should be plenty for
+ this grammar, which is not right recursive. Beware setting it too
+ high, since that might cause problems on machines whose
+ implementations have lame stack-overflow checking. */
+#define YYMAXDEPTH 20
+#define YYINITDEPTH YYMAXDEPTH
+
+/* Since the code of parse-datetime.y is not included in the Emacs executable
+ itself, there is no need to #define static in this file. Even if
+ the code were included in the Emacs executable, it probably
+ wouldn't do any harm to #undef it here; this will only cause
+ problems if we try to write to a static variable, which I don't
+ think this code needs to do. */
+#ifdef emacs
+# undef static
+#endif
+
+#include <inttypes.h>
+#include <c-ctype.h>
+#include <stdarg.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "gettext.h"
+
+#define _(str) gettext (str)
+
+/* Bison's skeleton tests _STDLIB_H, while some stdlib.h headers
+ use _STDLIB_H_ as witness. Map the latter to the one bison uses. */
+/* FIXME: this is temporary. Remove when we have a mechanism to ensure
+ that the version we're using is fixed, too. */
+#ifdef _STDLIB_H_
+# undef _STDLIB_H
+# define _STDLIB_H 1
+#endif
+
+/* Shift A right by B bits portably, by dividing A by 2**B and
+ truncating towards minus infinity. A and B should be free of side
+ effects, and B should be in the range 0 <= B <= INT_BITS - 2, where
+ INT_BITS is the number of useful bits in an int. GNU code can
+ assume that INT_BITS is at least 32.
+
+ ISO C99 says that A >> B is implementation-defined if A < 0. Some
+ implementations (e.g., UNICOS 9.0 on a Cray Y-MP EL) don't shift
+ right in the usual way when A < 0, so SHR falls back on division if
+ ordinary A >> B doesn't seem to be the usual signed shift. */
+#define SHR(a, b) \
+ (-1 >> 1 == -1 \
+ ? (a) >> (b) \
+ : (a) / (1 << (b)) - ((a) % (1 << (b)) < 0))
+
+#define HOUR(x) (60 * 60 * (x))
+
+#define STREQ(a, b) (strcmp (a, b) == 0)
+
+/* Verify that time_t is an integer as POSIX requires, and that every
+ time_t value fits in intmax_t. Please file a bug report if these
+ assumptions are false on your platform. */
+verify (TYPE_IS_INTEGER (time_t));
+verify (!TYPE_SIGNED (time_t) || INTMAX_MIN <= TYPE_MINIMUM (time_t));
+verify (TYPE_MAXIMUM (time_t) <= INTMAX_MAX);
+
+/* True if N is out of range for time_t. */
+static bool
+time_overflow (intmax_t n)
+{
+ return ! ((TYPE_SIGNED (time_t) ? TYPE_MINIMUM (time_t) <= n : 0 <= n)
+ && n <= TYPE_MAXIMUM (time_t));
+}
+
+/* Convert a possibly-signed character to an unsigned character. This is
+ a bit safer than casting to unsigned char, since it catches some type
+ errors that the cast doesn't. */
+static unsigned char to_uchar (char ch) { return ch; }
+
+static void _GL_ATTRIBUTE_FORMAT ((__printf__, 1, 2))
+dbg_printf (char const *msg, ...)
+{
+ va_list args;
+ /* TODO: use gnulib's 'program_name' instead? */
+ fputs ("date: ", stderr);
+
+ va_start (args, msg);
+ vfprintf (stderr, msg, args);
+ va_end (args);
+}
+
+
+/* An integer value, and the number of digits in its textual
+ representation. */
+typedef struct
+{
+ bool negative;
+ intmax_t value;
+ idx_t digits;
+} textint;
+
+/* An entry in the lexical lookup table. */
+typedef struct
+{
+ char const *name;
+ int type;
+ int value;
+} table;
+
+/* Meridian: am, pm, or 24-hour style. */
+enum { MERam, MERpm, MER24 };
+
+/* A reasonable upper bound for the buffer used in debug output. */
+enum { DBGBUFSIZE = 100 };
+
+enum { BILLION = 1000000000, LOG10_BILLION = 9 };
+
+/* Relative times. */
+typedef struct
+{
+ /* Relative year, month, day, hour, minutes, seconds, and nanoseconds. */
+ intmax_t year;
+ intmax_t month;
+ intmax_t day;
+ intmax_t hour;
+ intmax_t minutes;
+ intmax_t seconds;
+ int ns;
+} relative_time;
+
+#if HAVE_COMPOUND_LITERALS
+# define RELATIVE_TIME_0 ((relative_time) { 0, 0, 0, 0, 0, 0, 0 })
+#else
+static relative_time const RELATIVE_TIME_0;
+#endif
+
+/* Information passed to and from the parser. */
+typedef struct
+{
+ /* The input string remaining to be parsed. */
+ const char *input;
+
+ /* N, if this is the Nth Tuesday. */
+ intmax_t day_ordinal;
+
+ /* Day of week; Sunday is 0. */
+ int day_number;
+
+ /* tm_isdst flag for the local zone. */
+ int local_isdst;
+
+ /* Time zone, in seconds east of UT. */
+ int time_zone;
+
+ /* Style used for time. */
+ int meridian;
+
+ /* Gregorian year, month, day, hour, minutes, seconds, and nanoseconds. */
+ textint year;
+ intmax_t month;
+ intmax_t day;
+ intmax_t hour;
+ intmax_t minutes;
+ struct timespec seconds; /* includes nanoseconds */
+
+ /* Relative year, month, day, hour, minutes, seconds, and nanoseconds. */
+ relative_time rel;
+
+ /* Presence or counts of nonterminals of various flavors parsed so far. */
+ bool timespec_seen;
+ bool rels_seen;
+ idx_t dates_seen;
+ idx_t days_seen;
+ idx_t local_zones_seen;
+ idx_t dsts_seen;
+ idx_t times_seen;
+ idx_t zones_seen;
+ bool year_seen;
+
+#ifdef GNULIB_PARSE_DATETIME2
+ /* Print debugging output to stderr. */
+ bool parse_datetime_debug;
+#endif
+
+ /* Which of the 'seen' parts have been printed when debugging. */
+ bool debug_dates_seen;
+ bool debug_days_seen;
+ bool debug_local_zones_seen;
+ bool debug_times_seen;
+ bool debug_zones_seen;
+ bool debug_year_seen;
+
+ /* The user specified explicit ordinal day value. */
+ bool debug_ordinal_day_seen;
+
+ /* Table of local time zone abbreviations, terminated by a null entry. */
+ table local_time_zone_table[3];
+} parser_control;
+
+static bool
+debugging (parser_control const *pc)
+{
+#ifdef GNULIB_PARSE_DATETIME2
+ return pc->parse_datetime_debug;
+#else
+ return false;
+#endif
+}
+
+union YYSTYPE;
+static int yylex (union YYSTYPE *, parser_control *);
+static int yyerror (parser_control const *, char const *);
+static bool time_zone_hhmm (parser_control *, textint, intmax_t);
+
+/* Extract into *PC any date and time info from a string of digits
+ of the form e.g., YYYYMMDD, YYMMDD, HHMM, HH (and sometimes YYY,
+ YYYY, ...). */
+static void
+digits_to_date_time (parser_control *pc, textint text_int)
+{
+ if (pc->dates_seen && ! pc->year.digits
+ && ! pc->rels_seen && (pc->times_seen || 2 < text_int.digits))
+ {
+ pc->year_seen = true;
+ pc->year = text_int;
+ }
+ else
+ {
+ if (4 < text_int.digits)
+ {
+ pc->dates_seen++;
+ pc->day = text_int.value % 100;
+ pc->month = (text_int.value / 100) % 100;
+ pc->year.value = text_int.value / 10000;
+ pc->year.digits = text_int.digits - 4;
+ }
+ else
+ {
+ pc->times_seen++;
+ if (text_int.digits <= 2)
+ {
+ pc->hour = text_int.value;
+ pc->minutes = 0;
+ }
+ else
+ {
+ pc->hour = text_int.value / 100;
+ pc->minutes = text_int.value % 100;
+ }
+ pc->seconds.tv_sec = 0;
+ pc->seconds.tv_nsec = 0;
+ pc->meridian = MER24;
+ }
+ }
+}
+
+/* Increment PC->rel by FACTOR * REL (FACTOR is 1 or -1). Return true
+ if successful, false if an overflow occurred. */
+static bool
+apply_relative_time (parser_control *pc, relative_time rel, int factor)
+{
+ if (factor < 0
+ ? (INT_SUBTRACT_WRAPV (pc->rel.ns, rel.ns, &pc->rel.ns)
+ | INT_SUBTRACT_WRAPV (pc->rel.seconds, rel.seconds, &pc->rel.seconds)
+ | INT_SUBTRACT_WRAPV (pc->rel.minutes, rel.minutes, &pc->rel.minutes)
+ | INT_SUBTRACT_WRAPV (pc->rel.hour, rel.hour, &pc->rel.hour)
+ | INT_SUBTRACT_WRAPV (pc->rel.day, rel.day, &pc->rel.day)
+ | INT_SUBTRACT_WRAPV (pc->rel.month, rel.month, &pc->rel.month)
+ | INT_SUBTRACT_WRAPV (pc->rel.year, rel.year, &pc->rel.year))
+ : (INT_ADD_WRAPV (pc->rel.ns, rel.ns, &pc->rel.ns)
+ | INT_ADD_WRAPV (pc->rel.seconds, rel.seconds, &pc->rel.seconds)
+ | INT_ADD_WRAPV (pc->rel.minutes, rel.minutes, &pc->rel.minutes)
+ | INT_ADD_WRAPV (pc->rel.hour, rel.hour, &pc->rel.hour)
+ | INT_ADD_WRAPV (pc->rel.day, rel.day, &pc->rel.day)
+ | INT_ADD_WRAPV (pc->rel.month, rel.month, &pc->rel.month)
+ | INT_ADD_WRAPV (pc->rel.year, rel.year, &pc->rel.year)))
+ return false;
+ pc->rels_seen = true;
+ return true;
+}
+
+/* Set PC-> hour, minutes, seconds and nanoseconds members from arguments. */
+static void
+set_hhmmss (parser_control *pc, intmax_t hour, intmax_t minutes,
+ time_t sec, int nsec)
+{
+ pc->hour = hour;
+ pc->minutes = minutes;
+ pc->seconds.tv_sec = sec;
+ pc->seconds.tv_nsec = nsec;
+}
+
+/* Return a textual representation of the day ordinal/number values
+ in the parser_control struct (e.g., "last wed", "this tues", "thu"). */
+static const char *
+str_days (parser_control *pc, char *buffer, int n)
+{
+ /* TODO: use relative_time_table for reverse lookup. */
+ static char const ordinal_values[][11] = {
+ "last",
+ "this",
+ "next/first",
+ "(SECOND)", /* SECOND is commented out in relative_time_table. */
+ "third",
+ "fourth",
+ "fifth",
+ "sixth",
+ "seventh",
+ "eight",
+ "ninth",
+ "tenth",
+ "eleventh",
+ "twelfth"
+ };
+
+ static char const days_values[][4] = {
+ "Sun",
+ "Mon",
+ "Tue",
+ "Wed",
+ "Thu",
+ "Fri",
+ "Sat"
+ };
+
+ int len;
+
+ /* Don't add an ordinal prefix if the user didn't specify it
+ (e.g., "this wed" vs "wed"). */
+ if (pc->debug_ordinal_day_seen)
+ {
+ /* Use word description if possible (e.g., -1 = last, 3 = third). */
+ len = (-1 <= pc->day_ordinal && pc->day_ordinal <= 12
+ ? snprintf (buffer, n, "%s", ordinal_values[pc->day_ordinal + 1])
+ : snprintf (buffer, n, "%"PRIdMAX, pc->day_ordinal));
+ }
+ else
+ {
+ buffer[0] = '\0';
+ len = 0;
+ }
+
+ /* Add the day name */
+ if (0 <= pc->day_number && pc->day_number <= 6 && 0 <= len && len < n)
+ snprintf (buffer + len, n - len, &" %s"[len == 0],
+ days_values[pc->day_number]);
+ else
+ {
+ /* invalid day_number value - should never happen */
+ }
+ return buffer;
+}
+
+/* Convert a time zone to its string representation. */
+
+enum { TIME_ZONE_BUFSIZE = INT_STRLEN_BOUND (intmax_t) + sizeof ":MM:SS" } ;
+
+static char const *
+time_zone_str (int time_zone, char time_zone_buf[TIME_ZONE_BUFSIZE])
+{
+ char *p = time_zone_buf;
+ char sign = time_zone < 0 ? '-' : '+';
+ int hour = abs (time_zone / (60 * 60));
+ p += sprintf (time_zone_buf, "%c%02d", sign, hour);
+ int offset_from_hour = abs (time_zone % (60 * 60));
+ if (offset_from_hour != 0)
+ {
+ int mm = offset_from_hour / 60;
+ int ss = offset_from_hour % 60;
+ *p++ = ':';
+ *p++ = '0' + mm / 10;
+ *p++ = '0' + mm % 10;
+ if (ss)
+ {
+ *p++ = ':';
+ *p++ = '0' + ss / 10;
+ *p++ = '0' + ss % 10;
+ }
+ *p = '\0';
+ }
+ return time_zone_buf;
+}
+
+/* debugging: print the current time in the parser_control structure.
+ The parser will increment "*_seen" members for those which were parsed.
+ This function will print only newly seen parts. */
+static void
+debug_print_current_time (char const *item, parser_control *pc)
+{
+ bool space = false;
+
+ if (!debugging (pc))
+ return;
+
+ /* no newline, more items printed below */
+ dbg_printf (_("parsed %s part: "), item);
+
+ if (pc->dates_seen && !pc->debug_dates_seen)
+ {
+ /*TODO: use pc->year.negative? */
+ fprintf (stderr, "(Y-M-D) %04"PRIdMAX"-%02"PRIdMAX"-%02"PRIdMAX,
+ pc->year.value, pc->month, pc->day);
+ pc->debug_dates_seen = true;
+ space = true;
+ }
+
+ if (pc->year_seen != pc->debug_year_seen)
+ {
+ if (space)
+ fputc (' ', stderr);
+ fprintf (stderr, _("year: %04"PRIdMAX), pc->year.value);
+
+ pc->debug_year_seen = pc->year_seen;
+ space = true;
+ }
+
+ if (pc->times_seen && !pc->debug_times_seen)
+ {
+ intmax_t sec = pc->seconds.tv_sec;
+ fprintf (stderr, &" %02"PRIdMAX":%02"PRIdMAX":%02"PRIdMAX[!space],
+ pc->hour, pc->minutes, sec);
+ if (pc->seconds.tv_nsec != 0)
+ {
+ int nsec = pc->seconds.tv_nsec;
+ fprintf (stderr, ".%09d", nsec);
+ }
+ if (pc->meridian == MERpm)
+ fputs ("pm", stderr);
+
+ pc->debug_times_seen = true;
+ space = true;
+ }
+
+ if (pc->days_seen && !pc->debug_days_seen)
+ {
+ if (space)
+ fputc (' ', stderr);
+ char tmp[DBGBUFSIZE];
+ fprintf (stderr, _("%s (day ordinal=%"PRIdMAX" number=%d)"),
+ str_days (pc, tmp, sizeof tmp),
+ pc->day_ordinal, pc->day_number);
+ pc->debug_days_seen = true;
+ space = true;
+ }
+
+ /* local zone strings only change the DST settings,
+ not the timezone value. If seen, inform about the DST. */
+ if (pc->local_zones_seen && !pc->debug_local_zones_seen)
+ {
+ fprintf (stderr, &" isdst=%d%s"[!space],
+ pc->local_isdst, pc->dsts_seen ? " DST" : "");
+ pc->debug_local_zones_seen = true;
+ space = true;
+ }
+
+ if (pc->zones_seen && !pc->debug_zones_seen)
+ {
+ char time_zone_buf[TIME_ZONE_BUFSIZE];
+ fprintf (stderr, &" UTC%s"[!space],
+ time_zone_str (pc->time_zone, time_zone_buf));
+ pc->debug_zones_seen = true;
+ space = true;
+ }
+
+ if (pc->timespec_seen)
+ {
+ intmax_t sec = pc->seconds.tv_sec;
+ if (space)
+ fputc (' ', stderr);
+ fprintf (stderr, _("number of seconds: %"PRIdMAX), sec);
+ }
+
+ fputc ('\n', stderr);
+}
+
+/* Debugging: print the current relative values. */
+
+static bool
+print_rel_part (bool space, intmax_t val, char const *name)
+{
+ if (val == 0)
+ return space;
+ fprintf (stderr, &" %+"PRIdMAX" %s"[!space], val, name);
+ return true;
+}
+
+static void
+debug_print_relative_time (char const *item, parser_control const *pc)
+{
+ bool space = false;
+
+ if (!debugging (pc))
+ return;
+
+ /* no newline, more items printed below */
+ dbg_printf (_("parsed %s part: "), item);
+
+ if (pc->rel.year == 0 && pc->rel.month == 0 && pc->rel.day == 0
+ && pc->rel.hour == 0 && pc->rel.minutes == 0 && pc->rel.seconds == 0
+ && pc->rel.ns == 0)
+ {
+ /* Special case: relative time of this/today/now */
+ fputs (_("today/this/now\n"), stderr);
+ return;
+ }
+
+ space = print_rel_part (space, pc->rel.year, "year(s)");
+ space = print_rel_part (space, pc->rel.month, "month(s)");
+ space = print_rel_part (space, pc->rel.day, "day(s)");
+ space = print_rel_part (space, pc->rel.hour, "hour(s)");
+ space = print_rel_part (space, pc->rel.minutes, "minutes");
+ space = print_rel_part (space, pc->rel.seconds, "seconds");
+ print_rel_part (space, pc->rel.ns, "nanoseconds");
+
+ fputc ('\n', stderr);
+}
+
+
+
+
+#line 636 "parse-datetime.c"
+
+# ifndef YY_CAST
+# ifdef __cplusplus
+# define YY_CAST(Type, Val) static_cast<Type> (Val)
+# define YY_REINTERPRET_CAST(Type, Val) reinterpret_cast<Type> (Val)
+# else
+# define YY_CAST(Type, Val) ((Type) (Val))
+# define YY_REINTERPRET_CAST(Type, Val) ((Type) (Val))
+# endif
+# endif
+# ifndef YY_NULLPTR
+# if defined __cplusplus
+# if 201103L <= __cplusplus
+# define YY_NULLPTR nullptr
+# else
+# define YY_NULLPTR 0
+# endif
+# else
+# define YY_NULLPTR ((void*)0)
+# endif
+# endif
+
+#include "parse-datetime-gen.h"
+/* Symbol kind. */
+enum yysymbol_kind_t
+{
+ YYSYMBOL_YYEMPTY = -2,
+ YYSYMBOL_YYEOF = 0, /* "end of file" */
+ YYSYMBOL_YYerror = 1, /* error */
+ YYSYMBOL_YYUNDEF = 2, /* "invalid token" */
+ YYSYMBOL_tAGO = 3, /* tAGO */
+ YYSYMBOL_tDST = 4, /* tDST */
+ YYSYMBOL_tYEAR_UNIT = 5, /* tYEAR_UNIT */
+ YYSYMBOL_tMONTH_UNIT = 6, /* tMONTH_UNIT */
+ YYSYMBOL_tHOUR_UNIT = 7, /* tHOUR_UNIT */
+ YYSYMBOL_tMINUTE_UNIT = 8, /* tMINUTE_UNIT */
+ YYSYMBOL_tSEC_UNIT = 9, /* tSEC_UNIT */
+ YYSYMBOL_tDAY_UNIT = 10, /* tDAY_UNIT */
+ YYSYMBOL_tDAY_SHIFT = 11, /* tDAY_SHIFT */
+ YYSYMBOL_tDAY = 12, /* tDAY */
+ YYSYMBOL_tDAYZONE = 13, /* tDAYZONE */
+ YYSYMBOL_tLOCAL_ZONE = 14, /* tLOCAL_ZONE */
+ YYSYMBOL_tMERIDIAN = 15, /* tMERIDIAN */
+ YYSYMBOL_tMONTH = 16, /* tMONTH */
+ YYSYMBOL_tORDINAL = 17, /* tORDINAL */
+ YYSYMBOL_tZONE = 18, /* tZONE */
+ YYSYMBOL_tSNUMBER = 19, /* tSNUMBER */
+ YYSYMBOL_tUNUMBER = 20, /* tUNUMBER */
+ YYSYMBOL_tSDECIMAL_NUMBER = 21, /* tSDECIMAL_NUMBER */
+ YYSYMBOL_tUDECIMAL_NUMBER = 22, /* tUDECIMAL_NUMBER */
+ YYSYMBOL_23_ = 23, /* '@' */
+ YYSYMBOL_24_T_ = 24, /* 'T' */
+ YYSYMBOL_25_ = 25, /* ':' */
+ YYSYMBOL_26_ = 26, /* ',' */
+ YYSYMBOL_27_ = 27, /* '/' */
+ YYSYMBOL_YYACCEPT = 28, /* $accept */
+ YYSYMBOL_spec = 29, /* spec */
+ YYSYMBOL_timespec = 30, /* timespec */
+ YYSYMBOL_items = 31, /* items */
+ YYSYMBOL_item = 32, /* item */
+ YYSYMBOL_datetime = 33, /* datetime */
+ YYSYMBOL_iso_8601_datetime = 34, /* iso_8601_datetime */
+ YYSYMBOL_time = 35, /* time */
+ YYSYMBOL_iso_8601_time = 36, /* iso_8601_time */
+ YYSYMBOL_o_zone_offset = 37, /* o_zone_offset */
+ YYSYMBOL_zone_offset = 38, /* zone_offset */
+ YYSYMBOL_local_zone = 39, /* local_zone */
+ YYSYMBOL_zone = 40, /* zone */
+ YYSYMBOL_day = 41, /* day */
+ YYSYMBOL_date = 42, /* date */
+ YYSYMBOL_iso_8601_date = 43, /* iso_8601_date */
+ YYSYMBOL_rel = 44, /* rel */
+ YYSYMBOL_relunit = 45, /* relunit */
+ YYSYMBOL_relunit_snumber = 46, /* relunit_snumber */
+ YYSYMBOL_dayshift = 47, /* dayshift */
+ YYSYMBOL_seconds = 48, /* seconds */
+ YYSYMBOL_signed_seconds = 49, /* signed_seconds */
+ YYSYMBOL_unsigned_seconds = 50, /* unsigned_seconds */
+ YYSYMBOL_number = 51, /* number */
+ YYSYMBOL_hybrid = 52, /* hybrid */
+ YYSYMBOL_o_colon_minutes = 53 /* o_colon_minutes */
+};
+typedef enum yysymbol_kind_t yysymbol_kind_t;
+
+
+
+
+#ifdef short
+# undef short
+#endif
+
+/* On compilers that do not define __PTRDIFF_MAX__ etc., make sure
+ <limits.h> and (if available) <stdint.h> are included
+ so that the code can choose integer types of a good width. */
+
+#ifndef __PTRDIFF_MAX__
+# include <limits.h> /* INFRINGES ON USER NAME SPACE */
+# if defined __STDC_VERSION__ && 199901 <= __STDC_VERSION__
+# include <stdint.h> /* INFRINGES ON USER NAME SPACE */
+# define YY_STDINT_H
+# endif
+#endif
+
+/* Narrow types that promote to a signed type and that can represent a
+ signed or unsigned integer of at least N bits. In tables they can
+ save space and decrease cache pressure. Promoting to a signed type
+ helps avoid bugs in integer arithmetic. */
+
+#ifdef __INT_LEAST8_MAX__
+typedef __INT_LEAST8_TYPE__ yytype_int8;
+#elif defined YY_STDINT_H
+typedef int_least8_t yytype_int8;
+#else
+typedef signed char yytype_int8;
+#endif
+
+#ifdef __INT_LEAST16_MAX__
+typedef __INT_LEAST16_TYPE__ yytype_int16;
+#elif defined YY_STDINT_H
+typedef int_least16_t yytype_int16;
+#else
+typedef short yytype_int16;
+#endif
+
+/* Work around bug in HP-UX 11.23, which defines these macros
+ incorrectly for preprocessor constants. This workaround can likely
+ be removed in 2023, as HPE has promised support for HP-UX 11.23
+ (aka HP-UX 11i v2) only through the end of 2022; see Table 2 of
+ <https://h20195.www2.hpe.com/V2/getpdf.aspx/4AA4-7673ENW.pdf>. */
+#ifdef __hpux
+# undef UINT_LEAST8_MAX
+# undef UINT_LEAST16_MAX
+# define UINT_LEAST8_MAX 255
+# define UINT_LEAST16_MAX 65535
+#endif
+
+#if defined __UINT_LEAST8_MAX__ && __UINT_LEAST8_MAX__ <= __INT_MAX__
+typedef __UINT_LEAST8_TYPE__ yytype_uint8;
+#elif (!defined __UINT_LEAST8_MAX__ && defined YY_STDINT_H \
+ && UINT_LEAST8_MAX <= INT_MAX)
+typedef uint_least8_t yytype_uint8;
+#elif !defined __UINT_LEAST8_MAX__ && UCHAR_MAX <= INT_MAX
+typedef unsigned char yytype_uint8;
+#else
+typedef short yytype_uint8;
+#endif
+
+#if defined __UINT_LEAST16_MAX__ && __UINT_LEAST16_MAX__ <= __INT_MAX__
+typedef __UINT_LEAST16_TYPE__ yytype_uint16;
+#elif (!defined __UINT_LEAST16_MAX__ && defined YY_STDINT_H \
+ && UINT_LEAST16_MAX <= INT_MAX)
+typedef uint_least16_t yytype_uint16;
+#elif !defined __UINT_LEAST16_MAX__ && USHRT_MAX <= INT_MAX
+typedef unsigned short yytype_uint16;
+#else
+typedef int yytype_uint16;
+#endif
+
+#ifndef YYPTRDIFF_T
+# if defined __PTRDIFF_TYPE__ && defined __PTRDIFF_MAX__
+# define YYPTRDIFF_T __PTRDIFF_TYPE__
+# define YYPTRDIFF_MAXIMUM __PTRDIFF_MAX__
+# elif defined PTRDIFF_MAX
+# ifndef ptrdiff_t
+# include <stddef.h> /* INFRINGES ON USER NAME SPACE */
+# endif
+# define YYPTRDIFF_T ptrdiff_t
+# define YYPTRDIFF_MAXIMUM PTRDIFF_MAX
+# else
+# define YYPTRDIFF_T long
+# define YYPTRDIFF_MAXIMUM LONG_MAX
+# endif
+#endif
+
+#ifndef YYSIZE_T
+# ifdef __SIZE_TYPE__
+# define YYSIZE_T __SIZE_TYPE__
+# elif defined size_t
+# define YYSIZE_T size_t
+# elif defined __STDC_VERSION__ && 199901 <= __STDC_VERSION__
+# include <stddef.h> /* INFRINGES ON USER NAME SPACE */
+# define YYSIZE_T size_t
+# else
+# define YYSIZE_T unsigned
+# endif
+#endif
+
+#define YYSIZE_MAXIMUM \
+ YY_CAST (YYPTRDIFF_T, \
+ (YYPTRDIFF_MAXIMUM < YY_CAST (YYSIZE_T, -1) \
+ ? YYPTRDIFF_MAXIMUM \
+ : YY_CAST (YYSIZE_T, -1)))
+
+#define YYSIZEOF(X) YY_CAST (YYPTRDIFF_T, sizeof (X))
+
+
+/* Stored state numbers (used for stacks). */
+typedef yytype_int8 yy_state_t;
+
+/* State numbers in computations. */
+typedef int yy_state_fast_t;
+
+#ifndef YY_
+# if defined YYENABLE_NLS && YYENABLE_NLS
+# if ENABLE_NLS
+# include <libintl.h> /* INFRINGES ON USER NAME SPACE */
+# define YY_(Msgid) dgettext ("bison-runtime", Msgid)
+# endif
+# endif
+# ifndef YY_
+# define YY_(Msgid) Msgid
+# endif
+#endif
+
+
+#ifndef YY_ATTRIBUTE_PURE
+# if defined __GNUC__ && 2 < __GNUC__ + (96 <= __GNUC_MINOR__)
+# define YY_ATTRIBUTE_PURE __attribute__ ((__pure__))
+# else
+# define YY_ATTRIBUTE_PURE
+# endif
+#endif
+
+#ifndef YY_ATTRIBUTE_UNUSED
+# if defined __GNUC__ && 2 < __GNUC__ + (7 <= __GNUC_MINOR__)
+# define YY_ATTRIBUTE_UNUSED __attribute__ ((__unused__))
+# else
+# define YY_ATTRIBUTE_UNUSED
+# endif
+#endif
+
+/* Suppress unused-variable warnings by "using" E. */
+#if ! defined lint || defined __GNUC__
+# define YY_USE(E) ((void) (E))
+#else
+# define YY_USE(E) /* empty */
+#endif
+
+/* Suppress an incorrect diagnostic about yylval being uninitialized. */
+#if defined __GNUC__ && ! defined __ICC && 406 <= __GNUC__ * 100 + __GNUC_MINOR__
+# if __GNUC__ * 100 + __GNUC_MINOR__ < 407
+# define YY_IGNORE_MAYBE_UNINITIALIZED_BEGIN \
+ _Pragma ("GCC diagnostic push") \
+ _Pragma ("GCC diagnostic ignored \"-Wuninitialized\"")
+# else
+# define YY_IGNORE_MAYBE_UNINITIALIZED_BEGIN \
+ _Pragma ("GCC diagnostic push") \
+ _Pragma ("GCC diagnostic ignored \"-Wuninitialized\"") \
+ _Pragma ("GCC diagnostic ignored \"-Wmaybe-uninitialized\"")
+# endif
+# define YY_IGNORE_MAYBE_UNINITIALIZED_END \
+ _Pragma ("GCC diagnostic pop")
+#else
+# define YY_INITIAL_VALUE(Value) Value
+#endif
+#ifndef YY_IGNORE_MAYBE_UNINITIALIZED_BEGIN
+# define YY_IGNORE_MAYBE_UNINITIALIZED_BEGIN
+# define YY_IGNORE_MAYBE_UNINITIALIZED_END
+#endif
+#ifndef YY_INITIAL_VALUE
+# define YY_INITIAL_VALUE(Value) /* Nothing. */
+#endif
+
+#if defined __cplusplus && defined __GNUC__ && ! defined __ICC && 6 <= __GNUC__
+# define YY_IGNORE_USELESS_CAST_BEGIN \
+ _Pragma ("GCC diagnostic push") \
+ _Pragma ("GCC diagnostic ignored \"-Wuseless-cast\"")
+# define YY_IGNORE_USELESS_CAST_END \
+ _Pragma ("GCC diagnostic pop")
+#endif
+#ifndef YY_IGNORE_USELESS_CAST_BEGIN
+# define YY_IGNORE_USELESS_CAST_BEGIN
+# define YY_IGNORE_USELESS_CAST_END
+#endif
+
+
+#define YY_ASSERT(E) ((void) (0 && (E)))
+
+#if !defined yyoverflow
+
+/* The parser invokes alloca or malloc; define the necessary symbols. */
+
+# ifdef YYSTACK_USE_ALLOCA
+# if YYSTACK_USE_ALLOCA
+# ifdef __GNUC__
+# define YYSTACK_ALLOC __builtin_alloca
+# elif defined __BUILTIN_VA_ARG_INCR
+# include <alloca.h> /* INFRINGES ON USER NAME SPACE */
+# elif defined _AIX
+# define YYSTACK_ALLOC __alloca
+# elif defined _MSC_VER
+# include <malloc.h> /* INFRINGES ON USER NAME SPACE */
+# define alloca _alloca
+# else
+# define YYSTACK_ALLOC alloca
+# if ! defined _ALLOCA_H && ! defined EXIT_SUCCESS
+# include <stdlib.h> /* INFRINGES ON USER NAME SPACE */
+ /* Use EXIT_SUCCESS as a witness for stdlib.h. */
+# ifndef EXIT_SUCCESS
+# define EXIT_SUCCESS 0
+# endif
+# endif
+# endif
+# endif
+# endif
+
+# ifdef YYSTACK_ALLOC
+ /* Pacify GCC's 'empty if-body' warning. */
+# define YYSTACK_FREE(Ptr) do { /* empty */; } while (0)
+# ifndef YYSTACK_ALLOC_MAXIMUM
+ /* The OS might guarantee only one guard page at the bottom of the stack,
+ and a page size can be as small as 4096 bytes. So we cannot safely
+ invoke alloca (N) if N exceeds 4096. Use a slightly smaller number
+ to allow for a few compiler-allocated temporary stack slots. */
+# define YYSTACK_ALLOC_MAXIMUM 4032 /* reasonable circa 2006 */
+# endif
+# else
+# define YYSTACK_ALLOC YYMALLOC
+# define YYSTACK_FREE YYFREE
+# ifndef YYSTACK_ALLOC_MAXIMUM
+# define YYSTACK_ALLOC_MAXIMUM YYSIZE_MAXIMUM
+# endif
+# if (defined __cplusplus && ! defined EXIT_SUCCESS \
+ && ! ((defined YYMALLOC || defined malloc) \
+ && (defined YYFREE || defined free)))
+# include <stdlib.h> /* INFRINGES ON USER NAME SPACE */
+# ifndef EXIT_SUCCESS
+# define EXIT_SUCCESS 0
+# endif
+# endif
+# ifndef YYMALLOC
+# define YYMALLOC malloc
+# if ! defined malloc && ! defined EXIT_SUCCESS
+void *malloc (YYSIZE_T); /* INFRINGES ON USER NAME SPACE */
+# endif
+# endif
+# ifndef YYFREE
+# define YYFREE free
+# if ! defined free && ! defined EXIT_SUCCESS
+void free (void *); /* INFRINGES ON USER NAME SPACE */
+# endif
+# endif
+# endif
+#endif /* !defined yyoverflow */
+
+#if (! defined yyoverflow \
+ && (! defined __cplusplus \
+ || (defined YYSTYPE_IS_TRIVIAL && YYSTYPE_IS_TRIVIAL)))
+
+/* A type that is properly aligned for any stack member. */
+union yyalloc
+{
+ yy_state_t yyss_alloc;
+ YYSTYPE yyvs_alloc;
+};
+
+/* The size of the maximum gap between one aligned stack and the next. */
+# define YYSTACK_GAP_MAXIMUM (YYSIZEOF (union yyalloc) - 1)
+
+/* The size of an array large to enough to hold all stacks, each with
+ N elements. */
+# define YYSTACK_BYTES(N) \
+ ((N) * (YYSIZEOF (yy_state_t) + YYSIZEOF (YYSTYPE)) \
+ + YYSTACK_GAP_MAXIMUM)
+
+# define YYCOPY_NEEDED 1
+
+/* Relocate STACK from its old location to the new one. The
+ local variables YYSIZE and YYSTACKSIZE give the old and new number of
+ elements in the stack, and YYPTR gives the new location of the
+ stack. Advance YYPTR to a properly aligned location for the next
+ stack. */
+# define YYSTACK_RELOCATE(Stack_alloc, Stack) \
+ do \
+ { \
+ YYPTRDIFF_T yynewbytes; \
+ YYCOPY (&yyptr->Stack_alloc, Stack, yysize); \
+ Stack = &yyptr->Stack_alloc; \
+ yynewbytes = yystacksize * YYSIZEOF (*Stack) + YYSTACK_GAP_MAXIMUM; \
+ yyptr += yynewbytes / YYSIZEOF (*yyptr); \
+ } \
+ while (0)
+
+#endif
+
+#if defined YYCOPY_NEEDED && YYCOPY_NEEDED
+/* Copy COUNT objects from SRC to DST. The source and destination do
+ not overlap. */
+# ifndef YYCOPY
+# if defined __GNUC__ && 1 < __GNUC__
+# define YYCOPY(Dst, Src, Count) \
+ __builtin_memcpy (Dst, Src, YY_CAST (YYSIZE_T, (Count)) * sizeof (*(Src)))
+# else
+# define YYCOPY(Dst, Src, Count) \
+ do \
+ { \
+ YYPTRDIFF_T yyi; \
+ for (yyi = 0; yyi < (Count); yyi++) \
+ (Dst)[yyi] = (Src)[yyi]; \
+ } \
+ while (0)
+# endif
+# endif
+#endif /* !YYCOPY_NEEDED */
+
+/* YYFINAL -- State number of the termination state. */
+#define YYFINAL 12
+/* YYLAST -- Last index in YYTABLE. */
+#define YYLAST 112
+
+/* YYNTOKENS -- Number of terminals. */
+#define YYNTOKENS 28
+/* YYNNTS -- Number of nonterminals. */
+#define YYNNTS 26
+/* YYNRULES -- Number of rules. */
+#define YYNRULES 91
+/* YYNSTATES -- Number of states. */
+#define YYNSTATES 114
+
+/* YYMAXUTOK -- Last valid token kind. */
+#define YYMAXUTOK 277
+
+
+/* YYTRANSLATE(TOKEN-NUM) -- Symbol number corresponding to TOKEN-NUM
+ as returned by yylex, with out-of-bounds checking. */
+#define YYTRANSLATE(YYX) \
+ (0 <= (YYX) && (YYX) <= YYMAXUTOK \
+ ? YY_CAST (yysymbol_kind_t, yytranslate[YYX]) \
+ : YYSYMBOL_YYUNDEF)
+
+/* YYTRANSLATE[TOKEN-NUM] -- Symbol number corresponding to TOKEN-NUM
+ as returned by yylex. */
+static const yytype_int8 yytranslate[] =
+{
+ 0, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 26, 2, 2, 27, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 25, 2,
+ 2, 2, 2, 2, 23, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 24, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+ 2, 2, 2, 2, 2, 2, 1, 2, 3, 4,
+ 5, 6, 7, 8, 9, 10, 11, 12, 13, 14,
+ 15, 16, 17, 18, 19, 20, 21, 22
+};
+
+#if YYDEBUG
+/* YYRLINE[YYN] -- Source line where rule number YYN was defined. */
+static const yytype_int16 yyrline[] =
+{
+ 0, 603, 603, 604, 608, 616, 618, 622, 627, 632,
+ 637, 642, 647, 652, 656, 660, 667, 671, 675, 680,
+ 685, 690, 694, 699, 704, 711, 713, 717, 742, 744,
+ 754, 756, 758, 763, 768, 771, 773, 778, 783, 788,
+ 794, 803, 808, 841, 849, 857, 862, 868, 873, 879,
+ 883, 893, 895, 897, 902, 904, 906, 908, 910, 912,
+ 914, 917, 920, 922, 924, 926, 928, 930, 932, 934,
+ 936, 938, 940, 942, 944, 948, 950, 952, 955, 957,
+ 959, 964, 968, 968, 971, 972, 978, 979, 985, 990,
+ 1001, 1002
+};
+#endif
+
+/** Accessing symbol of state STATE. */
+#define YY_ACCESSING_SYMBOL(State) YY_CAST (yysymbol_kind_t, yystos[State])
+
+#if YYDEBUG || 0
+/* The user-facing name of the symbol whose (internal) number is
+ YYSYMBOL. No bounds checking. */
+static const char *yysymbol_name (yysymbol_kind_t yysymbol) YY_ATTRIBUTE_UNUSED;
+
+/* YYTNAME[SYMBOL-NUM] -- String name of the symbol SYMBOL-NUM.
+ First, the terminals, then, starting at YYNTOKENS, nonterminals. */
+static const char *const yytname[] =
+{
+ "\"end of file\"", "error", "\"invalid token\"", "tAGO", "tDST",
+ "tYEAR_UNIT", "tMONTH_UNIT", "tHOUR_UNIT", "tMINUTE_UNIT", "tSEC_UNIT",
+ "tDAY_UNIT", "tDAY_SHIFT", "tDAY", "tDAYZONE", "tLOCAL_ZONE",
+ "tMERIDIAN", "tMONTH", "tORDINAL", "tZONE", "tSNUMBER", "tUNUMBER",
+ "tSDECIMAL_NUMBER", "tUDECIMAL_NUMBER", "'@'", "'T'", "':'", "','",
+ "'/'", "$accept", "spec", "timespec", "items", "item", "datetime",
+ "iso_8601_datetime", "time", "iso_8601_time", "o_zone_offset",
+ "zone_offset", "local_zone", "zone", "day", "date", "iso_8601_date",
+ "rel", "relunit", "relunit_snumber", "dayshift", "seconds",
+ "signed_seconds", "unsigned_seconds", "number", "hybrid",
+ "o_colon_minutes", YY_NULLPTR
+};
+
+static const char *
+yysymbol_name (yysymbol_kind_t yysymbol)
+{
+ return yytname[yysymbol];
+}
+#endif
+
+#define YYPACT_NINF (-93)
+
+#define yypact_value_is_default(Yyn) \
+ ((Yyn) == YYPACT_NINF)
+
+#define YYTABLE_NINF (-1)
+
+#define yytable_value_is_error(Yyn) \
+ 0
+
+/* YYPACT[STATE-NUM] -- Index in YYTABLE of the portion describing
+ STATE-NUM. */
+static const yytype_int8 yypact[] =
+{
+ 38, 27, 77, -93, 46, -93, -93, -93, -93, -93,
+ -93, -93, -93, -93, -93, -93, -93, -93, -93, -93,
+ 62, -93, 82, -3, 66, 3, 74, -4, 83, 84,
+ 75, -93, -93, -93, -93, -93, -93, -93, -93, -93,
+ 71, -93, 93, -93, -93, -93, -93, -93, -93, 78,
+ 72, -93, -93, -93, -93, -93, -93, -93, -93, 25,
+ -93, -93, -93, -93, -93, -93, -93, -93, -93, -93,
+ -93, -93, -93, -93, -93, 21, 19, 79, 80, -93,
+ -93, -93, -93, -93, 81, -93, -93, 85, 86, -93,
+ -93, -93, -93, -93, -6, 76, 17, -93, -93, -93,
+ -93, 87, 69, -93, -93, 88, 89, -1, -93, 18,
+ -93, -93, 69, 91
+};
+
+/* YYDEFACT[STATE-NUM] -- Default reduction number in state STATE-NUM.
+ Performed when YYTABLE does not specify something else to do. Zero
+ means the default is an error. */
+static const yytype_int8 yydefact[] =
+{
+ 5, 0, 0, 2, 3, 85, 87, 84, 86, 4,
+ 82, 83, 1, 56, 59, 65, 68, 73, 62, 81,
+ 37, 35, 28, 0, 0, 30, 0, 88, 0, 0,
+ 31, 6, 7, 16, 8, 21, 9, 10, 12, 11,
+ 49, 13, 52, 74, 53, 14, 15, 38, 29, 0,
+ 45, 54, 57, 63, 66, 69, 60, 39, 36, 90,
+ 32, 75, 76, 78, 79, 80, 77, 55, 58, 64,
+ 67, 70, 61, 40, 18, 47, 90, 0, 0, 22,
+ 89, 71, 72, 33, 0, 51, 44, 0, 0, 34,
+ 43, 48, 50, 27, 25, 41, 0, 17, 46, 91,
+ 19, 90, 0, 23, 26, 0, 0, 25, 42, 25,
+ 20, 24, 0, 25
+};
+
+/* YYPGOTO[NTERM-NUM]. */
+static const yytype_int8 yypgoto[] =
+{
+ -93, -93, -93, -93, -93, -93, -93, -93, 20, -68,
+ -27, -93, -93, -93, -93, -93, -93, -93, 60, -93,
+ -93, -93, -92, -93, -93, 43
+};
+
+/* YYDEFGOTO[NTERM-NUM]. */
+static const yytype_int8 yydefgoto[] =
+{
+ 0, 2, 3, 4, 31, 32, 33, 34, 35, 103,
+ 104, 36, 37, 38, 39, 40, 41, 42, 43, 44,
+ 9, 10, 11, 45, 46, 93
+};
+
+/* YYTABLE[YYPACT[STATE-NUM]] -- What to do in state STATE-NUM. If
+ positive, shift that token. If negative, reduce the rule whose
+ number is the opposite. If YYTABLE_NINF, syntax error. */
+static const yytype_int8 yytable[] =
+{
+ 79, 67, 68, 69, 70, 71, 72, 58, 73, 100,
+ 107, 74, 75, 101, 110, 76, 49, 50, 101, 102,
+ 113, 77, 59, 78, 61, 62, 63, 64, 65, 66,
+ 61, 62, 63, 64, 65, 66, 101, 101, 92, 111,
+ 90, 91, 106, 112, 88, 111, 5, 6, 7, 8,
+ 88, 13, 14, 15, 16, 17, 18, 19, 20, 21,
+ 22, 1, 23, 24, 25, 26, 27, 28, 29, 79,
+ 30, 51, 52, 53, 54, 55, 56, 12, 57, 61,
+ 62, 63, 64, 65, 66, 60, 48, 80, 47, 6,
+ 83, 8, 81, 82, 26, 84, 85, 86, 87, 94,
+ 95, 96, 89, 105, 97, 98, 99, 0, 108, 109,
+ 101, 0, 88
+};
+
+static const yytype_int8 yycheck[] =
+{
+ 27, 5, 6, 7, 8, 9, 10, 4, 12, 15,
+ 102, 15, 16, 19, 15, 19, 19, 20, 19, 25,
+ 112, 25, 19, 27, 5, 6, 7, 8, 9, 10,
+ 5, 6, 7, 8, 9, 10, 19, 19, 19, 107,
+ 19, 20, 25, 25, 25, 113, 19, 20, 21, 22,
+ 25, 5, 6, 7, 8, 9, 10, 11, 12, 13,
+ 14, 23, 16, 17, 18, 19, 20, 21, 22, 96,
+ 24, 5, 6, 7, 8, 9, 10, 0, 12, 5,
+ 6, 7, 8, 9, 10, 25, 4, 27, 26, 20,
+ 30, 22, 9, 9, 19, 24, 3, 19, 26, 20,
+ 20, 20, 59, 27, 84, 20, 20, -1, 20, 20,
+ 19, -1, 25
+};
+
+/* YYSTOS[STATE-NUM] -- The symbol kind of the accessing symbol of
+ state STATE-NUM. */
+static const yytype_int8 yystos[] =
+{
+ 0, 23, 29, 30, 31, 19, 20, 21, 22, 48,
+ 49, 50, 0, 5, 6, 7, 8, 9, 10, 11,
+ 12, 13, 14, 16, 17, 18, 19, 20, 21, 22,
+ 24, 32, 33, 34, 35, 36, 39, 40, 41, 42,
+ 43, 44, 45, 46, 47, 51, 52, 26, 4, 19,
+ 20, 5, 6, 7, 8, 9, 10, 12, 4, 19,
+ 46, 5, 6, 7, 8, 9, 10, 5, 6, 7,
+ 8, 9, 10, 12, 15, 16, 19, 25, 27, 38,
+ 46, 9, 9, 46, 24, 3, 19, 26, 25, 53,
+ 19, 20, 19, 53, 20, 20, 20, 36, 20, 20,
+ 15, 19, 25, 37, 38, 27, 25, 50, 20, 20,
+ 15, 37, 25, 50
+};
+
+/* YYR1[RULE-NUM] -- Symbol kind of the left-hand side of rule RULE-NUM. */
+static const yytype_int8 yyr1[] =
+{
+ 0, 28, 29, 29, 30, 31, 31, 32, 32, 32,
+ 32, 32, 32, 32, 32, 32, 33, 34, 35, 35,
+ 35, 35, 36, 36, 36, 37, 37, 38, 39, 39,
+ 40, 40, 40, 40, 40, 40, 40, 41, 41, 41,
+ 41, 42, 42, 42, 42, 42, 42, 42, 42, 42,
+ 43, 44, 44, 44, 45, 45, 45, 45, 45, 45,
+ 45, 45, 45, 45, 45, 45, 45, 45, 45, 45,
+ 45, 45, 45, 45, 45, 46, 46, 46, 46, 46,
+ 46, 47, 48, 48, 49, 49, 50, 50, 51, 52,
+ 53, 53
+};
+
+/* YYR2[RULE-NUM] -- Number of symbols on the right-hand side of rule RULE-NUM. */
+static const yytype_int8 yyr2[] =
+{
+ 0, 2, 1, 1, 2, 0, 2, 1, 1, 1,
+ 1, 1, 1, 1, 1, 1, 1, 3, 2, 4,
+ 6, 1, 2, 4, 6, 0, 1, 2, 1, 2,
+ 1, 1, 2, 2, 3, 1, 2, 1, 2, 2,
+ 2, 3, 5, 3, 3, 2, 4, 2, 3, 1,
+ 3, 2, 1, 1, 2, 2, 1, 2, 2, 1,
+ 2, 2, 1, 2, 2, 1, 2, 2, 1, 2,
+ 2, 2, 2, 1, 1, 2, 2, 2, 2, 2,
+ 2, 1, 1, 1, 1, 1, 1, 1, 1, 2,
+ 0, 2
+};
+
+
+enum { YYENOMEM = -2 };
+
+#define yyerrok (yyerrstatus = 0)
+#define yyclearin (yychar = YYEMPTY)
+
+#define YYACCEPT goto yyacceptlab
+#define YYABORT goto yyabortlab
+#define YYERROR goto yyerrorlab
+#define YYNOMEM goto yyexhaustedlab
+
+
+#define YYRECOVERING() (!!yyerrstatus)
+
+#define YYBACKUP(Token, Value) \
+ do \
+ if (yychar == YYEMPTY) \
+ { \
+ yychar = (Token); \
+ yylval = (Value); \
+ YYPOPSTACK (yylen); \
+ yystate = *yyssp; \
+ goto yybackup; \
+ } \
+ else \
+ { \
+ yyerror (pc, YY_("syntax error: cannot back up")); \
+ YYERROR; \
+ } \
+ while (0)
+
+/* Backward compatibility with an undocumented macro.
+ Use YYerror or YYUNDEF. */
+#define YYERRCODE YYUNDEF
+
+
+/* Enable debugging if requested. */
+#if YYDEBUG
+
+# ifndef YYFPRINTF
+# include <stdio.h> /* INFRINGES ON USER NAME SPACE */
+# define YYFPRINTF fprintf
+# endif
+
+# define YYDPRINTF(Args) \
+do { \
+ if (yydebug) \
+ YYFPRINTF Args; \
+} while (0)
+
+
+
+
+# define YY_SYMBOL_PRINT(Title, Kind, Value, Location) \
+do { \
+ if (yydebug) \
+ { \
+ YYFPRINTF (stderr, "%s ", Title); \
+ yy_symbol_print (stderr, \
+ Kind, Value, pc); \
+ YYFPRINTF (stderr, "\n"); \
+ } \
+} while (0)
+
+
+/*-----------------------------------.
+| Print this symbol's value on YYO. |
+`-----------------------------------*/
+
+static void
+yy_symbol_value_print (FILE *yyo,
+ yysymbol_kind_t yykind, YYSTYPE const * const yyvaluep, parser_control *pc)
+{
+ FILE *yyoutput = yyo;
+ YY_USE (yyoutput);
+ YY_USE (pc);
+ if (!yyvaluep)
+ return;
+ YY_IGNORE_MAYBE_UNINITIALIZED_BEGIN
+ YY_USE (yykind);
+ YY_IGNORE_MAYBE_UNINITIALIZED_END
+}
+
+
+/*---------------------------.
+| Print this symbol on YYO. |
+`---------------------------*/
+
+static void
+yy_symbol_print (FILE *yyo,
+ yysymbol_kind_t yykind, YYSTYPE const * const yyvaluep, parser_control *pc)
+{
+ YYFPRINTF (yyo, "%s %s (",
+ yykind < YYNTOKENS ? "token" : "nterm", yysymbol_name (yykind));
+
+ yy_symbol_value_print (yyo, yykind, yyvaluep, pc);
+ YYFPRINTF (yyo, ")");
+}
+
+/*------------------------------------------------------------------.
+| yy_stack_print -- Print the state stack from its BOTTOM up to its |
+| TOP (included). |
+`------------------------------------------------------------------*/
+
+static void
+yy_stack_print (yy_state_t *yybottom, yy_state_t *yytop)
+{
+ YYFPRINTF (stderr, "Stack now");
+ for (; yybottom <= yytop; yybottom++)
+ {
+ int yybot = *yybottom;
+ YYFPRINTF (stderr, " %d", yybot);
+ }
+ YYFPRINTF (stderr, "\n");
+}
+
+# define YY_STACK_PRINT(Bottom, Top) \
+do { \
+ if (yydebug) \
+ yy_stack_print ((Bottom), (Top)); \
+} while (0)
+
+
+/*------------------------------------------------.
+| Report that the YYRULE is going to be reduced. |
+`------------------------------------------------*/
+
+static void
+yy_reduce_print (yy_state_t *yyssp, YYSTYPE *yyvsp,
+ int yyrule, parser_control *pc)
+{
+ int yylno = yyrline[yyrule];
+ int yynrhs = yyr2[yyrule];
+ int yyi;
+ YYFPRINTF (stderr, "Reducing stack by rule %d (line %d):\n",
+ yyrule - 1, yylno);
+ /* The symbols being reduced. */
+ for (yyi = 0; yyi < yynrhs; yyi++)
+ {
+ YYFPRINTF (stderr, " $%d = ", yyi + 1);
+ yy_symbol_print (stderr,
+ YY_ACCESSING_SYMBOL (+yyssp[yyi + 1 - yynrhs]),
+ &yyvsp[(yyi + 1) - (yynrhs)], pc);
+ YYFPRINTF (stderr, "\n");
+ }
+}
+
+# define YY_REDUCE_PRINT(Rule) \
+do { \
+ if (yydebug) \
+ yy_reduce_print (yyssp, yyvsp, Rule, pc); \
+} while (0)
+
+/* Nonzero means print parse trace. It is left uninitialized so that
+ multiple parsers can coexist. */
+int yydebug;
+#else /* !YYDEBUG */
+# define YYDPRINTF(Args) ((void) 0)
+# define YY_SYMBOL_PRINT(Title, Kind, Value, Location)
+# define YY_STACK_PRINT(Bottom, Top)
+# define YY_REDUCE_PRINT(Rule)
+#endif /* !YYDEBUG */
+
+
+/* YYINITDEPTH -- initial size of the parser's stacks. */
+#ifndef YYINITDEPTH
+# define YYINITDEPTH 200
+#endif
+
+/* YYMAXDEPTH -- maximum size the stacks can grow to (effective only
+ if the built-in stack extension method is used).
+
+ Do not make this value too large; the results are undefined if
+ YYSTACK_ALLOC_MAXIMUM < YYSTACK_BYTES (YYMAXDEPTH)
+ evaluated with infinite-precision integer arithmetic. */
+
+#ifndef YYMAXDEPTH
+# define YYMAXDEPTH 10000
+#endif
+
+
+
+
+
+
+/*-----------------------------------------------.
+| Release the memory associated to this symbol. |
+`-----------------------------------------------*/
+
+static void
+yydestruct (const char *yymsg,
+ yysymbol_kind_t yykind, YYSTYPE *yyvaluep, parser_control *pc)
+{
+ YY_USE (yyvaluep);
+ YY_USE (pc);
+ if (!yymsg)
+ yymsg = "Deleting";
+ YY_SYMBOL_PRINT (yymsg, yykind, yyvaluep, yylocationp);
+
+ YY_IGNORE_MAYBE_UNINITIALIZED_BEGIN
+ YY_USE (yykind);
+ YY_IGNORE_MAYBE_UNINITIALIZED_END
+}
+
+
+
+
+
+
+/*----------.
+| yyparse. |
+`----------*/
+
+int
+yyparse (parser_control *pc)
+{
+/* Lookahead token kind. */
+int yychar;
+
+
+/* The semantic value of the lookahead symbol. */
+/* Default value used for initialization, for pacifying older GCCs
+ or non-GCC compilers. */
+YY_INITIAL_VALUE (static YYSTYPE yyval_default;)
+YYSTYPE yylval YY_INITIAL_VALUE (= yyval_default);
+
+ /* Number of syntax errors so far. */
+ int yynerrs = 0;
+
+ yy_state_fast_t yystate = 0;
+ /* Number of tokens to shift before error messages enabled. */
+ int yyerrstatus = 0;
+
+ /* Refer to the stacks through separate pointers, to allow yyoverflow
+ to reallocate them elsewhere. */
+
+ /* Their size. */
+ YYPTRDIFF_T yystacksize = YYINITDEPTH;
+
+ /* The state stack: array, bottom, top. */
+ yy_state_t yyssa[YYINITDEPTH];
+ yy_state_t *yyss = yyssa;
+ yy_state_t *yyssp = yyss;
+
+ /* The semantic value stack: array, bottom, top. */
+ YYSTYPE yyvsa[YYINITDEPTH];
+ YYSTYPE *yyvs = yyvsa;
+ YYSTYPE *yyvsp = yyvs;
+
+ int yyn;
+ /* The return value of yyparse. */
+ int yyresult;
+ /* Lookahead symbol kind. */
+ yysymbol_kind_t yytoken = YYSYMBOL_YYEMPTY;
+ /* The variables used to return semantic value and location from the
+ action routines. */
+ YYSTYPE yyval;
+
+
+
+#define YYPOPSTACK(N) (yyvsp -= (N), yyssp -= (N))
+
+ /* The number of symbols on the RHS of the reduced rule.
+ Keep to zero when no symbol should be popped. */
+ int yylen = 0;
+
+ YYDPRINTF ((stderr, "Starting parse\n"));
+
+ yychar = YYEMPTY; /* Cause a token to be read. */
+
+ goto yysetstate;
+
+
+/*------------------------------------------------------------.
+| yynewstate -- push a new state, which is found in yystate. |
+`------------------------------------------------------------*/
+yynewstate:
+ /* In all cases, when you get here, the value and location stacks
+ have just been pushed. So pushing a state here evens the stacks. */
+ yyssp++;
+
+
+/*--------------------------------------------------------------------.
+| yysetstate -- set current state (the top of the stack) to yystate. |
+`--------------------------------------------------------------------*/
+yysetstate:
+ YYDPRINTF ((stderr, "Entering state %d\n", yystate));
+ YY_ASSERT (0 <= yystate && yystate < YYNSTATES);
+ YY_IGNORE_USELESS_CAST_BEGIN
+ *yyssp = YY_CAST (yy_state_t, yystate);
+ YY_IGNORE_USELESS_CAST_END
+ YY_STACK_PRINT (yyss, yyssp);
+
+ if (yyss + yystacksize - 1 <= yyssp)
+#if !defined yyoverflow && !defined YYSTACK_RELOCATE
+ YYNOMEM;
+#else
+ {
+ /* Get the current used size of the three stacks, in elements. */
+ YYPTRDIFF_T yysize = yyssp - yyss + 1;
+
+# if defined yyoverflow
+ {
+ /* Give user a chance to reallocate the stack. Use copies of
+ these so that the &'s don't force the real ones into
+ memory. */
+ yy_state_t *yyss1 = yyss;
+ YYSTYPE *yyvs1 = yyvs;
+
+ /* Each stack pointer address is followed by the size of the
+ data in use in that stack, in bytes. This used to be a
+ conditional around just the two extra args, but that might
+ be undefined if yyoverflow is a macro. */
+ yyoverflow (YY_("memory exhausted"),
+ &yyss1, yysize * YYSIZEOF (*yyssp),
+ &yyvs1, yysize * YYSIZEOF (*yyvsp),
+ &yystacksize);
+ yyss = yyss1;
+ yyvs = yyvs1;
+ }
+# else /* defined YYSTACK_RELOCATE */
+ /* Extend the stack our own way. */
+ if (YYMAXDEPTH <= yystacksize)
+ YYNOMEM;
+ yystacksize *= 2;
+ if (YYMAXDEPTH < yystacksize)
+ yystacksize = YYMAXDEPTH;
+
+ {
+ yy_state_t *yyss1 = yyss;
+ union yyalloc *yyptr =
+ YY_CAST (union yyalloc *,
+ YYSTACK_ALLOC (YY_CAST (YYSIZE_T, YYSTACK_BYTES (yystacksize))));
+ if (! yyptr)
+ YYNOMEM;
+ YYSTACK_RELOCATE (yyss_alloc, yyss);
+ YYSTACK_RELOCATE (yyvs_alloc, yyvs);
+# undef YYSTACK_RELOCATE
+ if (yyss1 != yyssa)
+ YYSTACK_FREE (yyss1);
+ }
+# endif
+
+ yyssp = yyss + yysize - 1;
+ yyvsp = yyvs + yysize - 1;
+
+ YY_IGNORE_USELESS_CAST_BEGIN
+ YYDPRINTF ((stderr, "Stack size increased to %ld\n",
+ YY_CAST (long, yystacksize)));
+ YY_IGNORE_USELESS_CAST_END
+
+ if (yyss + yystacksize - 1 <= yyssp)
+ YYABORT;
+ }
+#endif /* !defined yyoverflow && !defined YYSTACK_RELOCATE */
+
+
+ if (yystate == YYFINAL)
+ YYACCEPT;
+
+ goto yybackup;
+
+
+/*-----------.
+| yybackup. |
+`-----------*/
+yybackup:
+ /* Do appropriate processing given the current state. Read a
+ lookahead token if we need one and don't already have one. */
+
+ /* First try to decide what to do without reference to lookahead token. */
+ yyn = yypact[yystate];
+ if (yypact_value_is_default (yyn))
+ goto yydefault;
+
+ /* Not known => get a lookahead token if don't already have one. */
+
+ /* YYCHAR is either empty, or end-of-input, or a valid lookahead. */
+ if (yychar == YYEMPTY)
+ {
+ YYDPRINTF ((stderr, "Reading a token\n"));
+ yychar = yylex (&yylval, pc);
+ }
+
+ if (yychar <= YYEOF)
+ {
+ yychar = YYEOF;
+ yytoken = YYSYMBOL_YYEOF;
+ YYDPRINTF ((stderr, "Now at end of input.\n"));
+ }
+ else if (yychar == YYerror)
+ {
+ /* The scanner already issued an error message, process directly
+ to error recovery. But do not keep the error token as
+ lookahead, it is too special and may lead us to an endless
+ loop in error recovery. */
+ yychar = YYUNDEF;
+ yytoken = YYSYMBOL_YYerror;
+ goto yyerrlab1;
+ }
+ else
+ {
+ yytoken = YYTRANSLATE (yychar);
+ YY_SYMBOL_PRINT ("Next token is", yytoken, &yylval, &yylloc);
+ }
+
+ /* If the proper action on seeing token YYTOKEN is to reduce or to
+ detect an error, take that action. */
+ yyn += yytoken;
+ if (yyn < 0 || YYLAST < yyn || yycheck[yyn] != yytoken)
+ goto yydefault;
+ yyn = yytable[yyn];
+ if (yyn <= 0)
+ {
+ if (yytable_value_is_error (yyn))
+ goto yyerrlab;
+ yyn = -yyn;
+ goto yyreduce;
+ }
+
+ /* Count tokens shifted since error; after three, turn off error
+ status. */
+ if (yyerrstatus)
+ yyerrstatus--;
+
+ /* Shift the lookahead token. */
+ YY_SYMBOL_PRINT ("Shifting", yytoken, &yylval, &yylloc);
+ yystate = yyn;
+ YY_IGNORE_MAYBE_UNINITIALIZED_BEGIN
+ *++yyvsp = yylval;
+ YY_IGNORE_MAYBE_UNINITIALIZED_END
+
+ /* Discard the shifted token. */
+ yychar = YYEMPTY;
+ goto yynewstate;
+
+
+/*-----------------------------------------------------------.
+| yydefault -- do the default action for the current state. |
+`-----------------------------------------------------------*/
+yydefault:
+ yyn = yydefact[yystate];
+ if (yyn == 0)
+ goto yyerrlab;
+ goto yyreduce;
+
+
+/*-----------------------------.
+| yyreduce -- do a reduction. |
+`-----------------------------*/
+yyreduce:
+ /* yyn is the number of a rule to reduce with. */
+ yylen = yyr2[yyn];
+
+ /* If YYLEN is nonzero, implement the default value of the action:
+ '$$ = $1'.
+
+ Otherwise, the following line sets YYVAL to garbage.
+ This behavior is undocumented and Bison
+ users should not rely upon it. Assigning to YYVAL
+ unconditionally makes the parser a bit smaller, and it avoids a
+ GCC warning that YYVAL may be used uninitialized. */
+ yyval = yyvsp[1-yylen];
+
+
+ YY_REDUCE_PRINT (yyn);
+ switch (yyn)
+ {
+ case 4: /* timespec: '@' seconds */
+#line 609 "parse-datetime.y"
+ {
+ pc->seconds = (yyvsp[0].timespec);
+ pc->timespec_seen = true;
+ debug_print_current_time (_("number of seconds"), pc);
+ }
+#line 1771 "parse-datetime.c"
+ break;
+
+ case 7: /* item: datetime */
+#line 623 "parse-datetime.y"
+ {
+ pc->times_seen++; pc->dates_seen++;
+ debug_print_current_time (_("datetime"), pc);
+ }
+#line 1780 "parse-datetime.c"
+ break;
+
+ case 8: /* item: time */
+#line 628 "parse-datetime.y"
+ {
+ pc->times_seen++;
+ debug_print_current_time (_("time"), pc);
+ }
+#line 1789 "parse-datetime.c"
+ break;
+
+ case 9: /* item: local_zone */
+#line 633 "parse-datetime.y"
+ {
+ pc->local_zones_seen++;
+ debug_print_current_time (_("local_zone"), pc);
+ }
+#line 1798 "parse-datetime.c"
+ break;
+
+ case 10: /* item: zone */
+#line 638 "parse-datetime.y"
+ {
+ pc->zones_seen++;
+ debug_print_current_time (_("zone"), pc);
+ }
+#line 1807 "parse-datetime.c"
+ break;
+
+ case 11: /* item: date */
+#line 643 "parse-datetime.y"
+ {
+ pc->dates_seen++;
+ debug_print_current_time (_("date"), pc);
+ }
+#line 1816 "parse-datetime.c"
+ break;
+
+ case 12: /* item: day */
+#line 648 "parse-datetime.y"
+ {
+ pc->days_seen++;
+ debug_print_current_time (_("day"), pc);
+ }
+#line 1825 "parse-datetime.c"
+ break;
+
+ case 13: /* item: rel */
+#line 653 "parse-datetime.y"
+ {
+ debug_print_relative_time (_("relative"), pc);
+ }
+#line 1833 "parse-datetime.c"
+ break;
+
+ case 14: /* item: number */
+#line 657 "parse-datetime.y"
+ {
+ debug_print_current_time (_("number"), pc);
+ }
+#line 1841 "parse-datetime.c"
+ break;
+
+ case 15: /* item: hybrid */
+#line 661 "parse-datetime.y"
+ {
+ debug_print_relative_time (_("hybrid"), pc);
+ }
+#line 1849 "parse-datetime.c"
+ break;
+
+ case 18: /* time: tUNUMBER tMERIDIAN */
+#line 676 "parse-datetime.y"
+ {
+ set_hhmmss (pc, (yyvsp[-1].textintval).value, 0, 0, 0);
+ pc->meridian = (yyvsp[0].intval);
+ }
+#line 1858 "parse-datetime.c"
+ break;
+
+ case 19: /* time: tUNUMBER ':' tUNUMBER tMERIDIAN */
+#line 681 "parse-datetime.y"
+ {
+ set_hhmmss (pc, (yyvsp[-3].textintval).value, (yyvsp[-1].textintval).value, 0, 0);
+ pc->meridian = (yyvsp[0].intval);
+ }
+#line 1867 "parse-datetime.c"
+ break;
+
+ case 20: /* time: tUNUMBER ':' tUNUMBER ':' unsigned_seconds tMERIDIAN */
+#line 686 "parse-datetime.y"
+ {
+ set_hhmmss (pc, (yyvsp[-5].textintval).value, (yyvsp[-3].textintval).value, (yyvsp[-1].timespec).tv_sec, (yyvsp[-1].timespec).tv_nsec);
+ pc->meridian = (yyvsp[0].intval);
+ }
+#line 1876 "parse-datetime.c"
+ break;
+
+ case 22: /* iso_8601_time: tUNUMBER zone_offset */
+#line 695 "parse-datetime.y"
+ {
+ set_hhmmss (pc, (yyvsp[-1].textintval).value, 0, 0, 0);
+ pc->meridian = MER24;
+ }
+#line 1885 "parse-datetime.c"
+ break;
+
+ case 23: /* iso_8601_time: tUNUMBER ':' tUNUMBER o_zone_offset */
+#line 700 "parse-datetime.y"
+ {
+ set_hhmmss (pc, (yyvsp[-3].textintval).value, (yyvsp[-1].textintval).value, 0, 0);
+ pc->meridian = MER24;
+ }
+#line 1894 "parse-datetime.c"
+ break;
+
+ case 24: /* iso_8601_time: tUNUMBER ':' tUNUMBER ':' unsigned_seconds o_zone_offset */
+#line 705 "parse-datetime.y"
+ {
+ set_hhmmss (pc, (yyvsp[-5].textintval).value, (yyvsp[-3].textintval).value, (yyvsp[-1].timespec).tv_sec, (yyvsp[-1].timespec).tv_nsec);
+ pc->meridian = MER24;
+ }
+#line 1903 "parse-datetime.c"
+ break;
+
+ case 27: /* zone_offset: tSNUMBER o_colon_minutes */
+#line 718 "parse-datetime.y"
+ {
+ pc->zones_seen++;
+ if (! time_zone_hhmm (pc, (yyvsp[-1].textintval), (yyvsp[0].intval))) YYABORT;
+ }
+#line 1912 "parse-datetime.c"
+ break;
+
+ case 28: /* local_zone: tLOCAL_ZONE */
+#line 743 "parse-datetime.y"
+ { pc->local_isdst = (yyvsp[0].intval); }
+#line 1918 "parse-datetime.c"
+ break;
+
+ case 29: /* local_zone: tLOCAL_ZONE tDST */
+#line 745 "parse-datetime.y"
+ {
+ pc->local_isdst = 1;
+ pc->dsts_seen++;
+ }
+#line 1927 "parse-datetime.c"
+ break;
+
+ case 30: /* zone: tZONE */
+#line 755 "parse-datetime.y"
+ { pc->time_zone = (yyvsp[0].intval); }
+#line 1933 "parse-datetime.c"
+ break;
+
+ case 31: /* zone: 'T' */
+#line 757 "parse-datetime.y"
+ { pc->time_zone = -HOUR (7); }
+#line 1939 "parse-datetime.c"
+ break;
+
+ case 32: /* zone: tZONE relunit_snumber */
+#line 759 "parse-datetime.y"
+ { pc->time_zone = (yyvsp[-1].intval);
+ if (! apply_relative_time (pc, (yyvsp[0].rel), 1)) YYABORT;
+ debug_print_relative_time (_("relative"), pc);
+ }
+#line 1948 "parse-datetime.c"
+ break;
+
+ case 33: /* zone: 'T' relunit_snumber */
+#line 764 "parse-datetime.y"
+ { pc->time_zone = -HOUR (7);
+ if (! apply_relative_time (pc, (yyvsp[0].rel), 1)) YYABORT;
+ debug_print_relative_time (_("relative"), pc);
+ }
+#line 1957 "parse-datetime.c"
+ break;
+
+ case 34: /* zone: tZONE tSNUMBER o_colon_minutes */
+#line 769 "parse-datetime.y"
+ { if (! time_zone_hhmm (pc, (yyvsp[-1].textintval), (yyvsp[0].intval))) YYABORT;
+ if (INT_ADD_WRAPV (pc->time_zone, (yyvsp[-2].intval), &pc->time_zone)) YYABORT; }
+#line 1964 "parse-datetime.c"
+ break;
+
+ case 35: /* zone: tDAYZONE */
+#line 772 "parse-datetime.y"
+ { pc->time_zone = (yyvsp[0].intval) + 60 * 60; }
+#line 1970 "parse-datetime.c"
+ break;
+
+ case 36: /* zone: tZONE tDST */
+#line 774 "parse-datetime.y"
+ { pc->time_zone = (yyvsp[-1].intval) + 60 * 60; }
+#line 1976 "parse-datetime.c"
+ break;
+
+ case 37: /* day: tDAY */
+#line 779 "parse-datetime.y"
+ {
+ pc->day_ordinal = 0;
+ pc->day_number = (yyvsp[0].intval);
+ }
+#line 1985 "parse-datetime.c"
+ break;
+
+ case 38: /* day: tDAY ',' */
+#line 784 "parse-datetime.y"
+ {
+ pc->day_ordinal = 0;
+ pc->day_number = (yyvsp[-1].intval);
+ }
+#line 1994 "parse-datetime.c"
+ break;
+
+ case 39: /* day: tORDINAL tDAY */
+#line 789 "parse-datetime.y"
+ {
+ pc->day_ordinal = (yyvsp[-1].intval);
+ pc->day_number = (yyvsp[0].intval);
+ pc->debug_ordinal_day_seen = true;
+ }
+#line 2004 "parse-datetime.c"
+ break;
+
+ case 40: /* day: tUNUMBER tDAY */
+#line 795 "parse-datetime.y"
+ {
+ pc->day_ordinal = (yyvsp[-1].textintval).value;
+ pc->day_number = (yyvsp[0].intval);
+ pc->debug_ordinal_day_seen = true;
+ }
+#line 2014 "parse-datetime.c"
+ break;
+
+ case 41: /* date: tUNUMBER '/' tUNUMBER */
+#line 804 "parse-datetime.y"
+ {
+ pc->month = (yyvsp[-2].textintval).value;
+ pc->day = (yyvsp[0].textintval).value;
+ }
+#line 2023 "parse-datetime.c"
+ break;
+
+ case 42: /* date: tUNUMBER '/' tUNUMBER '/' tUNUMBER */
+#line 809 "parse-datetime.y"
+ {
+ /* Interpret as YYYY/MM/DD if the first value has 4 or more digits,
+ otherwise as MM/DD/YY.
+ The goal in recognizing YYYY/MM/DD is solely to support legacy
+ machine-generated dates like those in an RCS log listing. If
+ you want portability, use the ISO 8601 format. */
+ if (4 <= (yyvsp[-4].textintval).digits)
+ {
+ if (debugging (pc))
+ {
+ intmax_t digits = (yyvsp[-4].textintval).digits;
+ dbg_printf (_("warning: value %"PRIdMAX" has %"PRIdMAX" digits. "
+ "Assuming YYYY/MM/DD\n"),
+ (yyvsp[-4].textintval).value, digits);
+ }
+
+ pc->year = (yyvsp[-4].textintval);
+ pc->month = (yyvsp[-2].textintval).value;
+ pc->day = (yyvsp[0].textintval).value;
+ }
+ else
+ {
+ if (debugging (pc))
+ dbg_printf (_("warning: value %"PRIdMAX" has less than 4 digits. "
+ "Assuming MM/DD/YY[YY]\n"),
+ (yyvsp[-4].textintval).value);
+
+ pc->month = (yyvsp[-4].textintval).value;
+ pc->day = (yyvsp[-2].textintval).value;
+ pc->year = (yyvsp[0].textintval);
+ }
+ }
+#line 2060 "parse-datetime.c"
+ break;
+
+ case 43: /* date: tUNUMBER tMONTH tSNUMBER */
+#line 842 "parse-datetime.y"
+ {
+ /* E.g., 17-JUN-1992. */
+ pc->day = (yyvsp[-2].textintval).value;
+ pc->month = (yyvsp[-1].intval);
+ if (INT_SUBTRACT_WRAPV (0, (yyvsp[0].textintval).value, &pc->year.value)) YYABORT;
+ pc->year.digits = (yyvsp[0].textintval).digits;
+ }
+#line 2072 "parse-datetime.c"
+ break;
+
+ case 44: /* date: tMONTH tSNUMBER tSNUMBER */
+#line 850 "parse-datetime.y"
+ {
+ /* E.g., JUN-17-1992. */
+ pc->month = (yyvsp[-2].intval);
+ if (INT_SUBTRACT_WRAPV (0, (yyvsp[-1].textintval).value, &pc->day)) YYABORT;
+ if (INT_SUBTRACT_WRAPV (0, (yyvsp[0].textintval).value, &pc->year.value)) YYABORT;
+ pc->year.digits = (yyvsp[0].textintval).digits;
+ }
+#line 2084 "parse-datetime.c"
+ break;
+
+ case 45: /* date: tMONTH tUNUMBER */
+#line 858 "parse-datetime.y"
+ {
+ pc->month = (yyvsp[-1].intval);
+ pc->day = (yyvsp[0].textintval).value;
+ }
+#line 2093 "parse-datetime.c"
+ break;
+
+ case 46: /* date: tMONTH tUNUMBER ',' tUNUMBER */
+#line 863 "parse-datetime.y"
+ {
+ pc->month = (yyvsp[-3].intval);
+ pc->day = (yyvsp[-2].textintval).value;
+ pc->year = (yyvsp[0].textintval);
+ }
+#line 2103 "parse-datetime.c"
+ break;
+
+ case 47: /* date: tUNUMBER tMONTH */
+#line 869 "parse-datetime.y"
+ {
+ pc->day = (yyvsp[-1].textintval).value;
+ pc->month = (yyvsp[0].intval);
+ }
+#line 2112 "parse-datetime.c"
+ break;
+
+ case 48: /* date: tUNUMBER tMONTH tUNUMBER */
+#line 874 "parse-datetime.y"
+ {
+ pc->day = (yyvsp[-2].textintval).value;
+ pc->month = (yyvsp[-1].intval);
+ pc->year = (yyvsp[0].textintval);
+ }
+#line 2122 "parse-datetime.c"
+ break;
+
+ case 50: /* iso_8601_date: tUNUMBER tSNUMBER tSNUMBER */
+#line 884 "parse-datetime.y"
+ {
+ /* ISO 8601 format. YYYY-MM-DD. */
+ pc->year = (yyvsp[-2].textintval);
+ if (INT_SUBTRACT_WRAPV (0, (yyvsp[-1].textintval).value, &pc->month)) YYABORT;
+ if (INT_SUBTRACT_WRAPV (0, (yyvsp[0].textintval).value, &pc->day)) YYABORT;
+ }
+#line 2133 "parse-datetime.c"
+ break;
+
+ case 51: /* rel: relunit tAGO */
+#line 894 "parse-datetime.y"
+ { if (! apply_relative_time (pc, (yyvsp[-1].rel), (yyvsp[0].intval))) YYABORT; }
+#line 2139 "parse-datetime.c"
+ break;
+
+ case 52: /* rel: relunit */
+#line 896 "parse-datetime.y"
+ { if (! apply_relative_time (pc, (yyvsp[0].rel), 1)) YYABORT; }
+#line 2145 "parse-datetime.c"
+ break;
+
+ case 53: /* rel: dayshift */
+#line 898 "parse-datetime.y"
+ { if (! apply_relative_time (pc, (yyvsp[0].rel), 1)) YYABORT; }
+#line 2151 "parse-datetime.c"
+ break;
+
+ case 54: /* relunit: tORDINAL tYEAR_UNIT */
+#line 903 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).year = (yyvsp[-1].intval); }
+#line 2157 "parse-datetime.c"
+ break;
+
+ case 55: /* relunit: tUNUMBER tYEAR_UNIT */
+#line 905 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).year = (yyvsp[-1].textintval).value; }
+#line 2163 "parse-datetime.c"
+ break;
+
+ case 56: /* relunit: tYEAR_UNIT */
+#line 907 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).year = 1; }
+#line 2169 "parse-datetime.c"
+ break;
+
+ case 57: /* relunit: tORDINAL tMONTH_UNIT */
+#line 909 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).month = (yyvsp[-1].intval); }
+#line 2175 "parse-datetime.c"
+ break;
+
+ case 58: /* relunit: tUNUMBER tMONTH_UNIT */
+#line 911 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).month = (yyvsp[-1].textintval).value; }
+#line 2181 "parse-datetime.c"
+ break;
+
+ case 59: /* relunit: tMONTH_UNIT */
+#line 913 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).month = 1; }
+#line 2187 "parse-datetime.c"
+ break;
+
+ case 60: /* relunit: tORDINAL tDAY_UNIT */
+#line 915 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0;
+ if (INT_MULTIPLY_WRAPV ((yyvsp[-1].intval), (yyvsp[0].intval), &(yyval.rel).day)) YYABORT; }
+#line 2194 "parse-datetime.c"
+ break;
+
+ case 61: /* relunit: tUNUMBER tDAY_UNIT */
+#line 918 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0;
+ if (INT_MULTIPLY_WRAPV ((yyvsp[-1].textintval).value, (yyvsp[0].intval), &(yyval.rel).day)) YYABORT; }
+#line 2201 "parse-datetime.c"
+ break;
+
+ case 62: /* relunit: tDAY_UNIT */
+#line 921 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).day = (yyvsp[0].intval); }
+#line 2207 "parse-datetime.c"
+ break;
+
+ case 63: /* relunit: tORDINAL tHOUR_UNIT */
+#line 923 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).hour = (yyvsp[-1].intval); }
+#line 2213 "parse-datetime.c"
+ break;
+
+ case 64: /* relunit: tUNUMBER tHOUR_UNIT */
+#line 925 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).hour = (yyvsp[-1].textintval).value; }
+#line 2219 "parse-datetime.c"
+ break;
+
+ case 65: /* relunit: tHOUR_UNIT */
+#line 927 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).hour = 1; }
+#line 2225 "parse-datetime.c"
+ break;
+
+ case 66: /* relunit: tORDINAL tMINUTE_UNIT */
+#line 929 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).minutes = (yyvsp[-1].intval); }
+#line 2231 "parse-datetime.c"
+ break;
+
+ case 67: /* relunit: tUNUMBER tMINUTE_UNIT */
+#line 931 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).minutes = (yyvsp[-1].textintval).value; }
+#line 2237 "parse-datetime.c"
+ break;
+
+ case 68: /* relunit: tMINUTE_UNIT */
+#line 933 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).minutes = 1; }
+#line 2243 "parse-datetime.c"
+ break;
+
+ case 69: /* relunit: tORDINAL tSEC_UNIT */
+#line 935 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = (yyvsp[-1].intval); }
+#line 2249 "parse-datetime.c"
+ break;
+
+ case 70: /* relunit: tUNUMBER tSEC_UNIT */
+#line 937 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = (yyvsp[-1].textintval).value; }
+#line 2255 "parse-datetime.c"
+ break;
+
+ case 71: /* relunit: tSDECIMAL_NUMBER tSEC_UNIT */
+#line 939 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = (yyvsp[-1].timespec).tv_sec; (yyval.rel).ns = (yyvsp[-1].timespec).tv_nsec; }
+#line 2261 "parse-datetime.c"
+ break;
+
+ case 72: /* relunit: tUDECIMAL_NUMBER tSEC_UNIT */
+#line 941 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = (yyvsp[-1].timespec).tv_sec; (yyval.rel).ns = (yyvsp[-1].timespec).tv_nsec; }
+#line 2267 "parse-datetime.c"
+ break;
+
+ case 73: /* relunit: tSEC_UNIT */
+#line 943 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = 1; }
+#line 2273 "parse-datetime.c"
+ break;
+
+ case 75: /* relunit_snumber: tSNUMBER tYEAR_UNIT */
+#line 949 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).year = (yyvsp[-1].textintval).value; }
+#line 2279 "parse-datetime.c"
+ break;
+
+ case 76: /* relunit_snumber: tSNUMBER tMONTH_UNIT */
+#line 951 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).month = (yyvsp[-1].textintval).value; }
+#line 2285 "parse-datetime.c"
+ break;
+
+ case 77: /* relunit_snumber: tSNUMBER tDAY_UNIT */
+#line 953 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0;
+ if (INT_MULTIPLY_WRAPV ((yyvsp[-1].textintval).value, (yyvsp[0].intval), &(yyval.rel).day)) YYABORT; }
+#line 2292 "parse-datetime.c"
+ break;
+
+ case 78: /* relunit_snumber: tSNUMBER tHOUR_UNIT */
+#line 956 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).hour = (yyvsp[-1].textintval).value; }
+#line 2298 "parse-datetime.c"
+ break;
+
+ case 79: /* relunit_snumber: tSNUMBER tMINUTE_UNIT */
+#line 958 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).minutes = (yyvsp[-1].textintval).value; }
+#line 2304 "parse-datetime.c"
+ break;
+
+ case 80: /* relunit_snumber: tSNUMBER tSEC_UNIT */
+#line 960 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).seconds = (yyvsp[-1].textintval).value; }
+#line 2310 "parse-datetime.c"
+ break;
+
+ case 81: /* dayshift: tDAY_SHIFT */
+#line 965 "parse-datetime.y"
+ { (yyval.rel) = RELATIVE_TIME_0; (yyval.rel).day = (yyvsp[0].intval); }
+#line 2316 "parse-datetime.c"
+ break;
+
+ case 85: /* signed_seconds: tSNUMBER */
+#line 973 "parse-datetime.y"
+ { if (time_overflow ((yyvsp[0].textintval).value)) YYABORT;
+ (yyval.timespec).tv_sec = (yyvsp[0].textintval).value; (yyval.timespec).tv_nsec = 0; }
+#line 2323 "parse-datetime.c"
+ break;
+
+ case 87: /* unsigned_seconds: tUNUMBER */
+#line 980 "parse-datetime.y"
+ { if (time_overflow ((yyvsp[0].textintval).value)) YYABORT;
+ (yyval.timespec).tv_sec = (yyvsp[0].textintval).value; (yyval.timespec).tv_nsec = 0; }
+#line 2330 "parse-datetime.c"
+ break;
+
+ case 88: /* number: tUNUMBER */
+#line 986 "parse-datetime.y"
+ { digits_to_date_time (pc, (yyvsp[0].textintval)); }
+#line 2336 "parse-datetime.c"
+ break;
+
+ case 89: /* hybrid: tUNUMBER relunit_snumber */
+#line 991 "parse-datetime.y"
+ {
+ /* Hybrid all-digit and relative offset, so that we accept e.g.,
+ "YYYYMMDD +N days" as well as "YYYYMMDD N days". */
+ digits_to_date_time (pc, (yyvsp[-1].textintval));
+ if (! apply_relative_time (pc, (yyvsp[0].rel), 1)) YYABORT;
+ }
+#line 2347 "parse-datetime.c"
+ break;
+
+ case 90: /* o_colon_minutes: %empty */
+#line 1001 "parse-datetime.y"
+ { (yyval.intval) = -1; }
+#line 2353 "parse-datetime.c"
+ break;
+
+ case 91: /* o_colon_minutes: ':' tUNUMBER */
+#line 1003 "parse-datetime.y"
+ { (yyval.intval) = (yyvsp[0].textintval).value; }
+#line 2359 "parse-datetime.c"
+ break;
+
+
+#line 2363 "parse-datetime.c"
+
+ default: break;
+ }
+ /* User semantic actions sometimes alter yychar, and that requires
+ that yytoken be updated with the new translation. We take the
+ approach of translating immediately before every use of yytoken.
+ One alternative is translating here after every semantic action,
+ but that translation would be missed if the semantic action invokes
+ YYABORT, YYACCEPT, or YYERROR immediately after altering yychar or
+ if it invokes YYBACKUP. In the case of YYABORT or YYACCEPT, an
+ incorrect destructor might then be invoked immediately. In the
+ case of YYERROR or YYBACKUP, subsequent parser actions might lead
+ to an incorrect destructor call or verbose syntax error message
+ before the lookahead is translated. */
+ YY_SYMBOL_PRINT ("-> $$ =", YY_CAST (yysymbol_kind_t, yyr1[yyn]), &yyval, &yyloc);
+
+ YYPOPSTACK (yylen);
+ yylen = 0;
+
+ *++yyvsp = yyval;
+
+ /* Now 'shift' the result of the reduction. Determine what state
+ that goes to, based on the state we popped back to and the rule
+ number reduced by. */
+ {
+ const int yylhs = yyr1[yyn] - YYNTOKENS;
+ const int yyi = yypgoto[yylhs] + *yyssp;
+ yystate = (0 <= yyi && yyi <= YYLAST && yycheck[yyi] == *yyssp
+ ? yytable[yyi]
+ : yydefgoto[yylhs]);
+ }
+
+ goto yynewstate;
+
+
+/*--------------------------------------.
+| yyerrlab -- here on detecting error. |
+`--------------------------------------*/
+yyerrlab:
+ /* Make sure we have latest lookahead translation. See comments at
+ user semantic actions for why this is necessary. */
+ yytoken = yychar == YYEMPTY ? YYSYMBOL_YYEMPTY : YYTRANSLATE (yychar);
+ /* If not already recovering from an error, report this error. */
+ if (!yyerrstatus)
+ {
+ ++yynerrs;
+ yyerror (pc, YY_("syntax error"));
+ }
+
+ if (yyerrstatus == 3)
+ {
+ /* If just tried and failed to reuse lookahead token after an
+ error, discard it. */
+
+ if (yychar <= YYEOF)
+ {
+ /* Return failure if at end of input. */
+ if (yychar == YYEOF)
+ YYABORT;
+ }
+ else
+ {
+ yydestruct ("Error: discarding",
+ yytoken, &yylval, pc);
+ yychar = YYEMPTY;
+ }
+ }
+
+ /* Else will try to reuse lookahead token after shifting the error
+ token. */
+ goto yyerrlab1;
+
+
+/*---------------------------------------------------.
+| yyerrorlab -- error raised explicitly by YYERROR. |
+`---------------------------------------------------*/
+yyerrorlab:
+ /* Pacify compilers when the user code never invokes YYERROR and the
+ label yyerrorlab therefore never appears in user code. */
+ if (0)
+ YYERROR;
+ ++yynerrs;
+
+ /* Do not reclaim the symbols of the rule whose action triggered
+ this YYERROR. */
+ YYPOPSTACK (yylen);
+ yylen = 0;
+ YY_STACK_PRINT (yyss, yyssp);
+ yystate = *yyssp;
+ goto yyerrlab1;
+
+
+/*-------------------------------------------------------------.
+| yyerrlab1 -- common code for both syntax error and YYERROR. |
+`-------------------------------------------------------------*/
+yyerrlab1:
+ yyerrstatus = 3; /* Each real token shifted decrements this. */
+
+ /* Pop stack until we find a state that shifts the error token. */
+ for (;;)
+ {
+ yyn = yypact[yystate];
+ if (!yypact_value_is_default (yyn))
+ {
+ yyn += YYSYMBOL_YYerror;
+ if (0 <= yyn && yyn <= YYLAST && yycheck[yyn] == YYSYMBOL_YYerror)
+ {
+ yyn = yytable[yyn];
+ if (0 < yyn)
+ break;
+ }
+ }
+
+ /* Pop the current state because it cannot handle the error token. */
+ if (yyssp == yyss)
+ YYABORT;
+
+
+ yydestruct ("Error: popping",
+ YY_ACCESSING_SYMBOL (yystate), yyvsp, pc);
+ YYPOPSTACK (1);
+ yystate = *yyssp;
+ YY_STACK_PRINT (yyss, yyssp);
+ }
+
+ YY_IGNORE_MAYBE_UNINITIALIZED_BEGIN
+ *++yyvsp = yylval;
+ YY_IGNORE_MAYBE_UNINITIALIZED_END
+
+
+ /* Shift the error token. */
+ YY_SYMBOL_PRINT ("Shifting", YY_ACCESSING_SYMBOL (yyn), yyvsp, yylsp);
+
+ yystate = yyn;
+ goto yynewstate;
+
+
+/*-------------------------------------.
+| yyacceptlab -- YYACCEPT comes here. |
+`-------------------------------------*/
+yyacceptlab:
+ yyresult = 0;
+ goto yyreturnlab;
+
+
+/*-----------------------------------.
+| yyabortlab -- YYABORT comes here. |
+`-----------------------------------*/
+yyabortlab:
+ yyresult = 1;
+ goto yyreturnlab;
+
+
+/*-----------------------------------------------------------.
+| yyexhaustedlab -- YYNOMEM (memory exhaustion) comes here. |
+`-----------------------------------------------------------*/
+yyexhaustedlab:
+ yyerror (pc, YY_("memory exhausted"));
+ yyresult = 2;
+ goto yyreturnlab;
+
+
+/*----------------------------------------------------------.
+| yyreturnlab -- parsing is finished, clean up and return. |
+`----------------------------------------------------------*/
+yyreturnlab:
+ if (yychar != YYEMPTY)
+ {
+ /* Make sure we have latest lookahead translation. See comments at
+ user semantic actions for why this is necessary. */
+ yytoken = YYTRANSLATE (yychar);
+ yydestruct ("Cleanup: discarding lookahead",
+ yytoken, &yylval, pc);
+ }
+ /* Do not reclaim the symbols of the rule whose action triggered
+ this YYABORT or YYACCEPT. */
+ YYPOPSTACK (yylen);
+ YY_STACK_PRINT (yyss, yyssp);
+ while (yyssp != yyss)
+ {
+ yydestruct ("Cleanup: popping",
+ YY_ACCESSING_SYMBOL (+*yyssp), yyvsp, pc);
+ YYPOPSTACK (1);
+ }
+#ifndef yyoverflow
+ if (yyss != yyssa)
+ YYSTACK_FREE (yyss);
+#endif
+
+ return yyresult;
+}
+
+#line 1006 "parse-datetime.y"
+
+
+static table const meridian_table[] =
+{
+ { "AM", tMERIDIAN, MERam },
+ { "A.M.", tMERIDIAN, MERam },
+ { "PM", tMERIDIAN, MERpm },
+ { "P.M.", tMERIDIAN, MERpm },
+ { NULL, 0, 0 }
+};
+
+static table const dst_table[] =
+{
+ { "DST", tDST, 0 }
+};
+
+static table const month_and_day_table[] =
+{
+ { "JANUARY", tMONTH, 1 },
+ { "FEBRUARY", tMONTH, 2 },
+ { "MARCH", tMONTH, 3 },
+ { "APRIL", tMONTH, 4 },
+ { "MAY", tMONTH, 5 },
+ { "JUNE", tMONTH, 6 },
+ { "JULY", tMONTH, 7 },
+ { "AUGUST", tMONTH, 8 },
+ { "SEPTEMBER",tMONTH, 9 },
+ { "SEPT", tMONTH, 9 },
+ { "OCTOBER", tMONTH, 10 },
+ { "NOVEMBER", tMONTH, 11 },
+ { "DECEMBER", tMONTH, 12 },
+ { "SUNDAY", tDAY, 0 },
+ { "MONDAY", tDAY, 1 },
+ { "TUESDAY", tDAY, 2 },
+ { "TUES", tDAY, 2 },
+ { "WEDNESDAY",tDAY, 3 },
+ { "WEDNES", tDAY, 3 },
+ { "THURSDAY", tDAY, 4 },
+ { "THUR", tDAY, 4 },
+ { "THURS", tDAY, 4 },
+ { "FRIDAY", tDAY, 5 },
+ { "SATURDAY", tDAY, 6 },
+ { NULL, 0, 0 }
+};
+
+static table const time_units_table[] =
+{
+ { "YEAR", tYEAR_UNIT, 1 },
+ { "MONTH", tMONTH_UNIT, 1 },
+ { "FORTNIGHT",tDAY_UNIT, 14 },
+ { "WEEK", tDAY_UNIT, 7 },
+ { "DAY", tDAY_UNIT, 1 },
+ { "HOUR", tHOUR_UNIT, 1 },
+ { "MINUTE", tMINUTE_UNIT, 1 },
+ { "MIN", tMINUTE_UNIT, 1 },
+ { "SECOND", tSEC_UNIT, 1 },
+ { "SEC", tSEC_UNIT, 1 },
+ { NULL, 0, 0 }
+};
+
+/* Assorted relative-time words. */
+static table const relative_time_table[] =
+{
+ { "TOMORROW", tDAY_SHIFT, 1 },
+ { "YESTERDAY",tDAY_SHIFT, -1 },
+ { "TODAY", tDAY_SHIFT, 0 },
+ { "NOW", tDAY_SHIFT, 0 },
+ { "LAST", tORDINAL, -1 },
+ { "THIS", tORDINAL, 0 },
+ { "NEXT", tORDINAL, 1 },
+ { "FIRST", tORDINAL, 1 },
+/*{ "SECOND", tORDINAL, 2 }, */
+ { "THIRD", tORDINAL, 3 },
+ { "FOURTH", tORDINAL, 4 },
+ { "FIFTH", tORDINAL, 5 },
+ { "SIXTH", tORDINAL, 6 },
+ { "SEVENTH", tORDINAL, 7 },
+ { "EIGHTH", tORDINAL, 8 },
+ { "NINTH", tORDINAL, 9 },
+ { "TENTH", tORDINAL, 10 },
+ { "ELEVENTH", tORDINAL, 11 },
+ { "TWELFTH", tORDINAL, 12 },
+ { "AGO", tAGO, -1 },
+ { "HENCE", tAGO, 1 },
+ { NULL, 0, 0 }
+};
+
+/* The universal time zone table. These labels can be used even for
+ timestamps that would not otherwise be valid, e.g., GMT timestamps
+ oin London during summer. */
+static table const universal_time_zone_table[] =
+{
+ { "GMT", tZONE, HOUR ( 0) }, /* Greenwich Mean */
+ { "UT", tZONE, HOUR ( 0) }, /* Universal (Coordinated) */
+ { "UTC", tZONE, HOUR ( 0) },
+ { NULL, 0, 0 }
+};
+
+/* The time zone table. This table is necessarily incomplete, as time
+ zone abbreviations are ambiguous; e.g., Australians interpret "EST"
+ as Eastern time in Australia, not as US Eastern Standard Time.
+ You cannot rely on parse_datetime to handle arbitrary time zone
+ abbreviations; use numeric abbreviations like "-0500" instead. */
+static table const time_zone_table[] =
+{
+ { "WET", tZONE, HOUR ( 0) }, /* Western European */
+ { "WEST", tDAYZONE, HOUR ( 0) }, /* Western European Summer */
+ { "BST", tDAYZONE, HOUR ( 0) }, /* British Summer */
+ { "ART", tZONE, -HOUR ( 3) }, /* Argentina */
+ { "BRT", tZONE, -HOUR ( 3) }, /* Brazil */
+ { "BRST", tDAYZONE, -HOUR ( 3) }, /* Brazil Summer */
+ { "NST", tZONE, -(HOUR ( 3) + 30 * 60) }, /* Newfoundland Standard */
+ { "NDT", tDAYZONE,-(HOUR ( 3) + 30 * 60) }, /* Newfoundland Daylight */
+ { "AST", tZONE, -HOUR ( 4) }, /* Atlantic Standard */
+ { "ADT", tDAYZONE, -HOUR ( 4) }, /* Atlantic Daylight */
+ { "CLT", tZONE, -HOUR ( 4) }, /* Chile */
+ { "CLST", tDAYZONE, -HOUR ( 4) }, /* Chile Summer */
+ { "EST", tZONE, -HOUR ( 5) }, /* Eastern Standard */
+ { "EDT", tDAYZONE, -HOUR ( 5) }, /* Eastern Daylight */
+ { "CST", tZONE, -HOUR ( 6) }, /* Central Standard */
+ { "CDT", tDAYZONE, -HOUR ( 6) }, /* Central Daylight */
+ { "MST", tZONE, -HOUR ( 7) }, /* Mountain Standard */
+ { "MDT", tDAYZONE, -HOUR ( 7) }, /* Mountain Daylight */
+ { "PST", tZONE, -HOUR ( 8) }, /* Pacific Standard */
+ { "PDT", tDAYZONE, -HOUR ( 8) }, /* Pacific Daylight */
+ { "AKST", tZONE, -HOUR ( 9) }, /* Alaska Standard */
+ { "AKDT", tDAYZONE, -HOUR ( 9) }, /* Alaska Daylight */
+ { "HST", tZONE, -HOUR (10) }, /* Hawaii Standard */
+ { "HAST", tZONE, -HOUR (10) }, /* Hawaii-Aleutian Standard */
+ { "HADT", tDAYZONE, -HOUR (10) }, /* Hawaii-Aleutian Daylight */
+ { "SST", tZONE, -HOUR (12) }, /* Samoa Standard */
+ { "WAT", tZONE, HOUR ( 1) }, /* West Africa */
+ { "CET", tZONE, HOUR ( 1) }, /* Central European */
+ { "CEST", tDAYZONE, HOUR ( 1) }, /* Central European Summer */
+ { "MET", tZONE, HOUR ( 1) }, /* Middle European */
+ { "MEZ", tZONE, HOUR ( 1) }, /* Middle European */
+ { "MEST", tDAYZONE, HOUR ( 1) }, /* Middle European Summer */
+ { "MESZ", tDAYZONE, HOUR ( 1) }, /* Middle European Summer */
+ { "EET", tZONE, HOUR ( 2) }, /* Eastern European */
+ { "EEST", tDAYZONE, HOUR ( 2) }, /* Eastern European Summer */
+ { "CAT", tZONE, HOUR ( 2) }, /* Central Africa */
+ { "SAST", tZONE, HOUR ( 2) }, /* South Africa Standard */
+ { "EAT", tZONE, HOUR ( 3) }, /* East Africa */
+ { "MSK", tZONE, HOUR ( 3) }, /* Moscow */
+ { "MSD", tDAYZONE, HOUR ( 3) }, /* Moscow Daylight */
+ { "IST", tZONE, (HOUR ( 5) + 30 * 60) }, /* India Standard */
+ { "SGT", tZONE, HOUR ( 8) }, /* Singapore */
+ { "KST", tZONE, HOUR ( 9) }, /* Korea Standard */
+ { "JST", tZONE, HOUR ( 9) }, /* Japan Standard */
+ { "GST", tZONE, HOUR (10) }, /* Guam Standard */
+ { "NZST", tZONE, HOUR (12) }, /* New Zealand Standard */
+ { "NZDT", tDAYZONE, HOUR (12) }, /* New Zealand Daylight */
+ { NULL, 0, 0 }
+};
+
+/* Military time zone table.
+
+ RFC 822 got these backwards, but RFC 5322 makes the incorrect
+ treatment optional, so do them the right way here.
+
+ Note 'T' is a special case, as it is used as the separator in ISO
+ 8601 date and time of day representation. */
+static table const military_table[] =
+{
+ { "A", tZONE, HOUR ( 1) },
+ { "B", tZONE, HOUR ( 2) },
+ { "C", tZONE, HOUR ( 3) },
+ { "D", tZONE, HOUR ( 4) },
+ { "E", tZONE, HOUR ( 5) },
+ { "F", tZONE, HOUR ( 6) },
+ { "G", tZONE, HOUR ( 7) },
+ { "H", tZONE, HOUR ( 8) },
+ { "I", tZONE, HOUR ( 9) },
+ { "K", tZONE, HOUR (10) },
+ { "L", tZONE, HOUR (11) },
+ { "M", tZONE, HOUR (12) },
+ { "N", tZONE, -HOUR ( 1) },
+ { "O", tZONE, -HOUR ( 2) },
+ { "P", tZONE, -HOUR ( 3) },
+ { "Q", tZONE, -HOUR ( 4) },
+ { "R", tZONE, -HOUR ( 5) },
+ { "S", tZONE, -HOUR ( 6) },
+ { "T", 'T', 0 },
+ { "U", tZONE, -HOUR ( 8) },
+ { "V", tZONE, -HOUR ( 9) },
+ { "W", tZONE, -HOUR (10) },
+ { "X", tZONE, -HOUR (11) },
+ { "Y", tZONE, -HOUR (12) },
+ { "Z", tZONE, HOUR ( 0) },
+ { NULL, 0, 0 }
+};
+
+
+
+/* Convert a time zone expressed as HH:MM into an integer count of
+ seconds. If MM is negative, then S is of the form HHMM and needs
+ to be picked apart; otherwise, S is of the form HH. As specified in
+ https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap08.html#tag_08_03, allow
+ only valid TZ range, and consider first two digits as hours, if no
+ minutes specified. Return true if successful. */
+
+static bool
+time_zone_hhmm (parser_control *pc, textint s, intmax_t mm)
+{
+ intmax_t n_minutes;
+ bool overflow = false;
+
+ /* If the length of S is 1 or 2 and no minutes are specified,
+ interpret it as a number of hours. */
+ if (s.digits <= 2 && mm < 0)
+ s.value *= 100;
+
+ if (mm < 0)
+ n_minutes = (s.value / 100) * 60 + s.value % 100;
+ else
+ {
+ overflow |= INT_MULTIPLY_WRAPV (s.value, 60, &n_minutes);
+ overflow |= (s.negative
+ ? INT_SUBTRACT_WRAPV (n_minutes, mm, &n_minutes)
+ : INT_ADD_WRAPV (n_minutes, mm, &n_minutes));
+ }
+
+ if (overflow || ! (-24 * 60 <= n_minutes && n_minutes <= 24 * 60))
+ return false;
+ pc->time_zone = n_minutes * 60;
+ return true;
+}
+
+static int
+to_hour (intmax_t hours, int meridian)
+{
+ switch (meridian)
+ {
+ default: /* Pacify GCC. */
+ case MER24:
+ return 0 <= hours && hours < 24 ? hours : -1;
+ case MERam:
+ return 0 < hours && hours < 12 ? hours : hours == 12 ? 0 : -1;
+ case MERpm:
+ return 0 < hours && hours < 12 ? hours + 12 : hours == 12 ? 12 : -1;
+ }
+}
+
+enum { TM_YEAR_BASE = 1900 };
+enum { TM_YEAR_BUFSIZE = INT_BUFSIZE_BOUND (int) + 1 };
+
+/* Convert TM_YEAR, a year minus 1900, to a string that is numerically
+ correct even if subtracting 1900 would overflow. */
+
+static char const *
+tm_year_str (int tm_year, char buf[TM_YEAR_BUFSIZE])
+{
+ verify (TM_YEAR_BASE % 100 == 0);
+ sprintf (buf, &"-%02d%02d"[-TM_YEAR_BASE <= tm_year],
+ abs (tm_year / 100 + TM_YEAR_BASE / 100),
+ abs (tm_year % 100));
+ return buf;
+}
+
+/* Convert a text year number to a year minus 1900, working correctly
+ even if the input is in the range INT_MAX .. INT_MAX + 1900 - 1. */
+
+static bool
+to_tm_year (textint textyear, bool debug, int *tm_year)
+{
+ intmax_t year = textyear.value;
+
+ /* XPG4 suggests that years 00-68 map to 2000-2068, and
+ years 69-99 map to 1969-1999. */
+ if (0 <= year && textyear.digits == 2)
+ {
+ year += year < 69 ? 2000 : 1900;
+ if (debug)
+ dbg_printf (_("warning: adjusting year value %"PRIdMAX
+ " to %"PRIdMAX"\n"),
+ textyear.value, year);
+ }
+
+ if (year < 0
+ ? INT_SUBTRACT_WRAPV (-TM_YEAR_BASE, year, tm_year)
+ : INT_SUBTRACT_WRAPV (year, TM_YEAR_BASE, tm_year))
+ {
+ if (debug)
+ dbg_printf (_("error: out-of-range year %"PRIdMAX"\n"), year);
+ return false;
+ }
+
+ return true;
+}
+
+static table const * _GL_ATTRIBUTE_PURE
+lookup_zone (parser_control const *pc, char const *name)
+{
+ table const *tp;
+
+ for (tp = universal_time_zone_table; tp->name; tp++)
+ if (strcmp (name, tp->name) == 0)
+ return tp;
+
+ /* Try local zone abbreviations before those in time_zone_table, as
+ the local ones are more likely to be right. */
+ for (tp = pc->local_time_zone_table; tp->name; tp++)
+ if (strcmp (name, tp->name) == 0)
+ return tp;
+
+ for (tp = time_zone_table; tp->name; tp++)
+ if (strcmp (name, tp->name) == 0)
+ return tp;
+
+ return NULL;
+}
+
+#if ! HAVE_TM_GMTOFF
+/* Yield the difference between *A and *B,
+ measured in seconds, ignoring leap seconds.
+ The body of this function is taken directly from the GNU C Library;
+ see strftime.c. */
+static int
+tm_diff (const struct tm *a, const struct tm *b)
+{
+ /* Compute intervening leap days correctly even if year is negative.
+ Take care to avoid int overflow in leap day calculations,
+ but it's OK to assume that A and B are close to each other. */
+ int a4 = SHR (a->tm_year, 2) + SHR (TM_YEAR_BASE, 2) - ! (a->tm_year & 3);
+ int b4 = SHR (b->tm_year, 2) + SHR (TM_YEAR_BASE, 2) - ! (b->tm_year & 3);
+ int a100 = a4 / 25 - (a4 % 25 < 0);
+ int b100 = b4 / 25 - (b4 % 25 < 0);
+ int a400 = SHR (a100, 2);
+ int b400 = SHR (b100, 2);
+ int intervening_leap_days = (a4 - b4) - (a100 - b100) + (a400 - b400);
+ int years = a->tm_year - b->tm_year;
+ int days = (365 * years + intervening_leap_days
+ + (a->tm_yday - b->tm_yday));
+ return (60 * (60 * (24 * days + (a->tm_hour - b->tm_hour))
+ + (a->tm_min - b->tm_min))
+ + (a->tm_sec - b->tm_sec));
+}
+#endif /* ! HAVE_TM_GMTOFF */
+
+static table const *
+lookup_word (parser_control const *pc, char *word)
+{
+ char *p;
+ char *q;
+ idx_t wordlen;
+ table const *tp;
+ bool period_found;
+ bool abbrev;
+
+ /* Make it uppercase. */
+ for (p = word; *p; p++)
+ *p = c_toupper (to_uchar (*p));
+
+ for (tp = meridian_table; tp->name; tp++)
+ if (strcmp (word, tp->name) == 0)
+ return tp;
+
+ /* See if we have an abbreviation for a month. */
+ wordlen = strlen (word);
+ abbrev = wordlen == 3 || (wordlen == 4 && word[3] == '.');
+
+ for (tp = month_and_day_table; tp->name; tp++)
+ if ((abbrev ? strncmp (word, tp->name, 3) : strcmp (word, tp->name)) == 0)
+ return tp;
+
+ if ((tp = lookup_zone (pc, word)))
+ return tp;
+
+ if (strcmp (word, dst_table[0].name) == 0)
+ return dst_table;
+
+ for (tp = time_units_table; tp->name; tp++)
+ if (strcmp (word, tp->name) == 0)
+ return tp;
+
+ /* Strip off any plural and try the units table again. */
+ if (word[wordlen - 1] == 'S')
+ {
+ word[wordlen - 1] = '\0';
+ for (tp = time_units_table; tp->name; tp++)
+ if (strcmp (word, tp->name) == 0)
+ return tp;
+ word[wordlen - 1] = 'S'; /* For "this" in relative_time_table. */
+ }
+
+ for (tp = relative_time_table; tp->name; tp++)
+ if (strcmp (word, tp->name) == 0)
+ return tp;
+
+ /* Military time zones. */
+ if (wordlen == 1)
+ for (tp = military_table; tp->name; tp++)
+ if (word[0] == tp->name[0])
+ return tp;
+
+ /* Drop out any periods and try the time zone table again. */
+ for (period_found = false, p = q = word; (*p = *q); q++)
+ if (*q == '.')
+ period_found = true;
+ else
+ p++;
+ if (period_found && (tp = lookup_zone (pc, word)))
+ return tp;
+
+ return NULL;
+}
+
+static int
+yylex (union YYSTYPE *lvalp, parser_control *pc)
+{
+ unsigned char c;
+
+ for (;;)
+ {
+ while (c = *pc->input, c_isspace (c))
+ pc->input++;
+
+ if (c_isdigit (c) || c == '-' || c == '+')
+ {
+ char const *p = pc->input;
+ int sign;
+ if (c == '-' || c == '+')
+ {
+ sign = c == '-' ? -1 : 1;
+ while (c = *(pc->input = ++p), c_isspace (c))
+ continue;
+ if (! c_isdigit (c))
+ /* skip the '-' sign */
+ continue;
+ }
+ else
+ sign = 0;
+
+ time_t value = 0;
+ do
+ {
+ if (INT_MULTIPLY_WRAPV (value, 10, &value))
+ return '?';
+ if (INT_ADD_WRAPV (value, sign < 0 ? '0' - c : c - '0', &value))
+ return '?';
+ c = *++p;
+ }
+ while (c_isdigit (c));
+
+ if ((c == '.' || c == ',') && c_isdigit (p[1]))
+ {
+ time_t s = value;
+ int digits;
+
+ /* Accumulate fraction, to ns precision. */
+ p++;
+ int ns = *p++ - '0';
+ for (digits = 2; digits <= LOG10_BILLION; digits++)
+ {
+ ns *= 10;
+ if (c_isdigit (*p))
+ ns += *p++ - '0';
+ }
+
+ /* Skip excess digits, truncating toward -Infinity. */
+ if (sign < 0)
+ for (; c_isdigit (*p); p++)
+ if (*p != '0')
+ {
+ ns++;
+ break;
+ }
+ while (c_isdigit (*p))
+ p++;
+
+ /* Adjust to the timespec convention, which is that
+ tv_nsec is always a positive offset even if tv_sec is
+ negative. */
+ if (sign < 0 && ns)
+ {
+ if (INT_SUBTRACT_WRAPV (s, 1, &s))
+ return '?';
+ ns = BILLION - ns;
+ }
+
+ lvalp->timespec.tv_sec = s;
+ lvalp->timespec.tv_nsec = ns;
+ pc->input = p;
+ return sign ? tSDECIMAL_NUMBER : tUDECIMAL_NUMBER;
+ }
+ else
+ {
+ lvalp->textintval.negative = sign < 0;
+ lvalp->textintval.value = value;
+ lvalp->textintval.digits = p - pc->input;
+ pc->input = p;
+ return sign ? tSNUMBER : tUNUMBER;
+ }
+ }
+
+ if (c_isalpha (c))
+ {
+ char buff[20];
+ char *p = buff;
+ table const *tp;
+
+ do
+ {
+ if (p < buff + sizeof buff - 1)
+ *p++ = c;
+ c = *++pc->input;
+ }
+ while (c_isalpha (c) || c == '.');
+
+ *p = '\0';
+ tp = lookup_word (pc, buff);
+ if (! tp)
+ {
+ if (debugging (pc))
+ dbg_printf (_("error: unknown word '%s'\n"), buff);
+ return '?';
+ }
+ lvalp->intval = tp->value;
+ return tp->type;
+ }
+
+ if (c != '(')
+ return to_uchar (*pc->input++);
+
+ idx_t count = 0;
+ do
+ {
+ c = *pc->input++;
+ if (c == '\0')
+ return c;
+ if (c == '(')
+ count++;
+ else if (c == ')')
+ count--;
+ }
+ while (count != 0);
+ }
+}
+
+/* Do nothing if the parser reports an error. */
+static int
+yyerror (parser_control const *pc _GL_UNUSED,
+ char const *s _GL_UNUSED)
+{
+ return 0;
+}
+
+/* If *TM0 is the old and *TM1 is the new value of a struct tm after
+ passing it to mktime_z, return true if it's OK. It's not OK if
+ mktime failed or if *TM0 has out-of-range mainline members.
+ The caller should set TM1->tm_wday to -1 before calling mktime,
+ as a negative tm_wday is how mktime failure is inferred. */
+
+static bool
+mktime_ok (struct tm const *tm0, struct tm const *tm1)
+{
+ if (tm1->tm_wday < 0)
+ return false;
+
+ return ! ((tm0->tm_sec ^ tm1->tm_sec)
+ | (tm0->tm_min ^ tm1->tm_min)
+ | (tm0->tm_hour ^ tm1->tm_hour)
+ | (tm0->tm_mday ^ tm1->tm_mday)
+ | (tm0->tm_mon ^ tm1->tm_mon)
+ | (tm0->tm_year ^ tm1->tm_year));
+}
+
+/* Debugging: format a 'struct tm' into a buffer, taking the parser's
+ timezone information into account (if pc != NULL). */
+static char const *
+debug_strfdatetime (struct tm const *tm, parser_control const *pc,
+ char *buf, int n)
+{
+ /* TODO:
+ 1. find an optimal way to print date string in a clear and unambiguous
+ format. Currently, always add '(Y-M-D)' prefix.
+ Consider '2016y01m10d' or 'year(2016) month(01) day(10)'.
+
+ If the user needs debug printing, it means he/she already having
+ issues with the parsing - better to avoid formats that could
+ be mis-interpreted (e.g., just YYYY-MM-DD).
+
+ 2. Can strftime be used instead?
+ depends if it is portable and can print invalid dates on all systems.
+
+ 3. Print timezone information ?
+
+ 4. Print DST information ?
+
+ 5. Print nanosecond information ?
+
+ NOTE:
+ Printed date/time values might not be valid, e.g., '2016-02-31'
+ or '2016-19-2016' . These are the values as parsed from the user
+ string, before validation.
+ */
+ int m = nstrftime (buf, n, "(Y-M-D) %Y-%m-%d %H:%M:%S", tm, 0, 0);
+
+ /* If parser_control information was provided (for timezone),
+ and there's enough space in the buffer, add timezone info. */
+ if (pc && m < n && pc->zones_seen)
+ {
+ int tz = pc->time_zone;
+
+ /* Account for DST if tLOCAL_ZONE was seen. */
+ if (pc->local_zones_seen && !pc->zones_seen && 0 < pc->local_isdst)
+ tz += 60 * 60;
+
+ char time_zone_buf[TIME_ZONE_BUFSIZE];
+ snprintf (&buf[m], n - m, " TZ=%s", time_zone_str (tz, time_zone_buf));
+ }
+ return buf;
+}
+
+static char const *
+debug_strfdate (struct tm const *tm, char *buf, int n)
+{
+ char tm_year_buf[TM_YEAR_BUFSIZE];
+ snprintf (buf, n, "(Y-M-D) %s-%02d-%02d",
+ tm_year_str (tm->tm_year, tm_year_buf),
+ tm->tm_mon + 1, tm->tm_mday);
+ return buf;
+}
+
+static char const *
+debug_strftime (struct tm const *tm, char *buf, int n)
+{
+ snprintf (buf, n, "%02d:%02d:%02d", tm->tm_hour, tm->tm_min, tm->tm_sec);
+ return buf;
+}
+
+/* If mktime_ok failed, display the failed time values,
+ and provide possible hints. Example output:
+
+ date: error: invalid date/time value:
+ date: user provided time: '(Y-M-D) 2006-04-02 02:45:00'
+ date: normalized time: '(Y-M-D) 2006-04-02 03:45:00'
+ date: __
+ date: possible reasons:
+ date: non-existing due to daylight-saving time;
+ date: numeric values overflow;
+ date: missing timezone;
+ */
+static void
+debug_mktime_not_ok (struct tm const *tm0, struct tm const *tm1,
+ parser_control const *pc, bool time_zone_seen)
+{
+ /* TODO: handle t == -1 (as in 'mktime_ok'). */
+ char tmp[DBGBUFSIZE];
+ int i;
+ const bool eq_sec = (tm0->tm_sec == tm1->tm_sec);
+ const bool eq_min = (tm0->tm_min == tm1->tm_min);
+ const bool eq_hour = (tm0->tm_hour == tm1->tm_hour);
+ const bool eq_mday = (tm0->tm_mday == tm1->tm_mday);
+ const bool eq_month = (tm0->tm_mon == tm1->tm_mon);
+ const bool eq_year = (tm0->tm_year == tm1->tm_year);
+
+ const bool dst_shift = eq_sec && eq_min && !eq_hour
+ && eq_mday && eq_month && eq_year;
+
+ if (!debugging (pc))
+ return;
+
+ dbg_printf (_("error: invalid date/time value:\n"));
+ dbg_printf (_(" user provided time: '%s'\n"),
+ debug_strfdatetime (tm0, pc, tmp, sizeof tmp));
+ dbg_printf (_(" normalized time: '%s'\n"),
+ debug_strfdatetime (tm1, pc, tmp, sizeof tmp));
+ /* The format must be aligned with debug_strfdatetime and the two
+ DEBUG statements above. This string is not translated. */
+ i = snprintf (tmp, sizeof tmp,
+ " %4s %2s %2s %2s %2s %2s",
+ eq_year ? "" : "----",
+ eq_month ? "" : "--",
+ eq_mday ? "" : "--",
+ eq_hour ? "" : "--",
+ eq_min ? "" : "--",
+ eq_sec ? "" : "--");
+ /* Trim trailing whitespace. */
+ if (0 <= i)
+ {
+ if (sizeof tmp - 1 < i)
+ i = sizeof tmp - 1;
+ while (0 < i && tmp[i - 1] == ' ')
+ --i;
+ tmp[i] = '\0';
+ }
+ dbg_printf ("%s\n", tmp);
+
+ dbg_printf (_(" possible reasons:\n"));
+ if (dst_shift)
+ dbg_printf (_(" non-existing due to daylight-saving time;\n"));
+ if (!eq_mday && !eq_month)
+ dbg_printf (_(" invalid day/month combination;\n"));
+ dbg_printf (_(" numeric values overflow;\n"));
+ dbg_printf (" %s\n", (time_zone_seen ? _("incorrect timezone")
+ : _("missing timezone")));
+}
+
+/* Parse a date/time string, storing the resulting time value into *RESULT.
+ The string itself is pointed to by P. Return true if successful.
+ P can be an incomplete or relative time specification; if so, use
+ *NOW as the basis for the returned time. Default to timezone
+ TZDEFAULT, which corresponds to tzalloc (TZSTRING). */
+static bool
+parse_datetime_body (struct timespec *result, char const *p,
+ struct timespec const *now, unsigned int flags,
+ timezone_t tzdefault, char const *tzstring)
+{
+ struct tm tm;
+ struct tm tm0;
+ char time_zone_buf[TIME_ZONE_BUFSIZE];
+ char dbg_tm[DBGBUFSIZE];
+ bool ok = false;
+ char const *input_sentinel = p + strlen (p);
+ char *tz1alloc = NULL;
+
+ /* A reasonable upper bound for the size of ordinary TZ strings.
+ Use heap allocation if TZ's length exceeds this. */
+ enum { TZBUFSIZE = 100 };
+ char tz1buf[TZBUFSIZE];
+
+ struct timespec gettime_buffer;
+ if (! now)
+ {
+ gettime (&gettime_buffer);
+ now = &gettime_buffer;
+ }
+
+ time_t Start = now->tv_sec;
+ int Start_ns = now->tv_nsec;
+
+ unsigned char c;
+ while (c = *p, c_isspace (c))
+ p++;
+
+ timezone_t tz = tzdefault;
+
+ /* Store a local copy prior to first "goto". Without this, a prior use
+ below of RELATIVE_TIME_0 on the RHS might translate to an assignment-
+ to-temporary, which would trigger a -Wjump-misses-init warning. */
+ const relative_time rel_time_0 = RELATIVE_TIME_0;
+
+ if (strncmp (p, "TZ=\"", 4) == 0)
+ {
+ char const *tzbase = p + 4;
+ idx_t tzsize = 1;
+ char const *s;
+
+ for (s = tzbase; *s; s++, tzsize++)
+ if (*s == '\\')
+ {
+ s++;
+ if (! (*s == '\\' || *s == '"'))
+ break;
+ }
+ else if (*s == '"')
+ {
+ timezone_t tz1;
+ char *tz1string = tz1buf;
+ char *z;
+ if (TZBUFSIZE < tzsize)
+ {
+ tz1alloc = malloc (tzsize);
+ if (!tz1alloc)
+ goto fail;
+ tz1string = tz1alloc;
+ }
+ z = tz1string;
+ for (s = tzbase; *s != '"'; s++)
+ *z++ = *(s += *s == '\\');
+ *z = '\0';
+ tz1 = tzalloc (tz1string);
+ if (!tz1)
+ goto fail;
+ tz = tz1;
+ tzstring = tz1string;
+
+ p = s + 1;
+ while (c = *p, c_isspace (c))
+ p++;
+
+ break;
+ }
+ }
+
+ struct tm tmp;
+ if (! localtime_rz (tz, &now->tv_sec, &tmp))
+ goto fail;
+
+ /* As documented, be careful to treat the empty string just like
+ a date string of "0". Without this, an empty string would be
+ declared invalid when parsed during a DST transition. */
+ if (*p == '\0')
+ p = "0";
+
+ parser_control pc;
+ pc.input = p;
+#ifdef GNULIB_PARSE_DATETIME2
+ pc.parse_datetime_debug = (flags & PARSE_DATETIME_DEBUG) != 0;
+#endif
+ if (INT_ADD_WRAPV (tmp.tm_year, TM_YEAR_BASE, &pc.year.value))
+ {
+ if (debugging (&pc))
+ dbg_printf (_("error: initial year out of range\n"));
+ goto fail;
+ }
+ pc.year.digits = 0;
+ pc.month = tmp.tm_mon + 1;
+ pc.day = tmp.tm_mday;
+ pc.hour = tmp.tm_hour;
+ pc.minutes = tmp.tm_min;
+ pc.seconds.tv_sec = tmp.tm_sec;
+ pc.seconds.tv_nsec = Start_ns;
+ tm.tm_isdst = tmp.tm_isdst;
+
+ pc.meridian = MER24;
+ pc.rel = rel_time_0;
+ pc.timespec_seen = false;
+ pc.rels_seen = false;
+ pc.dates_seen = 0;
+ pc.days_seen = 0;
+ pc.times_seen = 0;
+ pc.local_zones_seen = 0;
+ pc.dsts_seen = 0;
+ pc.zones_seen = 0;
+ pc.year_seen = false;
+ pc.debug_dates_seen = false;
+ pc.debug_days_seen = false;
+ pc.debug_times_seen = false;
+ pc.debug_local_zones_seen = false;
+ pc.debug_zones_seen = false;
+ pc.debug_year_seen = false;
+ pc.debug_ordinal_day_seen = false;
+
+#if HAVE_STRUCT_TM_TM_ZONE
+ pc.local_time_zone_table[0].name = tmp.tm_zone;
+ pc.local_time_zone_table[0].type = tLOCAL_ZONE;
+ pc.local_time_zone_table[0].value = tmp.tm_isdst;
+ pc.local_time_zone_table[1].name = NULL;
+
+ /* Probe the names used in the next three calendar quarters, looking
+ for a tm_isdst different from the one we already have. */
+ {
+ int quarter;
+ for (quarter = 1; quarter <= 3; quarter++)
+ {
+ time_t probe;
+ if (INT_ADD_WRAPV (Start, quarter * (90 * 24 * 60 * 60), &probe))
+ break;
+ struct tm probe_tm;
+ if (localtime_rz (tz, &probe, &probe_tm) && probe_tm.tm_zone
+ && probe_tm.tm_isdst != pc.local_time_zone_table[0].value)
+ {
+ {
+ pc.local_time_zone_table[1].name = probe_tm.tm_zone;
+ pc.local_time_zone_table[1].type = tLOCAL_ZONE;
+ pc.local_time_zone_table[1].value = probe_tm.tm_isdst;
+ pc.local_time_zone_table[2].name = NULL;
+ }
+ break;
+ }
+ }
+ }
+#else
+#if HAVE_TZNAME
+ {
+# if !HAVE_DECL_TZNAME
+ extern char *tzname[];
+# endif
+ int i;
+ for (i = 0; i < 2; i++)
+ {
+ pc.local_time_zone_table[i].name = tzname[i];
+ pc.local_time_zone_table[i].type = tLOCAL_ZONE;
+ pc.local_time_zone_table[i].value = i;
+ }
+ pc.local_time_zone_table[i].name = NULL;
+ }
+#else
+ pc.local_time_zone_table[0].name = NULL;
+#endif
+#endif
+
+ if (pc.local_time_zone_table[0].name && pc.local_time_zone_table[1].name
+ && ! strcmp (pc.local_time_zone_table[0].name,
+ pc.local_time_zone_table[1].name))
+ {
+ /* This locale uses the same abbreviation for standard and
+ daylight times. So if we see that abbreviation, we don't
+ know whether it's daylight time. */
+ pc.local_time_zone_table[0].value = -1;
+ pc.local_time_zone_table[1].name = NULL;
+ }
+
+ if (yyparse (&pc) != 0)
+ {
+ if (debugging (&pc))
+ dbg_printf ((input_sentinel <= pc.input
+ ? _("error: parsing failed\n")
+ : _("error: parsing failed, stopped at '%s'\n")),
+ pc.input);
+ goto fail;
+ }
+
+
+ /* Determine effective timezone source. */
+
+ if (debugging (&pc))
+ {
+ dbg_printf (_("input timezone: "));
+
+ if (pc.timespec_seen)
+ fprintf (stderr, _("'@timespec' - always UTC"));
+ else if (pc.zones_seen)
+ fprintf (stderr, _("parsed date/time string"));
+ else if (tzstring)
+ {
+ if (tz != tzdefault)
+ fprintf (stderr, _("TZ=\"%s\" in date string"), tzstring);
+ else if (STREQ (tzstring, "UTC0"))
+ {
+ /* Special case: 'date -u' sets TZ="UTC0". */
+ fprintf (stderr, _("TZ=\"UTC0\" environment value or -u"));
+ }
+ else
+ fprintf (stderr, _("TZ=\"%s\" environment value"), tzstring);
+ }
+ else
+ fprintf (stderr, _("system default"));
+
+ /* Account for DST changes if tLOCAL_ZONE was seen.
+ local timezone only changes DST and is relative to the
+ default timezone.*/
+ if (pc.local_zones_seen && !pc.zones_seen && 0 < pc.local_isdst)
+ fprintf (stderr, ", dst");
+
+ if (pc.zones_seen)
+ fprintf (stderr, " (%s)", time_zone_str (pc.time_zone, time_zone_buf));
+
+ fputc ('\n', stderr);
+ }
+
+ if (pc.timespec_seen)
+ *result = pc.seconds;
+ else
+ {
+ if (1 < (pc.times_seen | pc.dates_seen | pc.days_seen | pc.dsts_seen
+ | (pc.local_zones_seen + pc.zones_seen)))
+ {
+ if (debugging (&pc))
+ {
+ if (pc.times_seen > 1)
+ dbg_printf ("error: seen multiple time parts\n");
+ if (pc.dates_seen > 1)
+ dbg_printf ("error: seen multiple date parts\n");
+ if (pc.days_seen > 1)
+ dbg_printf ("error: seen multiple days parts\n");
+ if (pc.dsts_seen > 1)
+ dbg_printf ("error: seen multiple daylight-saving parts\n");
+ if ((pc.local_zones_seen + pc.zones_seen) > 1)
+ dbg_printf ("error: seen multiple time-zone parts\n");
+ }
+ goto fail;
+ }
+
+ if (! to_tm_year (pc.year, debugging (&pc), &tm.tm_year)
+ || INT_ADD_WRAPV (pc.month, -1, &tm.tm_mon)
+ || INT_ADD_WRAPV (pc.day, 0, &tm.tm_mday))
+ {
+ if (debugging (&pc))
+ dbg_printf (_("error: year, month, or day overflow\n"));
+ goto fail;
+ }
+ if (pc.times_seen || (pc.rels_seen && ! pc.dates_seen && ! pc.days_seen))
+ {
+ tm.tm_hour = to_hour (pc.hour, pc.meridian);
+ if (tm.tm_hour < 0)
+ {
+ char const *mrd = (pc.meridian == MERam ? "am"
+ : pc.meridian == MERpm ?"pm" : "");
+ if (debugging (&pc))
+ dbg_printf (_("error: invalid hour %"PRIdMAX"%s\n"),
+ pc.hour, mrd);
+ goto fail;
+ }
+ tm.tm_min = pc.minutes;
+ tm.tm_sec = pc.seconds.tv_sec;
+ if (debugging (&pc))
+ dbg_printf ((pc.times_seen
+ ? _("using specified time as starting value: '%s'\n")
+ : _("using current time as starting value: '%s'\n")),
+ debug_strftime (&tm, dbg_tm, sizeof dbg_tm));
+ }
+ else
+ {
+ tm.tm_hour = tm.tm_min = tm.tm_sec = 0;
+ pc.seconds.tv_nsec = 0;
+ if (debugging (&pc))
+ dbg_printf ("warning: using midnight as starting time: 00:00:00\n");
+ }
+
+ /* Let mktime deduce tm_isdst if we have an absolute timestamp. */
+ if (pc.dates_seen | pc.days_seen | pc.times_seen)
+ tm.tm_isdst = -1;
+
+ /* But if the input explicitly specifies local time with or without
+ DST, give mktime that information. */
+ if (pc.local_zones_seen)
+ tm.tm_isdst = pc.local_isdst;
+
+ tm0.tm_sec = tm.tm_sec;
+ tm0.tm_min = tm.tm_min;
+ tm0.tm_hour = tm.tm_hour;
+ tm0.tm_mday = tm.tm_mday;
+ tm0.tm_mon = tm.tm_mon;
+ tm0.tm_year = tm.tm_year;
+ tm0.tm_isdst = tm.tm_isdst;
+ tm.tm_wday = -1;
+
+ Start = mktime_z (tz, &tm);
+
+ if (! mktime_ok (&tm0, &tm))
+ {
+ bool repaired = false;
+ bool time_zone_seen = pc.zones_seen != 0;
+ if (time_zone_seen)
+ {
+ /* Guard against falsely reporting errors near the time_t
+ boundaries when parsing times in other time zones. For
+ example, suppose the input string "1969-12-31 23:00:00 -0100",
+ the current time zone is 8 hours ahead of UTC, and the min
+ time_t value is 1970-01-01 00:00:00 UTC. Then the min
+ localtime value is 1970-01-01 08:00:00, and mktime will
+ therefore fail on 1969-12-31 23:00:00. To work around the
+ problem, set the time zone to 1 hour behind UTC temporarily
+ by setting TZ="XXX1:00" and try mktime again. */
+
+ char tz2buf[sizeof "XXX" - 1 + TIME_ZONE_BUFSIZE];
+ tz2buf[0] = tz2buf[1] = tz2buf[2] = 'X';
+ time_zone_str (pc.time_zone, &tz2buf[3]);
+ timezone_t tz2 = tzalloc (tz2buf);
+ if (!tz2)
+ {
+ if (debugging (&pc))
+ dbg_printf (_("error: tzalloc (\"%s\") failed\n"), tz2buf);
+ goto fail;
+ }
+ tm.tm_sec = tm0.tm_sec;
+ tm.tm_min = tm0.tm_min;
+ tm.tm_hour = tm0.tm_hour;
+ tm.tm_mday = tm0.tm_mday;
+ tm.tm_mon = tm0.tm_mon;
+ tm.tm_year = tm0.tm_year;
+ tm.tm_isdst = tm0.tm_isdst;
+ tm.tm_wday = -1;
+ Start = mktime_z (tz2, &tm);
+ repaired = mktime_ok (&tm0, &tm);
+ tzfree (tz2);
+ }
+
+ if (! repaired)
+ {
+ debug_mktime_not_ok (&tm0, &tm, &pc, time_zone_seen);
+ goto fail;
+ }
+ }
+
+ char dbg_ord[DBGBUFSIZE];
+
+ if (pc.days_seen && ! pc.dates_seen)
+ {
+ intmax_t dayincr;
+ if (INT_MULTIPLY_WRAPV ((pc.day_ordinal
+ - (0 < pc.day_ordinal
+ && tm.tm_wday != pc.day_number)),
+ 7, &dayincr)
+ || INT_ADD_WRAPV ((pc.day_number - tm.tm_wday + 7) % 7,
+ dayincr, &dayincr)
+ || INT_ADD_WRAPV (dayincr, tm.tm_mday, &tm.tm_mday))
+ Start = -1;
+ else
+ {
+ tm.tm_isdst = -1;
+ Start = mktime_z (tz, &tm);
+ }
+
+ if (Start == (time_t) -1)
+ {
+ if (debugging (&pc))
+ dbg_printf (_("error: day '%s' "
+ "(day ordinal=%"PRIdMAX" number=%d) "
+ "resulted in an invalid date: '%s'\n"),
+ str_days (&pc, dbg_ord, sizeof dbg_ord),
+ pc.day_ordinal, pc.day_number,
+ debug_strfdatetime (&tm, &pc, dbg_tm,
+ sizeof dbg_tm));
+ goto fail;
+ }
+
+ if (debugging (&pc))
+ dbg_printf (_("new start date: '%s' is '%s'\n"),
+ str_days (&pc, dbg_ord, sizeof dbg_ord),
+ debug_strfdatetime (&tm, &pc, dbg_tm, sizeof dbg_tm));
+
+ }
+
+ if (debugging (&pc))
+ {
+ if (!pc.dates_seen && !pc.days_seen)
+ dbg_printf (_("using current date as starting value: '%s'\n"),
+ debug_strfdate (&tm, dbg_tm, sizeof dbg_tm));
+
+ if (pc.days_seen && pc.dates_seen)
+ dbg_printf (_("warning: day (%s) ignored when explicit dates "
+ "are given\n"),
+ str_days (&pc, dbg_ord, sizeof dbg_ord));
+
+ dbg_printf (_("starting date/time: '%s'\n"),
+ debug_strfdatetime (&tm, &pc, dbg_tm, sizeof dbg_tm));
+ }
+
+ /* Add relative date. */
+ if (pc.rel.year | pc.rel.month | pc.rel.day)
+ {
+ if (debugging (&pc))
+ {
+ if ((pc.rel.year != 0 || pc.rel.month != 0) && tm.tm_mday != 15)
+ dbg_printf (_("warning: when adding relative months/years, "
+ "it is recommended to specify the 15th of the "
+ "months\n"));
+
+ if (pc.rel.day != 0 && tm.tm_hour != 12)
+ dbg_printf (_("warning: when adding relative days, "
+ "it is recommended to specify noon\n"));
+ }
+
+ int year, month, day;
+ if (INT_ADD_WRAPV (tm.tm_year, pc.rel.year, &year)
+ || INT_ADD_WRAPV (tm.tm_mon, pc.rel.month, &month)
+ || INT_ADD_WRAPV (tm.tm_mday, pc.rel.day, &day))
+ {
+ if (debugging (&pc))
+ dbg_printf (_("error: %s:%d\n"), __FILE__, __LINE__);
+ goto fail;
+ }
+ tm.tm_year = year;
+ tm.tm_mon = month;
+ tm.tm_mday = day;
+ tm.tm_hour = tm0.tm_hour;
+ tm.tm_min = tm0.tm_min;
+ tm.tm_sec = tm0.tm_sec;
+ tm.tm_isdst = tm0.tm_isdst;
+ Start = mktime_z (tz, &tm);
+ if (Start == (time_t) -1)
+ {
+ if (debugging (&pc))
+ dbg_printf (_("error: adding relative date resulted "
+ "in an invalid date: '%s'\n"),
+ debug_strfdatetime (&tm, &pc, dbg_tm,
+ sizeof dbg_tm));
+ goto fail;
+ }
+
+ if (debugging (&pc))
+ {
+ dbg_printf (_("after date adjustment "
+ "(%+"PRIdMAX" years, %+"PRIdMAX" months, "
+ "%+"PRIdMAX" days),\n"),
+ pc.rel.year, pc.rel.month, pc.rel.day);
+ dbg_printf (_(" new date/time = '%s'\n"),
+ debug_strfdatetime (&tm, &pc, dbg_tm,
+ sizeof dbg_tm));
+
+ /* Warn about crossing DST due to time adjustment.
+ Example: https://bugs.gnu.org/8357
+ env TZ=Europe/Helsinki \
+ date --debug \
+ -d 'Mon Mar 28 00:36:07 2011 EEST 1 day ago'
+
+ This case is different than DST changes due to time adjustment,
+ i.e., "1 day ago" vs "24 hours ago" are calculated in different
+ places.
+
+ 'tm0.tm_isdst' contains the DST of the input date,
+ 'tm.tm_isdst' is the normalized result after calling
+ mktime (&tm).
+ */
+ if (tm0.tm_isdst != -1 && tm.tm_isdst != tm0.tm_isdst)
+ dbg_printf (_("warning: daylight saving time changed after "
+ "date adjustment\n"));
+
+ /* Warn if the user did not ask to adjust days but mday changed,
+ or
+ user did not ask to adjust months/days but the month changed.
+
+ Example for first case:
+ 2016-05-31 + 1 month => 2016-06-31 => 2016-07-01.
+ User asked to adjust month, but the day changed from 31 to 01.
+
+ Example for second case:
+ 2016-02-29 + 1 year => 2017-02-29 => 2017-03-01.
+ User asked to adjust year, but the month changed from 02 to 03.
+ */
+ if (pc.rel.day == 0
+ && (tm.tm_mday != day
+ || (pc.rel.month == 0 && tm.tm_mon != month)))
+ {
+ dbg_printf (_("warning: month/year adjustment resulted in "
+ "shifted dates:\n"));
+ char tm_year_buf[TM_YEAR_BUFSIZE];
+ dbg_printf (_(" adjusted Y M D: %s %02d %02d\n"),
+ tm_year_str (year, tm_year_buf), month + 1, day);
+ dbg_printf (_(" normalized Y M D: %s %02d %02d\n"),
+ tm_year_str (tm.tm_year, tm_year_buf),
+ tm.tm_mon + 1, tm.tm_mday);
+ }
+ }
+
+ }
+
+ /* The only "output" of this if-block is an updated Start value,
+ so this block must follow others that clobber Start. */
+ if (pc.zones_seen)
+ {
+ bool overflow = false;
+#ifdef HAVE_TM_GMTOFF
+ long int utcoff = tm.tm_gmtoff;
+#else
+ time_t t = Start;
+ struct tm gmt;
+ int utcoff = (gmtime_r (&t, &gmt)
+ ? tm_diff (&tm, &gmt)
+ : (overflow = true, 0));
+#endif
+ intmax_t delta;
+ overflow |= INT_SUBTRACT_WRAPV (pc.time_zone, utcoff, &delta);
+ time_t t1;
+ overflow |= INT_SUBTRACT_WRAPV (Start, delta, &t1);
+ if (overflow)
+ {
+ if (debugging (&pc))
+ dbg_printf (_("error: timezone %d caused time_t overflow\n"),
+ pc.time_zone);
+ goto fail;
+ }
+ Start = t1;
+ }
+
+ if (debugging (&pc))
+ {
+ intmax_t Starti = Start;
+ dbg_printf (_("'%s' = %"PRIdMAX" epoch-seconds\n"),
+ debug_strfdatetime (&tm, &pc, dbg_tm, sizeof dbg_tm),
+ Starti);
+ }
+
+
+ /* Add relative hours, minutes, and seconds. On hosts that support
+ leap seconds, ignore the possibility of leap seconds; e.g.,
+ "+ 10 minutes" adds 600 seconds, even if one of them is a
+ leap second. Typically this is not what the user wants, but it's
+ too hard to do it the other way, because the time zone indicator
+ must be applied before relative times, and if mktime is applied
+ again the time zone will be lost. */
+ {
+ intmax_t orig_ns = pc.seconds.tv_nsec;
+ intmax_t sum_ns = orig_ns + pc.rel.ns;
+ int normalized_ns = (sum_ns % BILLION + BILLION) % BILLION;
+ int d4 = (sum_ns - normalized_ns) / BILLION;
+ intmax_t d1, t1, d2, t2, t3;
+ time_t t4;
+ if (INT_MULTIPLY_WRAPV (pc.rel.hour, 60 * 60, &d1)
+ || INT_ADD_WRAPV (Start, d1, &t1)
+ || INT_MULTIPLY_WRAPV (pc.rel.minutes, 60, &d2)
+ || INT_ADD_WRAPV (t1, d2, &t2)
+ || INT_ADD_WRAPV (t2, pc.rel.seconds, &t3)
+ || INT_ADD_WRAPV (t3, d4, &t4))
+ {
+ if (debugging (&pc))
+ dbg_printf (_("error: adding relative time caused an "
+ "overflow\n"));
+ goto fail;
+ }
+
+ result->tv_sec = t4;
+ result->tv_nsec = normalized_ns;
+
+ if (debugging (&pc)
+ && (pc.rel.hour | pc.rel.minutes | pc.rel.seconds | pc.rel.ns))
+ {
+ dbg_printf (_("after time adjustment (%+"PRIdMAX" hours, "
+ "%+"PRIdMAX" minutes, "
+ "%+"PRIdMAX" seconds, %+d ns),\n"),
+ pc.rel.hour, pc.rel.minutes, pc.rel.seconds,
+ pc.rel.ns);
+ intmax_t t4i = t4;
+ dbg_printf (_(" new time = %"PRIdMAX" epoch-seconds\n"), t4i);
+
+ /* Warn about crossing DST due to time adjustment.
+ Example: https://bugs.gnu.org/8357
+ env TZ=Europe/Helsinki \
+ date --debug \
+ -d 'Mon Mar 28 00:36:07 2011 EEST 24 hours ago'
+
+ This case is different than DST changes due to days adjustment,
+ i.e., "1 day ago" vs "24 hours ago" are calculated in different
+ places.
+
+ 'tm.tm_isdst' contains the date after date adjustment. */
+ struct tm lmt;
+ if (tm.tm_isdst != -1 && localtime_rz (tz, &result->tv_sec, &lmt)
+ && tm.tm_isdst != lmt.tm_isdst)
+ dbg_printf (_("warning: daylight saving time changed after "
+ "time adjustment\n"));
+ }
+ }
+ }
+
+ if (debugging (&pc))
+ {
+ /* Special case: using 'date -u' simply set TZ=UTC0 */
+ if (! tzstring)
+ dbg_printf (_("timezone: system default\n"));
+ else if (STREQ (tzstring, "UTC0"))
+ dbg_printf (_("timezone: Universal Time\n"));
+ else
+ dbg_printf (_("timezone: TZ=\"%s\" environment value\n"), tzstring);
+
+ intmax_t sec = result->tv_sec;
+ int nsec = result->tv_nsec;
+ dbg_printf (_("final: %"PRIdMAX".%09d (epoch-seconds)\n"),
+ sec, nsec);
+
+ struct tm gmt, lmt;
+ bool got_utc = !!gmtime_r (&result->tv_sec, &gmt);
+ if (got_utc)
+ dbg_printf (_("final: %s (UTC)\n"),
+ debug_strfdatetime (&gmt, NULL,
+ dbg_tm, sizeof dbg_tm));
+ if (localtime_rz (tz, &result->tv_sec, &lmt))
+ {
+#ifdef HAVE_TM_GMTOFF
+ bool got_utcoff = true;
+ long int utcoff = lmt.tm_gmtoff;
+#else
+ bool got_utcoff = got_utc;
+ int utcoff;
+ if (got_utcoff)
+ utcoff = tm_diff (&lmt, &gmt);
+#endif
+ if (got_utcoff)
+ dbg_printf (_("final: %s (UTC%s)\n"),
+ debug_strfdatetime (&lmt, NULL, dbg_tm, sizeof dbg_tm),
+ time_zone_str (utcoff, time_zone_buf));
+ else
+ dbg_printf (_("final: %s (unknown time zone offset)\n"),
+ debug_strfdatetime (&lmt, NULL, dbg_tm, sizeof dbg_tm));
+ }
+ }
+
+ ok = true;
+
+ fail:
+ if (tz != tzdefault)
+ tzfree (tz);
+ free (tz1alloc);
+ return ok;
+}
+
+#ifdef GNULIB_PARSE_DATETIME2
+/* Parse a date/time string, storing the resulting time value into *RESULT.
+ The string itself is pointed to by P. Return true if successful.
+ P can be an incomplete or relative time specification; if so, use
+ *NOW as the basis for the returned time. Default to timezone
+ TZDEFAULT, which corresponds to tzalloc (TZSTRING). */
+bool
+parse_datetime2 (struct timespec *result, char const *p,
+ struct timespec const *now, unsigned int flags,
+ timezone_t tzdefault, char const *tzstring)
+{
+ return parse_datetime_body (result, p, now, flags, tzdefault, tzstring);
+}
+#endif
+
+
+/* The plain interface: run with debug=false and the default timezone. */
+bool
+parse_datetime (struct timespec *result, char const *p,
+ struct timespec const *now)
+{
+ char const *tzstring = getenv ("TZ");
+ timezone_t tz = tzalloc (tzstring);
+ if (!tz)
+ return false;
+ bool ok = parse_datetime_body (result, p, now, 0, tz, tzstring);
+ tzfree (tz);
+ return ok;
+}
+
+#if TEST
+
+int
+main (int ac, char **av)
+{
+ char buff[BUFSIZ];
+
+ printf ("Enter date, or blank line to exit.\n\t> ");
+ fflush (stdout);
+
+ buff[BUFSIZ - 1] = '\0';
+ while (fgets (buff, BUFSIZ - 1, stdin) && buff[0])
+ {
+ struct timespec d;
+ struct tm const *tm;
+ if (! parse_datetime (&d, buff, NULL))
+ printf ("Bad format - couldn't convert.\n");
+ else if (! (tm = localtime (&d.tv_sec)))
+ {
+ intmax_t sec = d.tv_sec;
+ printf ("localtime (%"PRIdMAX") failed\n", sec);
+ }
+ else
+ {
+ int ns = d.tv_nsec;
+ char tm_year_buf[TM_YEAR_BUFSIZE];
+ printf ("%s-%02d-%02d %02d:%02d:%02d.%09d\n",
+ tm_year_str (tm->tm_year, tm_year_buf),
+ tm->tm_mon + 1, tm->tm_mday,
+ tm->tm_hour, tm->tm_min, tm->tm_sec, ns);
+ }
+ printf ("\t> ");
+ fflush (stdout);
+ }
+ return 0;
+}
+#endif /* TEST */